I
I
1
r
i
a Group of Authorities
I
I
I
i
I i i
ROY L. HARRl NGTON
I
I
I
I
Engineering Technical Pepartmelit Newport News Shipbuilding and
I
Published by
THE S ~ C I E OF ~ Y NAVAL ARCHITECTS AND MARINE ENGINEERS
I I I
One World Trade Center, Suite 1369, New York, N.Y. 10048
i
.
@ Copyright 1971 by The Society of Naval Architects and Marine Engineers Library of Congress Catalog Card No. 78472362 Printed in the United States of America Second Printing 1976 TMrd Printing 1980
Since 1942 and 1944 when the two volumes of MARINEENGINEERING were published, the basic body of knowledge constituting marine engineering has greatly increased. Recognizing was substantially out of date, the Society in 1964 underthat the original MARINEENGINEERING took the task of compiling a reviged edition. That same year a Control Committee was ap-' pointed by the president to guide the revision, carrying on the objective of the original work, that of producing a comprehensive treatise reflecting the important technical progress of the last several decades. Also, the intent is that this text should complement the Society's two companion volumes, Principles of Naval Architecture and Ship Design and Construction, which deal similarly with the subjects of naval architecture and ship construction practices. When the task of revising the original MARINEENGINEERING was undertaken, it was quickly found to be considerably larger in scope than anticipated. The original text had to be completely rewritten, not simply revised. At the putset, it was decided that, for ready use and reference, the text should be a single volume limited to about 850 pages. Therefore, discussion of engineering subjects.covered in other textbooks had to be greatly abbreviated. Every effort has been made, however, throughout the text to make reference to appropriate source material for the individual or self-taught reader as well as the resourceful teacher (who may in some cases prefer to use his own references). Each chapter is written by a separate author (or authors). The committee felt that this precept should be continued because of the advantages of professional specialization it affords. Some unevenness in style results, but this has been minimized by the technical editor. In May, 1968, Mr. Roy L. Harrington was selected as technical editor by the committee. Mr. Hanington received a Society scholarship in 1960 to pursue an M.S. degree in marine engineering and also has had twelve years of technical ship design experience in a major shipyard. With this background, plus his extensive literary capability, he was considered well equipped to bridge the academic and the practicing professional points of view of the Society members. This book is not intended to be either a handbook or conversely a definitive text on any specific engineering discipline which may be used in marine engineering. Its purpose is to acquaint a person already familiar with basic engineering fundamentals with the various engineering disciplines and applications which constitute marine engineering. The need for such a book becomes apparent when it is recognized that many practicing marine engineers have had little formal education in the field of marine engineering as such, but instead have come into it from other related engineering activities. The Control Committee appointed to guide the revision of MARINEENGINEERING consisted of: Ernst G. Frankel Jens T. Holm William E. Jacobsen John R. Kane John H. Lsscaster ' Lauren S. McCready Andrew I. McKee Laskar Wechsler John B. Woodward I11 Robert E. Yohe
There have been so many technological advancements since the original MARINEENQINEERwas published that the'content of this book bears little similarity to the original text. For example, in a manner of speaking, a nuclear power chapter has been substituted f o ~the old reciprocating steam engine chapter, and other differences are almost as dramatic. However, the same basic philosophy was used in writing both works except that, insofar as practicable, this text covers naval practice in addition to merchant practice. In order to ensure that this book is comprehensive and factual, and accurately represents the consensus of opinion of the marine industry as a whole, the chapters were subjected to a series of reviews. After the manuscripts were prepared by the authors and reviewed within their respective organizations, they were then reviewed by the editor, Control Committee, and selected members of the marine industry who were experts in each particular area. The entire Sociehy owes a large debt of gratitude to this last group as they were largely responsible for transforming good manuscripts into excellent manuscripts. With few exceptions, it is a gross injustice to suggest that the chapters have been prepared by only the authors indicated. In several cases, the contributions of single individuals who assisted were almost as large as that of the author; and in $1 cases, the comments and discussion provided by the Control Committee and other members of the marine industry were an invaluable asset. Mr. John Markert (author of Chapter 19) accurately expressed the sentiment of the chapter authors when he stated that the generous cooperhtion and assistance received from the numerous contributors, often persons not acquainted with the author, were a revelation; it should, however, be noted that such cooperation is characteristic of the marine fraternity. An accurate listing of those who assisted in the preparation of this book would include many names. Several hundred people made direct contributions (by assisting in the preparation of manuscripts, supplying reference material, reviewing manuscripts, or supplying illustration material); and when those who made indirect contributions are added, the number of names would become even larger. - It is, however, considered proper to acknowledge some of the contributions as follows: Mr. Catlin (Chapter 3) acknowledges the valuable contributions of Mr. George W. Kessler, vice resident. Babcock & Wilcox. Mr. L. E. Triggs, chief engineer, Marine Dept., Combustion ~ n ~ i h e e r i~nc., n ~ , Mr. W. I. signell, chief marine engineer, J. J. Henry Co., Inc., and Professor J. T. Holm, Webb Institute of Naval Architecture, in the development of the chapter dealing with boilers and combustion. Dr. Illies (Chapter 8) states that he received help from a large number of individuals while preparing the low-speed directrcoupled diesel engine chapter. The material that was made available by diesel engine manufacturers (MAN, Fiat, Sulzer, Burmeister and Wain, Gotaverken, Stork, and Doxford) was particularly helpful as was the valuable advice and personal assistance that ww provided by Mr. Klaus Knaack. Mr. Semar (Chapter 9) acknowledges the contributions made by Mr. W. S. Richardson, the Falk Corporation, Mr. Norman A. Smith, General Electric Company, and Mr. Frederic A. Thoma, DeLaval Turbine, Inc., in the development of the chapter on reduction gears. Mr. J. F. Sebald (Chapter 13) acknowledges the valuable contributions made by Mr. P. D. Gold of the Worthington Corporation, Mr. William J. Bow of the Foster Wheeler Corporation and Mr. J. J. Biese of the Ingersoll Rand Co. h providing illustrations and for their critical review of the manuscript. The cooperation of the Heat Exchange Institute and The American Society of Mechanical Engineers in permitting the publication of technical data and the technical support provided by Gilbert Associates, Inc. are also gratefully acknowledge4. Messrs. Smith and Nickerson (Chapter 16) gratefully acknowledge the assistance provided by Mr. A. Taplin of the Naval Ship Engineering Center, who prepared the active fin stabilizer section of the hull machinery chapter. Mr. Stephenson (Chapter 18) gratefully notes that the machinery arrangement illustrations and many of the piping diagram illustrations in the piping systems chapter were included with the permission of Mr. W. L. Baptie of American Mail Line, Ltd. The typical chapter author is a highly competent engineer who enjoys his field of specialization and has devoted the majority of his life to it. By studying the various chapters, it will become apparent that a book such as this is published only once per generation. INQ
\
,
I -
, $
t I
1
Division 1 Introductory
1'
Chapter I
J. R.
GENERAL CONSIDERATIONS I N MARINE ENGINEERING
KANE, Director of Engineering,
Newport News Shipbuilding and Dry Dock Company
1. Intrbduction .......................... 2. Concepts and Concept Formulation.. .... 3. Ship System Formulation.. ............. 4. Development of Main Propulsion System Requirements.. .....................
%. ~ a i Propulsion n Plent ~rade-offStudies. 6. Preliminary Design Considerations. . . . . . 7. Specifications. ........................ 8. Final Design and Working Plans. . . . . . . . 8 9. Tests and Trials. ..................... 1 2 5
$
11 18 31 33 35
Power Plants
I
Chapter I1
THERMODYNAMICS AND B E A T ENGINEERING
Jws T. HOLM, Professor, Webb Institute of Naval Architecture J. B. WOODWARD 111, Professor, University of Michigan
I
1. Review of Fundamentals. .............. 2. Heat Transfer in Boilers. . . . . . . . . . . . . . . 3. Internal Thermodynamics of the Steam Tutbine. ...........................
Page PW~ 38 4. External Thermodynamics of the Steam Turbine ............................ 55 5. ~herniod~namics of steam Cycles. . . . . . . 61 / 49 6. Waste Heat from Diesel and Gas Turbine i Engines ............................ 73
/
i b
Chapter I11
BOILERS AND COMBUSTION
EVERETT A. CATLIN,~ a x i n Engineer, e The Babcock & Wilcox Company 1. Classification of Marine Seam Generatom 2. Considerations in the Selection of a Boiler
Chapter I V
pa@ 78 3. Boiler Pesign . . . . . . . . . . . . . . . . . . . . . . . . . 94 90 4. Boiler Operation.. ..................... 125
NUCLEAR MARINE PROPULSION
ROBERT T. PENNINQMN, formerly Manager of Nuclear Maxine Engineering, Advanced Products Operation, General Electric Company
I
Page page 1. Basic Fundamentals. .................. 130 3. . Nuclear Propulsion Applications. . . . . . . . 149 2, Reactor Design Considerations. . . . . . . . . 138 vii
I STEAM TURB1,NES WILLIAMI. H. BUDD,Assistant to Manager of Engineering, Marine Systems, DeLaval Turbine, Inc.
unrtpucr v
Turbine Control.. .................... Rotors and Blades. .................. Norzlea, Diaphragms, and Stationary Blading.. .......................... Casings &adPackings. ................ Lubrication and Bearings. ............ Main Propulsion Turbine Operation. . . . Auxiliary Turbines. ..................
1. Nonreheat Main Propulsion Turbines. .. 2. Reheat Main Propulsion Turbines. ..... 3. Main Propulsion Turbine-Nuclear
Cycle. ............................. 4. Combined Steam and GaB Turbine Main Propulsion Cycles. ................. 5. Turbine Speed, Number of Stages, Dimensions. ....................... : .
1
Chapter X
ELECTRIC PROPULSION DRIVES
W. E. JACOBBEN, Manager, Marine Systems Engineering, General Electric Company 1. Introduction.. ........................ 334 2. The Diesel Direct-Current Drive System. 339 3. The Turbine Direct-Current Drive System 347
180 185
4. The Turbine Alternating-Current Drive
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
5. The Diesel Alternating-Current Drive
System. ........................... 356 6. Electric Couplings.. ................... 360
190 193 196 199
Chapter XI
201
PROPELLERS, SHAFTING, AND SHAFTrNG SYSTEM VIBRATION ANALYSIS
Assistant Chief Engineer, Newport News Shipbuilding and Dry Dock Company C. L. LONG, Chapter VI GAS TURBINES A. 0. WHITE, Manager, Advanced Applications Unit, Medium Electric Company
.
1. 2. 3. 4. 5.
Basic Considerations.. .................. Arrangement and Structural Details. . . . . . Accessories. ........................... Controls. ............................. Centrifugal Compressor Design. .........
206 213 218 219 222
6. 7. 8. 9.
Page
~k Turbine
Operation, General
Axial-Flow Compreseor Design. . . . . . . . . Turbine Design and Construction. ...... Combustion Systems. ................. Bearings, Seals, and Lubrication. .......
Introduction.. ........................ 362 b5. 2. Arrangement Considerations.. . . . . . . . . . . 365 e 6 . p - 3 . Shafting Loads.. . . . . . . . . . . . . . . . . . . . . . . 366 -7. ' '/4. Shafting Design. . . . . . . . . . . . . . . . . . . . . . . 372 8. b- 1.
I
1. Introdrtction. ......................... 246 2. aaracte$tics of Diesel Engines. . . . . . . . 251
Bearings.. ............................ Propellers. ........................... Torsional Vibration.. .................. Longitudinal Vibration.. ...............
379 384 388 393
9. Whirling Vibration.. ................... 397
225 229 235 239
Division 4 Auxiliary Co~aponents
MEDIUM AND HIGH-SPEED DIESEL ENGINES LASKARWECHBLER,Technical Director, Machinery Systems ~ i v i s i o n ,Naval ship ~ n ~ i n e e r i n ~
v
Page
PUMPS, FORCED-DRAFT BLOWERS, COMPRE$SORS, AND EJECTORS Supervisor, Centrifugal Pump Engineering Departmen;t, DeLaval Turbine, Inc. G. W. SOETE, Page
page
L=-
3. Marine Uses for Diesel Engines. . . . . . . . . 257 4. Design Considerations. ................ 261
1. Centrifugal Pumps.. ................... 401 4. Rotary Pumps.. ...................... 432 2. Reciprocating Steam Pumps. ........... 422 \.-5. Forced-@aft Blowers.. . . . . . . . . . . . . . . . . 436 3. Power Pumps.. ....................... 428 L. 6, Compressors.. ......................... 440
7. Ejectors.. ............................ 444
LOWSPEED DIRECT-COUPLED DIESEL ENGINES KURTILLIES, Professor, Technische Universitat Hannover Pege 1. survey of Principal &acteri&ics. ..... 280 2. Engine Subsystems. ................... 292
Chapter XI11
JOBEPH F. SEBALD, Consulting Engineer and Special Consultant to Gilbert Associates, Inc.
3. Overall Considerations. ................ 303
Transmissions REDUCTION GEARS
HAROLD W. SEE~AR, Manager, Technical Support, Marine Mechanical Dep&ment, westinghouse
MAIN AND AUXILIARY CONDENSERS
!*
Page
Page
1. General C~aracterhtics................. 450 2. Condenser Design. . . . . . . . . . . . . . . . . . . . . 456
3. Surface Condenser Performanm. . . . . . . . . 473 4. Performance Predictions from Design Geometry.. . . . . . . . . . . . . . . . . . . . . . . . . 478
Chapter X I V
HEAT EXCHANGERS
CHARLEB D. ROBE,Vice President, AquaXhem, Incorporated PHILIPLIU, Chief Thermal Design Consultant, Research and Development, Aqu*Chem, corporated
Electric Corporation page v
1. Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . 310 3. Gear Design. . . . . . . . . . . . . . . . . . . . . . . . . . 317 2. Tooth Design Factors. . . . . . . . . . . . . . . . . 313 v 4 . Applications.. . . . . . . . . . . . . . . . . . . . . . . . . 331
page
page
1. . Introduction. ......................... 488 2. Heat Transfer in Shell-and-Tube Heat Exchangers.. 496
3. Heat Exchanger Applications. .......... 514
In-
.......................
ix
Chapter XV
DISTILLING P U N T S
.
Division 6 Supporting Technology
C ~ I D.WROBE,Vice President, AqueChem, Incorporqted page
Page
1. Distilling Plant Designs. ............... 530
2. Distilling Plant Design Considerations. .. 550
Chapter XVI
-
HULL MACHINERY
IRVING W. SMITH,Mechanical Engineer, Office of Ship Construction, Maritime Administration ARCHERM. NICKERBON, JR.,Senior Engineer, J. E. Bowker Associates, Inc. v
Chapter XX
1. General Design Consideratioqs. .........
ptlge 564
L-
2.
'
570
WATT V. SMITH, Head, Friction and Wear Branch, Materials Department, Naval Ship Research and Development Laboratory, Annapolis, Maryland J. M. GRUBER,Vice President, Waukesha Industries Corporation
I I
,
1
page
Hull Machinery Installations.. ..........
page
I
ELECTRIC PLANTS
Chapter XVIII
1. Machinery Space Arrangement.. ........ 670
'2. Piping Design Details.. ................ 676 ;'3. Piping Systema........................ 682 "
ENVIRONMENTAL CONTROL
JOHN W. MARKERT, Professional S u p p o r t A i r Conditioning, Office of Construction Management, Public Buildings Service, General Services Administration 1. 2. 3. 4.
Page 2. Applications.. ........................ 796
CONSTRUCTION MATERIALS
W. LEE WILLIAMB, Assistant Bead, Materiala Department, Naval Ship Research and Development Laboratory, Annapolis, Maryland M. ROBERTGROSS,Head, Materials Engineering Branch, Materials Department, Naval Ship Research and Development Laboratory, Annapolis, Maryland 1. 2. 3. 4.
Prefacing Remarks. . . . . . . . . . . . . . . . . . . . Corrosion of Metals.. .................. Fatigue ............................... Behavior at Elevated Temperatures. ....
Page
Page
810 810 817 821
5. Applications of Materiala.. .............. 824 6. Glossary of Metallurgical Terms Used in Materials Engineering.. .............. 835
PIPING SYSTEMS
Manager, Piping Design '~epartment,Newport News Shipbuilding and Dry E. E. STEPHENBON, Dock Company ',
Pa%e 1. Automation System.. ................. 791
Chapter XXII
6. Lighting and Power Distribution. ... :... 640 7. Interior Communications.. ............. 654 8. Electronic Navigation and Radio Communication..................... 659 9. Wiring Application and Methods. ....... 663
605 607 614 621 635
AUTOMATION
W. 0. NICHOLS,Chief Engineer, Central Technical Division, Shipbuilding Department, Bethlehem Steel Corporation
Division 5 Sl~ipboard Systems
Introduction.. ........................ Generating Plants. .................... Switchboards and Panels. .............. Powe~Equipment. .................... Lighting Fixtures and Equipment. ......
Page
1. Review of Fundamentals. . . . . . . . . . . . . . . 770 ""2. Bearings.. ............................ 778 \--~ 3. Lubrication System. ................... 785
Chapter XXI
1. 2. 3. 4. 5.
BEARINGS AND LUBRICATION
Intraduction.. ........................ System General Requirements. ......... Design Criteria and L o 4 Components. .. Piping Systems.. ......................
710 718 726 734
5. 6. 7. 8.
Air Handling System Resign.. .......... Air Handling Equipment. .............. Beating and Cooling Equipment. ....... Refrigeration Equipment.. .............
745 756 763 766
Chapter XXIII
PETROLEUM FUELS
Manager, Technical Services, Marine Sales Department, Mobil Sales and CARLE. HABERMANN, Supply Corporation Page
Page
1. Fuel Manufactureand Characteristics. ... 842
2. Fuel Procurement. . . . . . . . . . . . . . . . . . . . . 853
page INDEX.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858
CHAPTER I
J. R. Kane
1
I
General Considerations in Marine Engineering
Section 1 Introduction The first efforts to apply mechanical power to the propulsion and operation of ships date back to the early oighteenth century, nearly concurrent with the start of the Industrial Revolution. By the beginning of the nineteenth century, almost a full century before the Wright brothers made their first sporadic flights in a glider at Kitty Hawk, ~ t e a m - ~ r o ~ e ships ied had become a commercial reality, and marine engineering was born. Considering such an early beginning, it mems paradoxical to have to say now, well along in the twentieth century, that it is difficult, if not impossible, to write a definitive text on the subject. Such is the case, however, since the field continues to enter new oras of activity and evolution. One of the reasons this subject is difficult to treat is that ships have never been simple products but, to the oontrary, require an exceptional number of specializations to plan, design, and build. Thus marine engineering is not as simply categorized as, for example, civil, machanical, electrical, or chemical engineering, but is an integrated engineering effort comprising parts of many ongineering disciplines directed to the development and dosign of systems of transport, warfare, exploration, and tlstural resource retrieval which have only one thing in earnmon; namely, that they operate in or upon the crurface of a body of water. The field of engineering activity designated as naval wrohitecture and marine engineering is concerned with at let~stthe following areas:
I
Ocean engineering. The conception, design, construction, and operation of vehicles, submersibles, and fixed or floating structures and their integration into systems for the conduct of oceanographic research, exploration of ocean resources, and the utilization of ocean resources are encom~assedin this categorv.
The division of responsibilities between the naval architect and the marine engineer differs from one activity to another. However, the marine engineer is, in general, responsible for the engineering systems required to propel, work, or fight the ship. More specifically, the marine engineer may be responsible for the main propulsion plant, the powering and mechanization aspects of ship functions such as steering, anchoring, cargo handling, heating, ventilation, air conditioning, electrical power generation and distribution, interior and exterior communication, and other related requirements. The naval architect, in general, is primarily concerned with the hydrodynamic and hull form characteristics of the ship, the structural design of the hull, the control aspects of the vehicle, habitability considerations and the ability to survive and endure in the service environment. The naval architect, assisted in appropriate areas by the marine engineer, is responsible for the overall arrangement or configuration of the ship extending to both the exterior and interior arrangements. I n addition, the naval architect is generally charged with the responsibility for the overall esthetics of the design, the interior decoration, and the general suitability and pleasing Inland waterway and ocean transportation. The con- quality of the architecture. usplion, design, construction, and operation of vehicles Certain aspects of the design of marine vehicles are utilizing the waterways and oceans, especially the ocean difficult gto clearly assign as the responsibility of either murfaces, for transportation of commodities, goods, and the naval architect or the marine engineer. The design personnel, are included in this category. The integration of propellers or propulsors is one of these, being in the of tho operation of these vehicles with land transport& minds of some a hydrodynamic device in the domain of tion via harbor and terminal facilities is an extremely the naval architect, and in the minds of others to be an hportant consideration. In the case of small boats, energy conversion device similar to pumps, turboymbts, and cruise ships, transportation may be secondary machinery, and the like, thus in the sphere of the marine h lsi~ureor sport as an objective. engineer. Hull vibration, excited by the propeller or by Naval engineering. This category includes the con- the main propulsion plant, is another such area. Noise aegt,ion, design, construction, and operation of naval reduction and shock hardening, in fact dynamic response rtcrfaoo ships and submarines and their integration into of structures or machinery in general, usually must be wsrf~bre systems. Means of appraising the military the joint responsibility of both the naval architect and effrotivenessof these systems and the optimal utilization the marine engineer. Cargo handling, cargo pumping of thoir properties are major considerations. systems, environmental control, habitability ,. hotel
2
I
MARINE ENGINEERING
1
services, and numerous other such aspects of ship design all involve joint responsibility and interfacing between the naval architect and the marine engineer. The traditional distinctiowbetween naval architecture and marine engineering in t k multifarious aspects of ship selection, design, construction, and operation are tending to disappear, to be replaced by broader concepts of systems engineering and analysis. Because of the multidisciplined nature of marine engineering and naval architecture, they have been particularly affected by the impact of the explosive growth of technology during recent years. Prevalent use of the electronic computer has been particularly influential, in that complex rnathematical analyses once considered prohibitively laborious are now routinely made. By providing the ability to rapidly conduct an increased number of computations, readily store and analyze data, and simultaneously
consider a larger number of factors, the computer makes mathematical simulation of complex problems feasible and is leading to a better optimization of designs. Furthermore, due to the period of large-scale industrial development into which we have entered, there is increasing acceptance of the principle of planned technology which affects systems of all sorts, including marine transportation, oceanography, and recovery of ocean or ocean-bottom resources. By surveying the series of inventions and innovations which have established the present state of the art of marine engineering, it becomes apparent that engineering in the ocean environment is characteristically a dynamic, continuously advancing technology. As a result, this text must be considered an interim report of the processeis that are developing in a broadening marine engineering field.
I
I
I
Section 2 Concepts and Concept Formulation 2.1 Early History. In about the year 1712, an enterprising blacksmith from Dartmoor, England, by the name of Thomas Newcomen, successfully developed a rudimentary steam engine for the purpose of pumping water out of mines. This engine consisted essentially of a single-acting piston working in a vertical open-topped cylinder. The piston was packed with hemp since the state of the metal-working art was very primitive and a tolerance of about one-sixteenth inch out of round or "the thickness of a thin sixpence" was about the best that could be expected. The piston was connected to one end of a rocker arm by a chain without a piston rod or guide. The differential working pressure was derived primarily from the vacuum which was created below the piston by water spray into the steam space a t the end of the upstroke. The steam and water valves were worked by hand. Some sixty years later, radical improvements were made by James Watt, whose name is more frequently associated with the early development of the steam engine. I n the course of time, numerous other.improvements followed, of which the most important was probably the double-acting inverted vertical engine which proved to have so many advantages that it has remained standard ever since. Accounts of the work of men such as Savery, Newcomen, Papin, and Watt in connection with the invention and development of steam engines are truly exciting [I, 2,3].l Despite the much earlier development of steam engines, their application to the propulsion of ships was not undertaken until about 1784. Attempts to adapt the early steam engines to ship propulsion were carried out almost simultaneously in America, Scotland, Numbers in brackets designate References at end of chapter.
and France, and a t least seven reasonably practical steamboats were developed before 1807 when Robert Fulton inaugurated the first commercially successful use of steam marine propulsion in the small wooden paddle wheel vessel Clermont [I]. The Clermont operated up the Hudson River from New York to Albany, a distance of 150 miles, in about 32 hr. Although paddle wheel vessels were promptly adopted for river service, twelve years elapsed after the launching of the Clermont before the steamer Savannah made the first ocean voyage from America to Europe. It should be noted, however, that even in this instance the machinery was not operated continuously during the outbound leg of the trip and the inbound leg was made under sail. The era of the paddle wheel steamships reached a climax about 50 years later when the steamship Great Eastern was built. This was a steel-hulled vessel almost 700 f t long and 22,000 tons burden, which is large even today for a cargo vessel, and which had the principal fault that it was too advanced for its time. The introduction of the screw propeller in 1837, which was a revolutionary development, similarly did not immediately displace sailing vessels. As late as 1860 the speed of the best clippers still exceeded that of any steams hi^ and the greater d art of the work a t sea continued td be accomilished inder sail. B y the year 1893, the year of the founding of The Society of Naval Architects and Marine Engineers, the screw propeller.. driven by a triple-expansion steam engine had become the,predominant means of propulsion of seagoing ships although t addle wheels were still used with river-and- excursion steamers. Steam was almost universally produced by Scotch boilers and coal was the
GENERAL' CONSIDERATIONS
c o w o n fuel. The steam turbine and diesel engine were yet to make their debut. The decade from 1893 to 1903 was a period rich in marine engineering development. The early reciprooating steam engine reached the point of development of the six-cylinder quadruple-expansion engines of 10,000 indicated horsepower supplied with steam by Scotch boilers a t 200 pounds pressure. The use of electric power generated by engine-driven "dynamos" a t 100 to 112 volts was increasing rapidly. Water tube boilers, which would eventually replace the Scotch boiler on the seas, had become established in England and in the United States. An important milestone in marine engineering was the development, by Sir Charles A. Parsons, of the first successful application of the steam turbine for marine propulsion; this was accomplished aboard the Turbinia, a small vessel similar to a torpedo boat. The rotative speed of the Turbinia's three series turbines was about 2000 rpm, and they were coupled directly to relatively primitive screw propellers in a triple shaft arrangement. Parsons was dismayed on his earliest trials to discover that the wheels more or less "bored a hole in the water," developing disappointingly low driving thrust. Much developmental work was necessary before this new prime mover was successfully adapted to the requirements of marine propulsion. In what must certainly be considered one of the earliest efforts at model tank testing of propellers, Parsons investigated the subject of cavitation and succeeded in redesigning his propellers (three per shaft were ultimately employed) such that in 1897 a t a naval review of the British fleet a t Spithead, England, the Turbinia astounded the British admirals by steaming past smoothly a t a speed of 34 knots, belching smoke like an angry bull tossing dust. Lord Kelvin described this development as "the greatest advance made in steam ongine practice since the time of James Watt" [4]. Prior to 1893, a number of internal-combustion engines were attempted using anything from gunpowder to gas. One of these was a radically different type of engine in which the combustion air charge was compressed to a pressure and temperature above the ignition point of the fuel; it was patented by Dr. Rudolf Diesel, a German engineer, in 1892. There were very serious Wculties to be overcome with the diesel engine, development proceeded slowly, and it was not until fifteen to sixteen years later that a successful commercial diesel enginc of 25 hp was produced. Once this had been achieved, however, rapid progress waq made, and in a few years many firms in Continental Europe were actively building diesel engines with as much as 500 hp per cylinder. Already a t that early date experimental cylinders of 2000 horsepower were under test. The challenge to the coal-fired low-pressure reciproaating steam engine came from the steam turbine and the 'dio~elengine about the same time a t the turn of the aantury. World War I retarded developments, however, etld maintained the supremacy of coal for a little while
3
longer. After the war, oil found preference either as diesel engine fuel or for raising steam. It also reduced crew requirements and made fuel storage an easier task. The historical developments noted in the foregoing were beginnings which, when viewed against the techniques and materials available a t the time, were magnificent conceptions. No effort has been made here to include the full roster of great names and pioneer events in marine engineering. However, some familiarity with the background of the early days in marine engineering is highly recommended for those entering this field to develop an appreciation of the hopes and disappointments, the dreams and disillusionments, and the blood and sweat which lie behind the present state of the art [I-81. 2.2 Broader Concepts-Systems Analysis. The concept which motivated the majority of the early attempts in marine engineering was quite simple; namely, to develop a superior system to overcome the vagaries of the wind and the feebleness of muscle power in the propulsion of ships. However, marine engineering today entails much broader system requirements and concepts than most developments of that time. By way of introduction, one particular historical undertaking is given special note since it contained, in a primitive way, elements of systems analysis. In 1776, a year which should strike a familiar note with most Americans, a Connecticut Yaxikee named David Bushnell built the Turtle, the first submersible craft to make an undersea attack during warfare. The Turtle of the American Revolution, so called because it could be likened to two turtle shells clamped together, was built of barrel staves and iron, contained ballast tanks which were flooded to submerge, and was moved by primitive spiral screws. Reference [8] contains an interesting description of the Turtle and its precocious concepts. The Turtle was not by any means the first successful submersible craft, but was one of the most significant, since among other things it was one of the earliest, and perhaps boldest, attempts to develop a military system involving an evolutionary marine vehicle. The operational concept of the Turtle d i e r e d somewhat from most other inventions of that era since it related in a primitive way to an entire system. It was intended that the pilot dive the vessel under the water in order to evade lookouts on an enemy vessel, attach a time-delayed explosive mine to the ehip's bottom, and make a safe escape. The initial target of the Turtle was Admiral Howe's 64-gun flagship, HMS Eagle. The story of this initial venture is fascinating; the Turtle did not in fact succeed, kt.it came perilously close to doing so. George Washington wrote to Thomas Jefferson a t the time of the Turtle, "I then thought and still think that it was an effort of genius, but that many things were necessary to be combined to expect much from the issue against an enemy who are always upon guard" [91. Although the development of the first ironclads, the Merrimac and the Monitor, almost a century later had probably a more revolutionary effect on the evolution of
GENERAL CONSIDERA'I'IONS
MARINE ENGINEERING
5
I
STATE SYSTEM OBJECTIVES
DETERMINE CONSTRAINTS
-\:
r
DELINEATE SYSTEM REQUIREMENTS DELINEATE DESIGN REQUIREMENTS
- I
i
DEVELOP DESIGN ALTERNATIVES
~r \
t
\ ln
= &
PERFORM TRADE-OFF STUDIES
-ESTABLISH OPTIMUM DESiGN
-
ii
L
1L
DELINEATE DETAILED SPECIFICATIONS DETAILED
Fig. 1 - Functional processes in a systems analysis .
.
warships, Bushnell's submarine is of special interest because of the singularity of its operational concept and its primacy. Actually it contained all the elements of a modern problem in concept formulation for a planned technological development: a mission objective or primary task, an analysis of the objective to establish specific operational requirements, trade-offs concerning alternative methods of accomplishing the mission, constraints imposed by limitations of techniques, materials, manpower, money, and time, and last but not least, the necessity of obtaining the interest and support of the controlling authority for what must have seemed, in this case, to be a radical venture. In the early historical stages of the basic engineering process, the concepts formed and the decisions made, although frequently ingenious, were of sufficiently narrow scope that a single individual could become intimately familiar with all facets of the undertaking. The stakes were high for a successful development; rugged individualism was the rule since society had not yet embraced the role of technological development, and support by the existing governing bodies was scanty or nonexistent. Success depended to a large extent upon
intuitive perception and upon chance. Today, in this age of institutionalized knowledge and electronic computers, such factors are still important, but are being largely transcended by systematized approaches and by team activity. The ship, which once was viewed as a highly subjective entity, possessed of feminine and almost human attributes, is now looked upon more objectively as a link in a transportation system, a military platform, or as a medium in a system of transferring people, commodities, national presence or authority, and the like from one point to another. From a functional point of view, a ship is a most complex vehicle which must be self-sustaining in its element for long periods of time with a high degree of confidence. A ship is perhaps the most multipurpose vehicle having more built-in functions than any other type; and, as a part of a transportation or military system, the ship envelope contains a greater variety of components than any other vehicle in the system. A ship's mechanical, electrical, and structural systems are quite complex and are further complicated by the fact that they must be environmentally oriented. Due to the complexity of ships and their interfaces in transportation networks, the design of optimum ship systems cannot practicably be undertaken in a random manner. The design of complex systems involving ships is best accomplished by utiliing the systems analysis approach [lo-141 as schematically illustrated by Fig. 1. I n this way, the design process can be organized in logical steps so as to ensure that, when completed, every facet of the design has been given proper treatment. As indicated in Fig. 1, a systems analysis is initiated by establishing a system objective. Beyond that point the systems analysis approach is a continuously iterative process with each of the functional processes possibly having an impact on those remaining. For example, referring to Fig. 1, the initial system objective could be to transport cargo between two points at a given rate and a t the lowest possible cost. Proceeding with this objective, constraints such as time and capital limitations must be established. Since the constraints may alter the original objective (e.g., preclude transporting cargo at the desired rate or make higher rates attractive), the original objective must be reevaluated. The various aspects of the design process continue until all factors in the analysis are compatible, at which time the design is complete. In more general terms, a combination of theory and facts (including a careful statement of the constraints upon the system) is used to ~roducean abstract study or model of the actual situation. The model, in turn, is combined with a set of aims to produce a plan of action or a proposed technical approach. Working with such analyses and with checks against experience and data gives rise to a body of correlated information which feeds back to modify the designs which are acceptable, the facts which are relevant, the controls which are efficient, is and the aims which are realistic. Systems engidng
I
1
1
the term for such a process when limited to basic engineering processes. Systems analysis is the more general term for the process when social and economic factors in addition to basic engineering processes are included. Operations research is the name of the process when operability, that is, the optimum deployment or utilization of components, men, and machines, is the principal objective. Work study is another term of related connotation, although in this case the emphasis is placed on optimum utilization of man, and reduction in manning requirements, by taking a fresh look at work patterns and habits that have come to be taken for granted. The objectives and constraints upon which the policy for systems analyses is based have differing motivations for military systems and for merchant marine transport systems; but in both instances they ultimately reduce to the same base-cost effectiveness. I n the case of merchant systems, the proposed system must be cost effective as compared to other potential investments in order to command the necessary venture capital under the free enterprise system, or they have to be justified for governmental support by subsidy. Military planners are charged with the national defense, but there is in fact a limit to the amount of money available for such purposes as there are more military systems competing for funds than can be supported by the funds available. Consequently, the analysis of military budgets becomes a process of identifying systems, or combinations of systems, which have the maximum military cost off ectiveness. Cost effectiveness seems simple to comprehend, but usually is difficult to quantify [15]. In general, the
expression denotes a measure of the degree to which the achievement of the tasks or missions of a system (e.g., revenue earned or national protection provided) has been maximized relative to the costs associated with the system. Since the effective life of a ship is approximately twenty to twenty-five years, a period long enough for economic and political factors to undergo substantial change, the projection of life cycle costs associated with ships is inherently less accurate than life cycle cost estimates made in connection with vehicles such as automobiles or aircraft which have a much shorter life cycle. When conducting life cycle cost analyses with ships, which are relatively long lived, considerably more importance must be attached to the events which occur during the early stages of the ship's life. There is little question that the basic vehicle will perform satisfactorily for a 25-year life; however, there have been' many cases in which ships have been reequipped, modernized, jumboized, converted, etc., a number of times during their lives. As a result, the credibility of projections for the first five or ten years of a ship's life are considerably better and are often given more weight than more distant forecasts. However, despite the uncertainties associated with long-range forecasts, attempts to project them are being made and a new branch of systems analysis termed assurance engineering has been developed to give numerical expression to characteristics such as reliability, maintainability, logistic aupport, operability, safety, and similar factors which augment the standard design performance estimates traditionally made. Also, producibility analyses, requiring a combination of design and industrial engineering skills, are sometimes made to assure a design best adapted to economy in construction.
Section 3 Ship System Formulation 9.1 Mode of Utilization. Before proceeding with a mview of the marine engineering phase of a ship system formulation, which as indicated by Fig. 1 does not oornmence until the broader aspects of the system have boen tentatively formulated, it is useful to review some af the broad considerations. In particular, the modes in which ships can be utilized and the payload and speed oharacteristics of ships are of great importance in that they must be compatible with the overall system oonsiraints.. From the viewpoint of utillation, marine vehicles mny be classified in the following three categories:
in terms of deadweight and cubic requirements, must be very carefully analyzed as the latter will have a controlling effect on the vessel configuration. (b) As a mobile fighting base. Seaborne bases for force groups, weapons systems, missiles, aircraft, or other sJrstems of warfare either tactical or strategic and either offensive or defensive are included in this group. In this instance, the design of the ship is subordinated to the military system and weapon requirements except for certain inescapable essentials such as seaworthiness, habitability, etc. Payload in this case will generally be defined in military terms relating to militaw effective(a) As a link in a tramportation ~ s t e m . Inthis case, ness, and the speed requirement will be a function of the payload, mean effective speed between t e d n a l s , turn- expected speed of the hostile forces and the successful mound time, and the number of vessels are the ~rimary accomplishment of the n~ission. vmiables and must be considered in relation to their (c) As a* special-purpose vehicle or platform. This gffeot on the initial and daily operating costs as well as category includes many diversified craft which have little tho other facets of the transportation system. Payload, in common except that they all work or operate in an
,
6
GENERAL Cob
MARINE ENGINEERING I
Table 1
A Comparison of Constraints Imposed upon Merchant and Military Ship Systems
Tramportation market potentiak cargo and/or passengers Type of tran ort system contexnplated:?ulk, break bulk, containerl passengeFcar o combinatmliquid and buk etc. Most likely itine terminal facilities, h a r b ~ h t a t i o m , c d limitatiom, and fueling ports Linking services: shore d@ribution systems, new termma1 facilities, cranes, and so on Competing services Socio/political considerations and union relations Economic projections, financial support, government subaidii&etc. Technologid development, state of the art G c t o bodies, such as ABS and U%G
Fig. 2
Specitlc power Venus speed for various vehicles
ocean or waterway environment and that much support for the systematic design of them is derived from the body of marine engineering knowledge obtained from less specialized vessels. Oceangoing tugs, salvage vessels, oceanographic research ships, submersibles, dredging vessels, yachts, ferryboats, towboats, pushers, barges, hydrofoil craft, surface effect ships, and many others are examples of such special-purpose craft. Category (c) does not lend itself to generalization beyond the fundamentals of naval architecture and marine engineering. Neither, one might conjecture, do (a) and (b). However, the constraints to be considered in determining system requirements so as to ensure a reasonably optimum design configuration do parallel between merchant and military applications to rs surprising extent as indicated by the comparison in Table 1. 3.2 Payload and Speed Considerations. I n addition to the constraints dealing with the mode of utiliiation, payload and speed considerations have a strong influence on the selection of the type of vehicle employed. Payload and speed constraints are important in that they restrict the types of vehicles which are feasible for parti~ularapplications. Figure 2, parts of w h i ~ hwere taken from references [16-201, is an informative com-
Type of war situation anticipated Tactics, strategy, mission pro-
,,
Most like1 operational locale, support8aaes, replenishment means, etc. Force pou compatibility, potential dies Enemy threat in weapons and ship types Socio/political considerations Fiscal environment and budg e t pressures ~ Technolo 'cal development, state o&he art Military specifications
parison of alternative means of transportation in that the feasiblerange of speed for the various types of vehicles becomes evident. Although payload considerations are still a factor, size restrictions are less stringent in connection with ships than with the alternative modes of transportation. An investigation of a systematic family of ships (a parametric study in which size is the principle characteristic that is varied) will demonstrate that ships are not sizelimited and can be built as large as one may wish without encountering limitations from the laws of physics. Dimensional analysis will show that geometrically similar ships of a diierent scale will float at the same proportionate draft since both the water displaced (buoyancy) and the weight of the ship tend to increase as the cube of the scale. A corollary conclusion from such systematic investigations is that displacement ships are not particularly weight-sensitive. Vehicles such as fixed-wing aircraft, hydrofoil craft, planing boats, and surface effect devices in general are weight-sensitive and size-limited as may be seen from a simple dimensional analysis. Such craft derive their support in flight from lifting surfaces of various types; when geometrically similar but larger versions of a prototype are considered, the weight of the craft, including its payload, increases approximately as the cube of the scale ratio while the area of the lifting surface increases only as the square. As a result, the unit pressure loading on the lifting surface increases directly with the scale. The increase in size of fixed-wing aircraft over the last several decades has been achieved largely by increasing the forward speed by almost an order of magnitude and by greatly refining and improving the lifting character-
istics of wings and fuselages by means of extensive research developments. As the speed in flight is increased, the basic configuration of the aircraft must be changed appropriately also, because, as compared with diplacement-type ships, vehicles in the aircraft or surface-effectsupported category tend to be size-limited and weight-sensitive. As may be evident from Fig. 2, the displacement type of vessel has very definite limitations with regard to the speed at which it can be efficiently driven. The inherent speed limitations for ships are most appropriately expressed in terms of the so-called speed-length ratio (the ship's speed in knots divided by the square root of the ship's length in feet) in conjunction with various ratios of the ship's dimensions such as the beam-draft ratio and the prismatic and block coefficients (see reference [21] for a comprehensive treatment of this subject). The most spectacular growth in the size of ships has been in tankera. During the early 19509s,the so-called supertankers were in the cargo deadweight range of 20,000 to 30,000 tons; whereas during the latter 19609s, tankers as large as 200,000 to 300,000 tons were being built with projected giants in the 1,000,000-ton range appearing feasible. The theoretical problem of optimizing a transport system would appear to be simply that of maximizing payload times mean effective speed from point to point while a t the same time minimizing initial costs and yearly operating costs. If this were the only consideration, ships would be in much greater favor as compared with aircraft than they are. Systems analyses of typical transport missions usually include another highly important factor which puts a great premium on higher speed; namely, flexibility, or the ability to be in the right place at the right time with the right payload. The great increase in the speed of communications and the resultant great increase in the rapidity of affairs in recent decades has resulted in a higher premium on speed and time in many instances whether justifiable or not. Aircraft, therefore, usually transport a substantial proportion of the people, special equipment, and lighter commodities in which cases speed is of great importance, while ships continue to carry the larger proportion of the heavy cargos and commodities and bulk cargos in both military and nonmilitary transoceanic routes. 3.3 Deflnition of Fundamental Requirements. The constraints imposed by the intended mode of utilization and requirements regarding payload and speed will Ittrgely define the fundamental requirements of the ship, and an analysis of the ship system can now be conducted for the purpose of establiahing a reasonably optimum aolution. All of the positive constraints upon configuration should be identified in the analysis, but as much freedom of selection retained as possible. Once the objective and the constraints have been clearly stated, tho analysis may often proceed to the development of a u~oful abstract model for the system. Parametric &dies, in which the prin~ipalindependent variables are
varied systematically, using the electronic computer as appropriate, are often made. The sensitivity of the system to variation of the independent variables begins to emerge and can be identified. Because of its value in decision-making, the sensitivity of system characteristics to such systematic variation of the system parameters is often specifically explored in a formalized sensitivity analysis. Exercise of such techniques should result in sufficient background to support decisions regarding a policy and a plan of action. This plan of action will generally result in decisions which will further limit the range of variables to be considered; for example, the range of the size and the numbers of ships required may be more confined, notional ship design arrangements may be selected, approximate manning requirements determined, first approximation of costs projected, and so forth. A description of some of the procedures which may be used during the preliminary design of a ship is given detailed treatment in references [22-31.1. In the case of cargo ships, the fundamental concept of the cargo transportation system must be established at this point as the design of the entire system is predicated on this decision. General cargo transportation systems which employ intermodal containers (i.e., systems in which cargo is packed in containers that are transported by trucks, ships, barges, and trains in any combination before being delivered and unpacked) are becoming increasingly popular. The use of intermodal containers offers several advantages, the major one being the minimization of the number of times the cargo must be handled on an individual basis with a corresponding reduction in damage, pilferage, and handling costs. The iterative process of assessment/adjustment described in the foregoing results in an initial design configuration baseline which is essentially a preliminary statement of the ship system requirements. Such ship system requirements include the followingfor a merchant vessel : a
a a
a
a
Payload (cargo/passenger capacity and description) Sustained sea speed and endurance Number of containers, holds, refrigerated spaces, etc., for balanced service Limits to overall diiensions such as length, draft, Beam, etc., for operability on required service Loading-discharging methods and capacities Hotel requirements such as heating, ventilation, air conditioning, galley, public spaces, power, and lighting Crew or manning requirements Automation and mechanization objectives Reliability and logistic support objectives Special requirements for navigation and communications Maneuverability (steering, handling, stopping, and backing) Anchoring and mooring
8
MARINE ENGINEERING POWER P
R
W
I
M I S S I O N PROFILE L ~
Main Propulsion System Shaft horsepower Propeller rpm Specific fuel consumption and bunker capacity Space and weight objectives Adaptability to ship configuration
REQUIREMENT A c n v SELECTION OST / EFFECTIVENESS
@QUIP AVAIL & CHUACTERlSllCS
I F € CYCLE COSTS
CONFIGURATION OI SYSTEMS
Auxiliary Ship Systems Power and lighting Steam-galley, deck, and heating systems Heating, ventilation, afid air conditioning Firefighting, bilge, and ballasting Fresh water
CONSUMAILES & tNDURANCE
Fig. 3
3.4
M & R AND LOGlSllCS SUPPORT
Hull Engineering Systems Anchor handling Steering engine and bridge telemetering control Cargo handling gear, such as winching systems, burtoning, and swinging boom Crane systems Bulk cargo systems, self-unloaders, etc. Container systems Palletized systems Tankering systems, such as cargo piping and pumps
Propulsion machinery preliminary design spiral
Speciflc
Marine
Engineering Requirements.
The broad requirements of the ship system as just established must be translated into specific performance capabilities by the naval architect and the marine engineer. Since the requirements established a t this point are broad (e.g., unmanned engine room), subsequent investigations may show that some of the requirements cannot feasibly be fulfilled; in which case, all considerations must again be re-evaluated. Most of the broad requirements of the ship system cannot be analyzed independently of the others; and further refinement of each, to a degree, involves yet another iterative design process which is analogous to a slowly closing spiral that gradually approaches a point of fixation. Figure 3, which was taken from reference [23], is a diagram of, this sort of iterative spiral. The marine engineer utilizes a procedure similar to that indicated by Fig. 3 when performing the design comparisons and trade-off studies required to establish specific design requirements in the area of his cognizance. Such specific design requirements will generally be of the following classifications:
Electronic and Navigation System Commupication, exterior and interior Radar Loran, Decca, RDF, etc., navigational aids Military electronics, sensors, command and control systems, weapons directors, tactical data systems, and electronic countermeasures The procedures which are used when designing the engineering aspects of a ship may best be illustrated by outlining the process of designing a ship from a marine engineering viewpoint. This is done in the following sections and is initiated by a review of the procedures used in developing the main propulsion system requirements.
Developme~~tof Main Propulsion System Requireme~~ts 4.1 Overall Considerations. The basic operating requirement for the main propulsion system is to propel the vessel a t the required sustained sea speed for the range (or endurance) required of the vessel and to provide stopping, backing, and maneuvering capabilities. I n the case of a military vessel, which rarely operates a t its maximum rating, the speed requirement may be partly stated in terms of a mqimum flank or burst speed, which need be sustained for only a short percentage of the operating life of the vessel, in conjunction with a
-
more efficient lower speed for long-range endurance. A further restriction is that the main propulsion system must fulfill all of the basic operating requirements at a cost within that allocated during the preliminary studies of the ship system; otherwise the preliminary studies must be re-evaluated. Many factors must be considered in selecting the main propulsion system. Reliability is of the utmost importance since the safety and security of the vessel will depend upon it. Specific fuel consumption, bunker
GENERAL
cot
capacity, type of fuel required, fuel availability, space and weight requirements, and the adaptability of the propulsion system to the overall ship configuration are closely related to the type of plant selected and must be evaluated. Comparative costs, that is, first costs and operational costs, are also major considerations in tradeoff studies. Before entering into the process of selecting the main propulsion plant, it is necessary that the power required for sustained operation and endurance be tentatively determined. Since the space and weight requirements for the propulsion plant can have a significant effect on the ship configuration, and since the dimensional and form characteristics of the hull and its approximate displacement are required in order to arrive at an estimate of the propulsive power required, it is apparent that the marine engineer must coordinate his activities with the naval architect from the earliest conceptual design stage in an iterative preliminary design process such as that discussed in the previous section and illustrated in the preliminary design spiral, Fig. 3. 4.2 Determination of Ship Resistance. The general subject of ship resistance falls within the domain of naval architecture as opposed to marine engineering. For this reason, a detailed treatment of the subject is left to reference [21]; but for completeness purposes, some of the considerations involved warrant a brief review. The most reliable means of determining the resistance of a ship is to construct a scaled model of the underwater portions of the ship and conduct model resistance tests at one of the towing tank installations. .However, for several reasons such a procedure is far from feasible during the preliminary design phase: one is that sufficient time is not available; another is that the ship dimensions frequently change during the preliminary design phase; and another is that repeated testing would be prohibitively expensive. When tentative values have been established for the ship payload, sustained sea speed, and principal dimensions, an approximate assessment of the ship's resistance aan feasibly be obtained by utilizing the results obtained from a series of tests with systematically varied hull forms. There are principally two such test series: the Taylor's Standard Series [32, 331 and the Series 60 [34]. The Speed and Power of Ships [32], which was the original presentation of the Taylor's Standard Series data, is in tm exceptionally clear and concise form for preliminary design purposes and is a classic that is extensively used by practically all design activities; if not used directly, it la a t least used as a standard for evaluating the relative merits of any particular ship configuration. Although the use of series test data to estimate the resistance of ships is straightforward, the process nevertheless entails a considerable amount of tedious labor. In the event that the accuracy of an estimate is somewhat I&a important than the rapidity with which it can be made, a statistical method similar to that developed by Johnson and Rumble [28] can appropriately be used. Johnson and Rumble developed a simple approximate
statistical method of estimating the weight, displacement, speed, power, and other principal characteristics of a wide variety of dry cargo ships and tankers by averaging plots of a substantial number of actual designs. A number of marine engineering design activities have reduced their data on existing design series to a similar basis such that it is suitable for programming on an electronic computer; this enables approximate investigations of the parametric type to be made rapidly. As noted previously, reference 1211 contains a detailed discussion of the methods which are employed to obtain resistance estimates for ships. 4.3 Selection of the Propulsor. Once the ship speed, requirements and resistance have been tentatively established, it is necessary to select the type of propulsor. With considerations restricted to the type of propulsor for the moment, as indicated by Fig. 4, which was taken from reference [35], some types are inherently more efficient than others for particular applications. The abscissa on Fig. 4 is in terms of the Taylor power coefficient,B,, which is defined as:
where
N = propeller rpm P = power, hp V4 = speed of advance, knots The efficiency of propulsiop devices, including jet propulsion, is presented in a somewhat similar manner in reference [36]. The selection of the propulsor may not be a simple process, particularly in marginal cases, because in order to establish the type of propulsor it may be necessary to a t least tacitly select the type of main propulsion machinery. For example, the gain in efficiency offered by selecting contrarotating propellers versus a Troost B Series propeller (discussed further in the following), for a cargo ship, must be assessed in light of the impact on the main propulsion machinery and shafting arrangements. Similarly, the selection of the number of propellers may be a multifaceted problem. I n general, vessels may be single, twin, triple, or quadruple screw. That is to say, the total power required to propel a vessel may be distributed (usually equally) between one, two, three, or four shafts and propellers. From the point of view of initial and operating costs, fewer numbers of propellers are preferred, but the magnitude of the ship effective horsepower requirements or restraints on the propeller diameter may force a multiple-screw arrangement because of excessive propeller loading and the attendant danger of cavitation associated with unduly small propeller diameters. I n addition, there may be other factors in a given case, such as less vulnerability, more maneuverability, or take-home capability in the case that propeller damage may be likely in service, which favor an arrangement with a larger number of propellers. )
GENERAL CONSIDERATIONS
MARINE ENGINEERING
-
BP
Fig. 4
Cornparim of opfimum ettlckncy valuer
There are several extensive systematic series of fixedpitch propellers which have been model-tested and are in a form convenient for design selection purposes. Of these, probably the most suitable for design approximation is the Troost B Series of three, four, five, six, and seven-bladed propellers although there are others which may be used [21]. I n the usual case, the maximum propeller diameter that will provide adequate propeller submergence for the operating draft of the vessel and provide ample tip clearances as well as adapt to the stern configuration of the vessel so as to minimize propeller blade frequency excitation forces may be used for propeller selection purposes. The propeller design established during the preliminary design phase is generally very close to that obtained from later, morerefined design studies. A trade-off study must be made between the propeller rpm which is required from a maximum propulsive efficiency viewpoint and propeller rpm constraints imposed by prime mover/transmission size, weight, and cost considerations. The propeller rpm which is necessary to achieve a maximum propulsive efficiency is frequently considerably lower than that which is feasible from the viewpoint of the prime mover/transmission (due to the greater torque and hence machinery size associated with lower propeller speeds). Furthermore, attainment of the maximum propulsive efficiency does not necessarily constitute the most cost-effective system. Propeller characteristics are in general such that the propeller can be designed to operate a t an rpm somewhat greater than that corresponding to the maximum propulsive efficiency without incurring a serious efficiency
fv diiemnf Wpcn of propulm
'
penalty. Whiie no significant penalty in efficiency is incurred with propeller rpm's slightly greater than that for peak efficiency, significant savings in the first costs, size, and weight of the prime mover/transmission can be realized due to the lower torque rating (with the power remaining the same). The most cost-effective propeller rpm is selected by conducting a trade-off study which balances the propulsive efficiency against the size, weight, and cost of the prime mover/transmission. 4.4 Establishment of Propulsion Plant Shaff Horsepower Rating. Good practice dictates that a ship's
propulsion plant be rated such that the desired ship speed can be attained with reserve shaft horsepower capabilities. Factors to be considered in establishing the reserve capability include fouling and roughening of the hull, roughening of the working sections of the propeller due to cavitation or erosion, and erosion and deposits on the internal flow passages and working elements of the prime mover and power plant parts; all of which result in a significant performance degradation (approximately 5 to 15 percent) in time. It is also important that the vessel have a reasonable ability to maintain speed in moderately rough seas and adverse weather conditions. The usual practice for providing such a margin is to utiliie the parameter sustained sea speed, which is defined as that speed which is obtained a t some percentage of the installed maximum shaft horsepower, during trials, a t design load draft, under favorable weather conditions, when the vessel and engines are new, and the hull is clean. The percentage (or the so-called service factor) of the maximum shaft horsepower used to establish the sustained sea speed is ordinarily taken to
be 0.80 for cargo ships, which may be continuously loaded during the various legs of a voyage, and 0.90 for tankers, which in general are loaded on the outgoing leg of a voyage and in b a a s t during the return leg. However, depending upon the itinerary, the type of maintenance that is predicated, and mean time between dry docking and overhauls contemplated, the service factor used in a particular case may be somewhat Werent. 4.5 Selection of Main Propulsion Plant. Considerations concerning the selection of tbe main prop h i o n plant cannot be deferred until the propulsor, propulsion plant rating, etc., have been established, which may be suggested by the order of this discussion. Instead, the type of main propulsion plant is generally assumed a t the time the type of propulsor is established. Nevertheless, a final review of the main propuleion plant selected is one of the last tasks accomplished. Selection of a main propulsion plant entails the marrying of a power geeerator/prime mover, a transmission system, a propulsor, other shipboard systems, and the ship's hull. A myriad of possible propulsion plant arrangements may be considered by the marine engineer in making the selection. As indicated in Pig. 5, even when the range of considerations ia confined to the mo8t popular drives for fixed-pitch and controllable-pitch propellers, tbe number of permutations open to the marine engineer is sizable. It may be noted from Fig. 5 (which neglects infrequently used arrangements such as, for instance, directdrive steam turbines or the out-of-date reciprocating steam engine) that in modem ships only large-bore, slow-speed diesel engines are directly connected to the propeller shaft. Transmission devicea such as mechanical speed-reducing gears or electrical generator/motor transmissions are otherwise required to make compatible the relatively high rpm necessary for an economical and small prime mover and the relatively low propeller rpm nece8sary for a high propulsive efficiency. In the case of steam turbines, medium and high-speed diesel engines, and gas turbines, the high rpm inherent in a compact prime mover design and the low speed suited to the marine propeller is reconciled with speed reduction geah. Gear ratios vary from relatively low values for medium-speed diesels up to approximately 50 to 1 for a compact turbine design. An electricd transmission has attractive features, dthough its first cost tepds to be somewhat high; in this owe, the prime mover drives a generator or alkrnator
STEAM TURBINE(S1 n ~ v ~ n LLEICUTS w ~ u ~
WITH
DIESEL ENGINES IEDIUI
8,,ED0n
IRCVCRSI*.l
MECHANICAL
c,,BINATloN
REDUCTION GEAR
A:."S::.lnIvEnwI*.l 4.N.D,%JY,"."ty& -
y;r;~~~;;~~~;;;;~~;~
DIRECT SHAFTCOUPLED
cnA'TrvCE~WtOWpEnpOn*AuC~~
INOM nEvCReIN0)
-
COMBINATION DIESEL ENGINE AND QAS TURBINE DIESEL, UEDIUI S ~ ~ E D O R U I ~ U S ~ E D ,
MECHANICAL REDUCT I O N GEAR1nEvEn.Iu.J
IncvCnsINeI
-
FIXED PROPELLCR
eA,~Un,lNE,Hc,vvOU
""'ICVEm8~*~)
-1
ELECTRIC DRIVE lnEvnn#lu.)
DIESEL ENGINC~S) 'OW ""O
InIVEn#luo)
'
C
GAS TURBINE unrw ourv on UIOU P C R ~ D R ~ A M C E - luo*nEVEnsI~@~
ELECTRIC DRIVE IneVcn8Iue) MECHANICAL REDUCTION GEAR
MECHANICAL REDUCTION GEAR
d--
CONTROLLABLE AND REVERSIBLE PITCH PROPELLER
DIESEL CNGINES .,,,,, 8,EEoon , ,,,,, tao* ~LVIIOIUSI
Fig. 5
Alternatives in the wlection of a main propulsion plant
which in turn drives a propulsion motor having a large number of poles which is either coupled directly to the propeller or drives the propeller through a low-ratio reduction gear. Electrical drives may be either a-c or d-c; an a-c transmission is somewbt favored since it is lighter and cheaper, but it involves special design considerations in order to provide satisfactory maneuvering torque characteristics and becomes more comple~than a d-c transmission especially when the 'prime movers are diesel engines which may be stalled if J o e too abruptly. Reveming may be accomplished by stopping and reverb ing a reversible engine, rts in the case of many reciprocating engines, or by adding reversing elements in the prime mover in the case of steam turbines. It is geperdljl impracticable to provide reversing elements in gas turbines, in which case a reversing capability must be either provided in the transmiwion system or in the propulsor itself. Reversing reduction gears for mch tralismissions are available up to quite subs$antial powers, and controllable and reversible-pitch propellers also have been used with dim1 or gas turbine drives. Electrical drives provide reversing by dynamic braking and ener@zing (plugging) the electric motor in the reverse direction.
Section 5 Main Propulsion Plant Trade-Off Studies 8.1 Fundamental Concepts. The design of the maahinery plant, like many other general design projects, I@y consists of a correlation of a number of units end
elements into a functioning system which gives a desired performance. This entails selecting components, adjusting each to the constraints imposed by all others, and
GENERAL CONSIDERATIONS
MARINE ENGINEERING
12
arranging them so as to achieve the required system performance, a satisfactory configuration, and an equitable life cycle cost. There are a number of design decisions which must be made in formulating a main propulsion plant design. For example, the prime mover must be selected with the major alternatives being a diesel engine, oil-fired steam turbine, nuclear-fueled steam turbine, gas turbine, a combined design, or a special design such as that required for surface-effect vehicles. And once the generic type t been established then the major characterof ~ i a nhas istics of the plant must be selected. Questions which must typically be answered in selecting the major design characteristics of the propulsion plant are: Should a &&el plant be high speed, medium speed, low speed, two cycle, four cycle, and the like? Or, in connection with a fossil-fueled steam turbine plant, should the boiler have natural circulation, forced circulation, or no recirculation at all (once-through type)? With gas turbine installations, there is the choice of simple or regenerative plants. Innovations in nuclear technology continue to provide new alternatives in the design of nuclear plants. The most controversial subject in marine engineering is the relative merits of the various types of main propulsion plants and each type of plant has its own advocates, who often exhibit excessive enthusiasm for their particular type. Since a variety of types is used more or less extensively in a number of ships, it can be c~ncludedfrom this fact alone that all types bave their ~ l a c pand that the only way to determine the most suitable choice of main machinery plant is to consider all of the factors involved in each particular application. The selection of a ship's main propulsion plant may be influenced by previous practice, as is the case with most complicated engineering systems. Ordinarily, pertinent plans and essential data relating to the machmery of other ships, some perhaps rather similar to the one in question, will be available. If this information is aivailable and in a proper form, first approximations can often be made without detailed study, thereby reducing the range and number of variables that must be given detailed consideration in the preliminary stage. There are many factons which should be considered in conducting trade-off studies involving the various types of main propulsion plants; the more important factors are : 1 Reliability 2 Maintainability 3 Space and arrangement requirements 4 Weight requirements 5 Type of fuel required (including fuel treatment) 6 Fuel consumption 7 Fractional power and transient performance 8 Interrelations with auxiljaries 9 Reversing capability 10 Operating personnel 11 Rating limitations 12 Costs
4
I n addition, however, the selection of the type of main propulsion plant can be influenced by intangible personal factors reflecting the backgrofind or personal preferences of those interested in the construction or operation of the vessel, and greatly influenced by the experience of the operating personnel available to them. The aforementioned factor8 will be given a more detailed discussion in turn. 5.2 Reliability. Of all the factors which must be considered in selecting the most suitable type of machinery, reliability in service is one of the most important and should be given proper emphasis. The design effort devoted to this consideration has been receiving increasing emphasis during recent years [37-44]. This has been attributed to the increasing'complexity of the more modern equipment and the increased reliability requirements which are associated with the trend toward reduced manning. Breakdown in the propelling machinery may mean the loss of ship availability (or even the loss of the vessel), which is a very serious matter for the owners and operators. Considerations other than reliability, such as fuel economy, weight, space, and first cost, which may seem to be important in the early stages of the design, later become surprisingly insignificant when compared with irritating and costly service interruptions which can result from inadequate reliability. Accordingly, developmental features should be proven ashore where failures are of little consequence as compared with failures at sea. The method of establishing ratings of the various power plant components should be analyzed for service and design margins so as to ensure the high degree of reliability required for the safety of the vessel. Assurance should be provided that reasonably conservative horsepower ratings are used for design purposes since in some cases there is a tendency for ratings to be stated as that obtained on block tests under ideal laboratory conditions as opposed to the lets-perfect environmental conditions that are encountered in marine service. Evaluating the service and design margins is d i c u l t ; the type of fuels and the pressures, temperatures, and pressure ratios used in the design have a significant effect on the plant reliability. However, realistic trade-off studies require that either the degree of conservatism be consistent between various candidate power plants or an allowance be made for the differences. 5.3 Mai~tainability. Both preventive maintenance and correctiye'maintenance requirements must be considered in selecting the type of machinery to be used in a propulsion plant [&50]. Preventive maintenance has a direct impact on manning levels and operating costs. If the equipment installed requires frequent preventative maintenance, such as greasing, packing, cleaning, and parts replacement, crew personnel must be provided to carry out these duties. This is an important consideration as the cost associated with one crew member over the l i e of a ship is a startling sum, particularly if he must be highly skilled; additionally, the cost of the materials required for preventive maintenance adds to
I I
operating costa and can become significant especially when special tools and equipment are required. Corrective maintenance must also be considered in light of the manning requirements (as regards both manpower and skill level), materials, and tools required. Furthermore, the various modes of equipment failure should be studied in order to identify the failure modes which would adversely affect the propulsion plant operation (the effect could be either in terms of performance degradation, corrective maintenance requirements, downtime, or a combination of these considerations). Failure modes which have unacceptably adverse effects should be further analyzed to identify methods of reducing the likelihood or consequences of their occurrence (e.g., by means of redundancy or selecting other design alternatives). 5.4 Spare and Arrangement Requirements. Some years ago the minimum space required for the machinery plant of a merchant ship was a relatively unimportant consideration due to the tonnage laws in effect a t that time. Formerly, if the actual propelling machinery space exceeded 13,percent of the groas tonnage of the ship, then 32 percent of the gross tonnage of the ship could be deducted in computing the net tonnage, which is the basis for tax assessments, harbor and canal dues, etc. As a result, a special effort was then made to ensure that the space required for the propelling machinery was a t least 13 percent of the gross tonnage of the ship. The tonnage laws have subsequently been modified, however, and such an artificial condition no longer exists. I n most ship desigd configurations, an intensive effort is made to minimize the space required for the propulsion plant. In general, the space required for the machinery space is considered to be deducted from that which can be used for other purposes (e.g., carrying cargo); and a maximum effort is accordingly made to restrain the dimensions of the machinery space. In some ships, such as tankers, this is not as critical a factor. Minimum space requirements are almost impossible to generalize satisfactorily for different types of power plants. There is no substitute for making at least a preliminary ship arrangement layout to determine the effect of the power plant on the overall machinerv mace oonfiguration. In order to illustrate general dff%nces in this respect between principal propulsion plant types, representative machinery arrangements in typical merohant vessels are shown in Figs. 6,. 7,. 8, and 9 for a slteam turbine, diesel, nuclear, and a gas turbine plant respectively. There is a wide range of flexibility in the design of the propulsion plants illustrated; therefok, the oonfigurations shown should only be considered representative. 5.5 Weight Requirements; The importance of the weight of a main propulsion plant varies depending upon the particular application. I n the case of tankers, whose cargo capacity is limited by draft restrictions, the weight of the main propulsion machinery represents oargo foregone. Cargo vessels, on the other hand,
13
seldom operate at their full load draft; furthermore, they have chronic stability problems due in part to the extensive amount of cargo handling gear located high on the ship. As a result, the weight associated with the main propulsion machinery, as such, is mildly advantageous in that it improves the stability of the ship. I n general, naval vessels have chronic weight problems, particularly since the advent of the major emphasis on shock resistance; and shipboard equipment is carefully analyzed from the viewpoint of weight reduction. Representative pmpulsion plant weights (without fuel) are shown in Fig. 10, where the specific weight (the weight of the complete propulsion plant per unit of rated shaft horsepower) is plotted versus shaft horsepower rating. Representative propulsion plant weights, including fuel, versus the plant shaft horsepower rating are shown in Fig. 11. This plot permits a proper comparison to be made between petroleum-fueled plants and nuclear plants; for the latter the weight of fuel is not significant. Propulsion plant weights have been greatly reduced over the years. This trend is expected to continue, particularly as regards nuclear plants, due to the relatively large amount i5f research and development expended on this type of plant. 5.6 Type of Fuel Required. Although solid and gaseous fuels (coal, uranium, and natural gas) play important roles in worldwide energy production, by far the greatest proportion of the fuel buined aboard ships is petroleum fuels. Virtually all petroleum fuels are obtained by fractionating or cracking crude oils obtained from the world's various oil wells. There is a wide spectrum of petroleum fuels from which a choice may be made; some of the more important alternatives are given in Table 2. Table 2 TYPEOF DISTILLATE Light
Petroleum Distillates and Their Uses CLA~SIFICATION Intermediate naphthas Kerosene
Medium Heav
&~du&
Gas oil ~ubricatin~ oils Residual fuel oils Refinery sludges
COMMON UNRESTRICTED USES Aviation gasoline Motor gasoline Tractor fuel Gas turbine fuel Heating fuel Diesel fuel Not used as fuel Boiler fuel Refinery fuel
I n general, oils with higher viscosity are less expensive; however, an additional major consideration js that higher viscosity fuels have greater concentrations of impurities and harmful constituents. The fuel oil selected should be determined on the basis of the lowest overall cost, taking into consideration factors such as initial costs, handling costs, and equipment maintenance costs which can be attributed to the fuel. Factors which must be borne in mind, relative to handling and equipment costs, when selecting a petroleum fuel are fuel constituents, type of metals which will be
MARINE ENGINEERING
GENERAL CONSIDERATIONS
15
I
PLAN VlEW OF MACHINERY SPACE
PLAN VlEW OF MACHINERY SPACE
P L A N VlEW OF MACHINERY SPACE P L A N VlEW OF MACHINERY SPACE
ELEVATION
ELEVATION I BOILER 2. H.P. TURBINE 3. L.P TURBINE 4. REDUCTION GEAR 5. CENTRALCD~~TROLROOM S. STEAM TURBO-GENERATOR 7 WORKSHOP B. CONTAMINATED SThAM GENERATOR 9 DISTILLING PLANT
10. MAIN CONDENSER I I. MAlN CIRCULITING PUMP 12. 13. 14 15. 16. 17.
THRUST BEARIUG FORCED DRAFT FAN STEAM AIR HEATER UPTAKE DEAERATING FEED HEATER LUBE OIL SUMP TANK
Fig. 6 Steam turbine powor pknt
*
I. MAIN ENGINE
2. 3. 4 5. 6. 7. B.
THRUST BEARING MAIN ENGINE CONTROL CONSOLE MAIN SWITCHBOARD TuRBD-GENERATOR DIESEL GENERATOR DONKEY BOILER FUEL OIL PUMP
9 MAIN AIR COMPRESSOR 10 ENGINE I I. MAIN STARTING AIRFUEL TANUHEATER 12. DISTILLING PLANT I 3 LUBE OIL COOLER 14 AIR COMPRESSOR IS. WASTE HEAT BOILER I S EXHAUST SILENCER
Fig. 7 Low-speed diesel power plant
I. NUCLEAR STEAM GENERATOR 2. MAIN BLOWER TURBINE 3 AUX BLOWER 4. H P TURBINE 5. L.P TURBINE
6. 7. 8. S. 10.
REDUCTION GEAR TURBO-GENERATOR MAIN SWITCHBOARD WORKSHOP AIR COMPRESSOR
I I.
AIR TANK
12 13 14. 15. 1s 17. 18 19.
CONTROL CONSOLE DISTILLING PLANT CONTAMINATED STEAM GENERATOR DEMINERALIZER CAUSTIC B ACID STORAGE COMPONENT COOLIN0 SYSTEM LUBE CONDENSER OIL GRAVITY TANK MAIN
22 01
MAIN CIRCULATING THRUST BEARING PUMP
22. 23.
DEAERATING FEED TANK LUBE OIL SUMP TANK
I. MAIN ENGiNE
a. a. 4.
5. 6. 7.
a.
AIR INTAKE PLENUM EXHAUST DUCT COMPRESSOR GAS TURBINE REDUCTION GEAR LUBE OIL SUMP TANK STEAM TURBO-GENERAT~R
h.9
9. M I I N SWITCHBOARD 10. MAlN CONTROL CONSDLE I I. THRUST BEARING 12. PORT USE BOILER 13. DISTILLING PLANTS 14. DIESEL GENERATOR 15. WISTE HEAT BOILER 18. STEAM DRUM
Ggs turbine power plant
Fig. 8 Nuclear power plant
degradation being dependent upon the type of prime mover and its design parameters. It ia extremely important that fuel combustion technology be properly taken into account in any realistic appraisal of propulsion machinery life cycle costs and in the selection of an optimum fuel for a given set of circumstances. Much material has been published on economic oom~arisonsof ~uclearversus fossil fuels for shipboard we. These studies are clouded by the fact that the nuclear technology is subject to strong governmental Influence. The Atomic Energy Commission closely controls the manufacturing of nuclear fuels in the United Btates rigid licensing procedures; however, there several private firms which are engaged in the production of nuclear fuels.
I n the case of very large-capacity central station plants, where the cost of transporting coal is quite important, nuclear fuel has appromhed economic parity with f w i l fuels. However, ship power plants generally fall into a small-capacity category as compared to central station plants; consequently, widespread application of nuclear power in merchant ships will probably await further 'advancements in nuclear reactor practice and technology. Nuolear power for large naval ships is advantageous in that it eliminates the requirement of frequent refuelings, thereby aueenting the shipPs military effectivenew Nuclear power b p&iCularly advantageous in the case of submarines and has pmvided them with new dimensions of operability, submerged endurance, and military effectiveness.
16
GENERAL CONSIDERATIONS
MARINE ENGINEERING
The efficiency of gas turbine cycles is highly dependent upon factors such as the turbine inlet temperatures, the amount of regenerative heating, the pressure ratios, and methods of staging and matching the characteristics of the various compressors and turbines used. These are discussed in Chapter 6. 5.8
SHP RATING OF PROPULSION PLANT (THOUSANDSI
Fig. 12
I
I
201 I I I I 14 1 8 ' 2 2 26 50 34 38 42 SHP RATING OF PROPULSION PLANT (THOUSANDS)
Fig. 1 0 Specific weight of propulsion plants
8
SHP RATING OF PROPULSION PLANT (THOUSANDS)
Fig. 11
,
All-purpose fuel consumption
Weight of cargo ship propulsion machinery plus fuel for a 10,000mile voyage
As indicated in the foregoing, the selection of a fuel n e e success of the ship. An analysis of life cycle costs which fails to take the maintenance factors and other various aspects of the fuel selection into proper consideration would not be expected to be meaningful. 5.7 Fuel Consumption. Differing types of propulsion plants have inherently different thermal efficiencies and specific fuel consumption rates. A heat balance is the fundamental tool used i n determining the fuel consumption associated with a power plant, and it is given a detailed treatment in Chapter 2 for a steam turbine propulsion plant. Heat cycles related to other types of prime movers are discussed, to the extent deemed appropriate for a text of this sort, in the chapter applicable to the type of prime mover under consideration. The fuel consumption chmacteristics of various types
is a multifaceted process which may greatly i
of propulsion plants are expressed by Fig. 12, which illustrates the relationship between fuel consumption and size for the more usual propulsion plant alternatives. The fuel consumption indicated in Fig. 12 includes that required for the main propulsion plant, auxiliaries, and normal hotel loads; no allowance has been made for extraordinary service, such as the hotel load on passenger ships, cargo heating and tank cleaning on tankers, and cargo refrigeration. Figure 12 is not intended to be uaed as a substitute for detailed fuel consumption calculations; it is intended to illustrate only the general characteristics of the propulsion plant alternatives. Once the general type of propulsion plant has been tentatively selected, there are several design characteristics which may be selected to enhance the plant fuel consumption characteristics. For example, with regard to a steam turbine propulsion plant, regenerative feedwater heating using extraction steam or reheating of the steam in the boiler after a portion of expansion work has been extracted in the turbines typifies the methods by which the thermal efficiency of a steam cycle can be improved. In general, trade-off studies are required to determine the most appropriate steam cycle. Trade-off studies could consider such parameters as boiler superheater outlet pressure and temperature, condenser vacuum, main turbine efficiency, number of stages of regenerative feed heating, and selection of extraction points. In addition to the presentation made in Chapter 2, several excellent studies have been conducted and published which deal with the effect of cycle variations on machinery plant performance [51-571. These studies are useful in that they provide a sound basis upon which preliminary decisions can be made. Trade-off studies for the purpose of improving fuel economy should similarly be conducted with gas turbine or diesel propulsion plants. Cycles employing diesel engines tend to have higher thermal effioiencies than those employing steam turbines since the cycle works between greater temperature extremes; nevertheless, the overall efficiency of the total power plant can be improved by the use of waste-heat boilers or exhaust-gas turbines.
-
Fractional Power and Transient Performance.
Except for short periods when leaving or coming into port, most merchant vessels operate a t or near full power. Occasionally, the operating schedules include periods a t reduced speed that may be long enough to require special consideration, but such lowering of speed rarely goes below that corresponding to about one-half power. The case of naval vessels is entirely different. They are designed for high speeds for use on those occasions when speed is of great importance. However, most of the operating life of a naval vessel is spent a t moderate speeds, roughly about 60 percent of the maximum speed. Such cruising speeds require only about 20 percent of the normal power for which the machinery is designed. Good economy a t these low speeds is as important as at maximum speed, because it determines the cruising range of the vessel during many operations. I n high-powered naval vessels, therefore, specid provisions are made for economy at low ppwers. These usually include specially designed turbines (with cruising stages or stage arrangements which can be operated in series a t low powers and in parallel a t high powers), and auxiliary arrangements which are especially designed for economical operation at low powers. I n some instances the service requirements of a ship impose severe demands upon the propulsion plant. For oxample, special-purpose vessels may be required to operate for extended periods of time in an economical aruising mode, whereas upon command they may be mquired to reach maximum power in a matter of seconds, A special propulsion plant such as the combined-dieseland-gas-turbine arrangement described in reference [58] may be required to satisfy demands of this severity. 6.9 Interrelations with Auxiliaries. A considerable number of auxiliaries are required to serve the main tngines and for cargo support, cargo handling, ship kbndling, hotel load, and the like. Since in most instances there is a choice in selecting the type of prime mover for the auxiliary equipment, interrelations between the auxiliary equipment and the main propulsion plant must be considered in order to ensure that the dvorall ship is designed in the most effective mqnner. Auxiliaries can in general be driven by either steam or dectric power; when the main engines are driven by rteam, it may be desirable to also drive equipment such Y generators, pumps, and windlasses by steam. In the @$so of diesel and gas turbine drives, where steam is not Os readily available, electrically driven auxiliaries may be more appropriate. A supply of steam for heating purposes is required on moat vessels; the quantity depends on the type of vessel ~ n the d service for which it is intended. If the vessel is
17
steam driven, the supply is easily taken from the main boilers. For diesel or gas turbine driven ships, a boiler or boilers will have to be provided for that purpose. One economical method of doing this is to utilize the hot exhaust gases from the main engines by passing them through a boiler specially designed for this purpose. Such a boiler may also be provided with an oil burner to make up the deficiency, if any, and to operate in port when the main engines are shut down. I n tankers, where a large steam capacity is required for heating the cargo and rather large quantities of hot water are required for cleaning the cargo tanks, the boilers for steam-driven tankers may be significantly increased for this additional load. If the main propulsion plant is driven by a diesel or gas turbine, one or two large boilers may be required especially for this purpose. As may be seen, interrelations between the main machinery plant and the auxiliary equipment can be an essential consideration in the selection of the main propulsion plant. 5.10 Reversing Capability. The provision of means for stopping and reversing a ship is closely lrelated to the type of prime mover selected. Propulsion plants that utilize reciprocating steam engines, diesel engines, or electric motors present no problem in providing reversing capabilities because such components are intrinsically reversible. Steam turbines and gas turbines, on the other hand, cannot be directly reversed and require special provisions. The common solution with steam turbines is to provide special rows of astern blading in the exhaust end of the turbine (in the low-pressure region); in order to reverse, steam is admitted to the astern blading rather than the ahead blading. The solution with gas turbines is not as simple. It is generally not the practice to provide astern blading in gas turbines; therefore special provisions such as electric drives, reversing reduction gears, or reversible-pitch propellers must be provided. In cases where maneuverability requirements are severe (e.g., dredging vessels, tugboats, vessels which frequently pass through locks), controllable and reversible-pitch propellers may be used in conjunction with other types of prime movers [591. 5.1 1 Operating Personnel. The number and caliber of the personnel required to operate a main propulsion plant may be of major importance. Even though other considerf~tionsof a particular propulsion plant may be attractive, if difficulty is anticipated in obtaining suitable operating ~ersonnel, prudence may dictate that the plant be abandoned in deference to others. In the past, the general adoption of new types of machinery has been retarded as a consequence of this practical cogsideration. Over the years, fewer men have tended toward a seafaring life and as a result the total cost to man ships has risen sharply. An adequate number of highly trained men has not been available for ship manning and, as a result, propulsion plants have become increasingly more automated as a means of reducing the number of operating personnel required (see Chapter 21 for a
1+
MARINE ENGINEERING
years, there yet remains a limit to the size of diesel engine which is considered feasible. On the other hand, the rating of the propulsion plant, as such, does not impose a practical restraint on the size of a steam turbine
fig. 13 Relative imtalled cork of propulsion plank
J j c u ~ i o n of automation and controls). This is an effective means of reducing operating costs and is expe&d to continue. It should however be noted that automated ships will generally require more highly skilled operati~gpersonnel. This, in part, offsets the advantage of fewer personnel. turbine It is often said that the operation of machinery requires less engineeriog or mechanical skill than that required in connection with diesel engin% exceptto the extentthat This ie not entirely board maintenance of the main engine is carried outby to a shipboard personnel on diesel turbine &ips. The shorthigher degree than on treliability of steam turbines is usually considered and the turbine to be slightly better than diesel for short periods, that is, plant can sustain more maintenance of steamturbines can be postponed for short perioda in many instances. Diesel engines cannot be neglected without serious effects,and, flexibility of maintenance policies is not recommended for any typeof power plant, it is possibly less cmcial on the steam plant than the diesel. 5-12 Rating Lim;+dions. There are practical limits the power ranges in which the various which For example, typesof pmpu~sionplants are f-ible. which have been even though the rating of diesel has continued to increase over the installed
The ratings of propulsion machinery tend to be disCrete rather than continuous; consequently an additional rating limitation is imposed. As an example, gas turbine designs have been developed for a limited number of discrete ratings. If a gas turbine were desired with a rating different from those available, the cost associated with the development of such a special design would be pn>hibitive; the same situation exists, although to a ,gomewhat lesser extent due to the larger number of ratings available, with the other types of propulsion plants. 5.13 Costs. The installed cost, which is one of the most important considerations in making trade-off studies, is also the most volatile- Pro~ulsionplant price levels are strongly influenced by factors such as material and labor costs, the similarity of a plant with those previously produced, and ~ a n u f ~ t u r e rexisting 's work backlog, and therefore are subject to fiuctuations which depend on the current status of the industryof Nevertheless, the relative costs of the various plants along with the general relatiomhip of plant size and cost are illustrated in Fig- 13- The data presented in Figs. 10, 11, and 13 were largely taken from references
-
1% 611.
BY reviewing the factors enumerated in the foregoing which should be considered in selecting the type of marine propulsion plant, it may be noted that in every instance the fundamental issue is economics. There are three types of costs to be considered: initial (e.g., installed costs), recurring costs (e.g.9 fuel consumption), and contingency costs (e-g-,most aspects of B~ using a technique such as the present-value concept, the C O S ~ Sto be incurred in the future can be their present value So that all of the costs associated with the various design alternatives can be totaled and compared, in light of their contingencies, in arriving at the most advantageous alternative [62].
Setti011 6 Pnliminary Design Considerations 6.1 Introduction. Before the naval architect can firmly establish the dimensions, form, and charactervalues for the machinery space and istics of a weight, requirements, fuel consumption, and other engineehg quantities must be available to him. However, these quantities are dependent upon the vessel dimensions and form. I n order that the analysis may pmceed, tentative values must be selected initially and subsequently refined as the analysis progresses. Esti-
mates based on sophisticated procedures are warranted during the fomulative ~ h m e of s a design because the rapidly changing characteMcs of the supporting data are not commensurate with the accuracy of the calculation; overall methods of comparison which may involve the use of results from previous parametric studies or systematic ft3milies figuration are adequate and are more Preliminary design procedures differ so
,
GENERAL CONSIDERATIONS
19
one design organization to another that no routine pro- increase in initial pressure to increase the thermal codure can be described for this process. However, cycle efficiency 1 percent; or a 40 deg F increase in Home guides regarding specific methods of establishing temperature will have the same effect. Chapter 2 tho engineering features of a ship can be reviewed. But contains a detailed treatment of thermodynamics and it) order to proceed with s, typical example of further heat engineering considerations. dcsign selection steps, it becomes necessary to make It may be noted that the heat balance calculation is noveral presumptions. First, it is assumed that an well adapted to electronic computer calculation, permitoverall study similar to those described in Sections 2 ting parametric studies to be readily made. However, ~ m d3 has been used to establish the payload and s u 5 in providing component data to the computer, care must tained sea speed required of the vessel or vessels. be taken that it is reliable and accurate as the results will Second, it is assumed that the shaft horsepower required be no better than the data entered. The effect of the of the main propulsion plant has been established as following design variables on the thermal cycle efficiency, outlined in section 4. Lastly, main propulsion plant tempered by practical considerations, would normally trade-off studies, as described in Section 5, are con- be investigated at this point: nidered to have been conducted and, for the purpose of Boiler superheater outlet pressure and temperature this section, that a rather conventional cross-compound Condenser vacuum eared steam turbine propulsion plant has been identified Number of stages of regenerative feed heating and ILN the most advantageous type for the particular vessel best extraction points r~tldservice under consideration. Steam reheating in boiler Like other complicated engineering systems, much of a Main turbine efficiency nhip design is patterned after previous successful Turbogenerator efficiency (condensing versus practice. Ordinarily, pertinent plans of other ships, noncondensing) Nome perhaps rather similar to the one under consideraExhaust heat recovery from boiler stack gases Oio11,would be available. Also, essential data relating to Motor-driven versus steam-driven feed pumps Illlosevessels and important particulars of the machinery and auxiliaries ad auxiliaries, their characteristics, and their ratings Utilization of and balancing out of excess auxiliary would normally be available. If this information is exhaust steam properly compiled, it is often possible to make useful Desuperheated steam service requirements Arnt approximations without detailed study and thus Distillers, steam-air heaters, etc. reduce the range and number of variables that must be &on detailed study-to optimize a ship design. Of the foregoing design variables, the largest direct some of the more salient considerations in establishing gain in efficiency will come from increasing the boiler tho design of an engineering plant for a ship, in addition superheater outlet temperature and the boiler to the main propulsion plant trade-off studies described There are, however, several factors which cannot be 111 Roction 5, are reviewed in the following paragraphs. ignored; boiler design pressure must be increased in 6.2 Propulsion Plant Steam Cycle. The propulsion proper proportion with the temperature in order to plr~uthas been established to be of the steam turbine ensure that the turbine condition line does not lead to typo; however, the precise steam conditions and cycle excessive moisture in the exhaust end of the low~rrbngementwould warrant yet another review. The pressure turbine as an erosion problem could otherwise h e ~ tbalance calculation is the basic analysis tool for result. Furthermore, inerewing the boiler delurmining the effect of various steam cycles on the outlet temperature and the boiler efficiency beyond tharmal efficiency of the plant. Standard practices certain limits both lead to costly increases in either the atrd allowances which are recommended in the prepara- boiler design or its mainhnance, or both, which must be l ~ n l rof heat balances have been promulgated by the taken into account. When burning Bunker C residual ~ l l l p '~~a c h i n e r yCommittee of the Society and are fuel oil, eutectic combinations of oxides of vanadium, available in ~ e c h n i c a l kResearch Publication No. 3-1 1. sodium, and potassium can c a w slaggng and accelerated 111 tho absence of specSc component efficiencies and erosion of tubing at relatively low metal surface temurvioing allowances during preliminary design, the peratures. Thus if low-grade residual fuel is to be used, r@UXIlmendationsof this publication are most helpful. it must either be treated aboard ship, or the boiler must Many excellent parametric studies have been con- be specially designed to limit the metallic wall tempersduotml by various design agencies and several have been tures of the superheater tubes and supports; additionally, ~ublinhed[51, 52, 531 which cover the effect of steam the boiler should be designed for ready acceas into the @~adltions and cycle variations on machinery plant per- superheater for mechanical de-slagging, cleaning, and f@lmalce. These may be used as a guide during initial tube replacement. *l@otion and thus minimize the amount of detailed work Another factor which should give rise a cautious ah& must be carried out later during the more refined approach in moving to higher design initial pressures @@WO the design- Fmm parametric studies of this and temperatures is the increasing cost and difficulty in Rnl'tt ollc can derive some useful yardsticks for design assuring the safety and longevity of steam piping, n@lailiOn, such as, for example, that it takes an 85-psig joints, valves, fittings, manifolds, and pressure bound-
MARINE ENGINEERING
GENERAL CONSIDERATIONS
21
GENERAL CONSIDERATIONS
23
MARINE ENGINEERING
desisns vary widely depending upon the type of cargo handled [6&75]; however, some of the more common types are as follows: Winching system, burtoning or swinging booms for dry cargo, i-e., break-bulk cargo or palletized cargo systems Cargo crane systems, either shipboard or onehore ~ u l cargo k systems, such as self-unloadem utilizing either standardized Container containers which lift On/& or standard truck trailers which roll on/off B~~~~systems, utilizing hrges which either lift on/off or float on/off systems, utilizing cargo ~ i ~ ~ i tankering d piping, pumps, and so forth Barge raftlngsystems, ut&ing pushboats or
certain limik), radar su~eillance and warning collision hazards (also within cedain limits), data monitoring and recording of principal voyage data, weather reporting, sounding, and fire detection. Some of the facets of navigation which do not appear to be readily adaptable to automation are: docking and undocking; piloting in nanow channels) harbors, Or , territorial waters where local knowledge is emntial; planning and laying-out of best course and speed, taking into account all potential factors; decisions on slowing Or proceeding with due caution in poor visibility ;maneuvering to prevent collision, determination of safe sea speed, determination of best fix from position fixes, and judgment as to when to post lookouts in foul weatherEngine room control stations appear perfectly feasible to permit the handling of even complex plants by a single licensed officer. The gas turbine and the diesel
The relative of the in selecting the types of cargo handlhg gear, such as the winch desip aeociated with different rigging schemes, hydrau]$ally operated hatch covers, special types of cranes, elevators, conveyors, and cargo pumping systems, should be given a rigorous analysis during the preliminary desisn shge. Close cooperation between the naval archit& and the marine engineer is essential in such and power requirements analyees. The space, be estimated very early in the design of a ship as they may have an important impact on the deck arrangement, the size of the electrical generating plant, and indeed the configuration of the vessel itself. Ca%o refrigeration, cargo hesting, ballasting requiremenh related to cargo handling, buttemorthing, etc., are all imporbnt servke load factors which may result in peak loads not only on the electrical generating plant, but also on the main machinery plant. They must then be included in the design heat balances and electric load anslyses from the emliest stage of the design selection process. For a detailed discu$sion of dry, bulk, and liquid cargo handling systems, see Chapter 16. 6.17 Autorntion rnnd Mechanization. Automation and rnechaoieation of shipboard processes are important means of improving the efficiency of ship operation. These are subjects that are particularly well suited to system engineering analyses in that the cost of developmerit, manufacture, installation, and maintenance of such mechanized or automated equipment is readily compared to the cost of hand labor. However, close sight must be kept on the degree of reliability of autornation where it involves the safety and security of the vessel; furthemore, the training and adjustment of maritime labor to new conditions of operation must be rnnsidered in addition to simple engineering feasibility. There is potential for reducing the burden of bridge duty and reducing the number of operating personnel required for the saf. navigation of the ship by the intraduction of rnonitonng and control devices in a bridge coneole. Some of the facets of navigation which are adaptable to automation and semi-automation are: course steering, dead reckoning, position-fixing (within
engine appear particulady well adapted to automation becsuse of the Simplicity of their control- However, even the steamship with its more complicated plant has been automated to a surprbing degree and developments (see Chapter 21). in this direction continue to be 6.18 Dynamic Effecfso Dynamic effects, principally mechanical vibration but also noise and shock resistance, must be an integral aspect of the preliminary design process as the dynamic cha~acteristicsof the ship and the dynamic requirements for equipment am largely established during the preliminary design stages* The objective is to develop the design so that the desired dynamic Characteristicscan be achieved in an effective manner. Reafisticall~ conceived requirements with regard to dynamic effects require careful and adequate planning during the preliminary design stages in that they may be met without excessive dimculty or undue expense. especially important insofaras Vibration analyses
..
the design of the pmpulsion shafting system is concerned, and particularly its relationship to the excitation forces resulting from the propeller operating in a nonuniform wake. Propeller exciting forces are diicussed in detail in reference 1761 and main propulsion shafting systems in Chapter 11 of this text. As may be noted in the latter, the main propulsion shafting can vibrate in longitudinal, torsional) and lateral modes. Each mode of vibration must be dealt with during the early stages of design. Modes of vibration of the ship's hull as a whole (i.e., as a free-free beam) are discussed in reference 1771. T h m may be vertical, horizontal, torsional, or longtudinal and may occur separately or, in rare case*, coupled. The calculation and re diction of the hull vibration modes is quite complex since the hull girder is far from a Simple homogeneous beam. Hull vibration of this type, may be excited by s~nchronirationwith periodic harmonics of the ~ r o ~ e l l forces er acting either through the shafting, by the ~mpellerforce field interl acting with the hull afterbody, or both. ~ u lvibration may also be set up by unbalanced harmonic forces from the main machinery, and in some cases by impact excitation from slamming or ~eriodicwave encounter.
GENERAL CONSIDERATIONS
Tab* ITEM
Mg~lltlllmrated power
'r'ho(+(Lsteam condition ( !otldnnmr vacuum
listof Machinery for RATING
One set, cross compound, with astern element located in exh u tcasing end of low-pressure turbine 24,000 at 105 rpm 850 F 28.5 in. H 925 g 8t maximum rated power
a
24,000~~h,, cargo
ITEM Line Shaft Beanng8
gbr Diameter
Length Materid Stem Tube Bean'ng Type Length
RATING 8
Re laceable shell, ring oiled 21& in. In. 32%
Caet atex?] pedestal, cover and Oil lubricated 27 in. forward bearing 54 in. aft bearing Ductile iron and babbitt
Vertical, walk-in, five pas4 convection with automatic superheat control by desuperheater Coil in steam drum
20,000 Ib/hr from 875 p i g , 930 F to 775 psig, 575 F
345,000lb at 24,000 shp and 105 rpm
Aft of low-speed gear cssing
27% in.
2 4 f t 11Xin. Solid forged steel, ABS Gr. 2
Superheat Control De.guper& Descr~pt~on Coil in steam drum, steam Burnera after third superheater pms Number her boiler 3 Type Wide range ateam atomking
Air quantity, cfm Air temp in-out, deg F Stm Pr=.-temp, pslgdeg F Air press, drop, in. H ~ O
Ruting
116%
23,500
10048.3
29,400 100-275
62-453
62-453
0.6
1.0
GENERAL CONSIDERATIONS
MARIYE ENGINEERING Table 3 (continued) RATING
RATING
ITEM
~ l ~ ~ t ~ - m e c h adeck n i dmounted , 50 hp, 650 rpm, 230 volt d-c
20-ton cargo hoist 1 ~ t o cargo n hoist
14,500 lb at 105 fpm 18,000 lb at 85 fpm 8,800 lb at 185 fpm 14,200 lb at 116 fpm 8,800 lti at 85 fpm 1250 ft of 76 in. wire rope 800 f t of M in. wire rope
eretors
Drum storwe
Section 7 Specifications
,
31
MARINE ENGINEERING
GENkkAl CONSIDERA'I'IONS
Section 8 Final Design and Working Plans
16
Lifesaving Equip-
65 . Air Conditioningand
33
MARINE ENGINEERING
GENERAL CONSIDERATIONS
of contracts for ordinary merchant vessels where the plans must be developed in a short time. Where oOmposite Plans are not made, the elimination of
interferences and the treatment of wstems in accordance with their relative importance must be accomplished by the cooperation of the various design groups iivolved.
Section 9 The design and construction of a ship is culminated by Sea trials are conducted as a means of demonstrating broad array of tests which demonstrate that the ship is the adequacy and perfomance of those aspectsof a ship in accordance with contract requirements. At the lower which cannot be realistically tested at dockside. sea the test spectrum are those of a q ~ a l i @ - C ~ n t r ~trials l are bmadly classified into twogroups; namely, "ature which are conducted to ensure conformance of machinery trials and maneuvering trials. ~h~ former lnaterial properties to specified requirements, soundness deals with the mechanical and economical performance of cmtings, dimensional accuracy, and the like. Tests of the boders, the proeelling machinery and their nuch these are not Peculiar to marine equipment and auxiliaries, and tests of evapowtors and distillers, Ihu standard quality-control Practices of the manu- together with the anchor and steeringgear and Iaaturer Or are generally relied upon to other equipment which cannot be tested uader actual nrluure the adequacy of equipment in this regard. conditions at the dock. The latter involves calibration Shop and installation tests include those tests which of navigating equipment, the of the n o m a l l ~conducted in the shop after assembly or in ship, and the speed-power characteristics of the ship. tho ~esselat dockside after the installation of the Tests typically conducted during sea trials are as equipment or system to be tested is substantially follows: aamplete. These tests are conducted to prove correct Calibration of navigating equipment rflnombly and proper installation and to demonstrate Speed-power-rpm standardization tests that control and safety devices are functional and properly adjusted. References [81, 821 contain general Economy power teats Full-power endurance tests guidelines which may be used in connection with shop Ahead steej n g tests ihd installation tests for merchant ships; similar, Quick reversal astern and head reach although generally more exhaustive, test requirements Astern endurance tests I0r naval ships are invoked in the specifications prepared Astern steering tests for eech particular ship. Quick reversal ahead and stern reach Anchor windlass tests which strength is a major concern. 8.8 Electric Plant. The procedure for the final '*lo COnektion Of Detail pian'' The design of the electrical installation roughly parallels that A careful review is made of of a ship are made by a large number of for the other working simultaneously in several drafting departmentsthe probable electrical loads and the selected number The administration and practice of the dr*ting organizaand rating of ship,s service generators and emergency tion must aim at complete elimination of physical generators. Vendors, pmposals are obtained and r e interferences between various parts and at a design in viewed for correlation with the general design. which each element is treated in acc~rdancewith its The airing plans for power, lighting, and interior relativeimportance. W r e x a m ~ l e , a P o o r l e ~ o f v e n t ~ ~ ~ commu~cations mnsist of single line diapams and tion duct 4ould not be accepted merely because a deck arrangement plans. The single line diagrams in elementary form, the electrical interconnection perfect lead for a freshwater line or an electric cable is of the various parts of each system. The diagrams desired' sections of the of the cables and c o n d u c t o ~ It is Customary to make, for show the approximate machinery spaces, composite layouts showing everyalong the ship and through the decks. The deck thing in those Spaces; i.e., structure, machinery, arrangements show the wiring on each deck and the These may be to a correct location of all appliances, fixtures and fittings, Piping, ventilation, and scale larger than the ~ s u a arrangement l plans; and develop including radio and navigation equipment. possible interferences' "lVe to ing these wiring plans, consideration is given to carrying Or other large Occasionally, in the case of capacities and voltage drops, directness and simplicity of leads, protection, support, and accessibility. important vessels, such composite layouts are made cf 8.9 null Machinew. The marine engineer is usually practically all machinery spaces This procedure is, concerned with the deck machinery and other mechanical however, slow and costly and cannot be afforded in the
36
MARINE ENGINEERING
GENERAL CONSIDERATIONS
37
An Anal~si*' Naval Engineers Journal, 64 D. M. Mack-Florist and R. H~~~~~~~ dlAn Economio February 1965. Feasibility Study of U n i w States Bulk Carriers, 49 A' J' Ruffini~ standard Navy Maintenance Marine Technology,vol. 3, no. 2, ~ ~ r1966, i l and Management system (3-M System),JJ 65 W. j . Dormm, 'dcombimtion Bulk ,, Bureau Of ships Association of senior Engineers, March Marine Technology,vo~.3, no. 4, October 1966. 66 A. W. Feck andTankem J. 0. Sommerhalder, 'Cargo,, 50 A. Goldman and T. B. Slattery, Maintainability: pumping in M~~~~ and Bulk Carriers, A Majw of SYskm Efectiveness, John Wiley & ~ ~T r ~ i 4, no. ~ ~ July, 1967. ~ ~ ~ Sons, New York, 1964. 67 Leslie A. Harlander, "Further Developmenh of a 51 W' Giblon and Cheater W' "Effect container System1961. for the West Coast-Hawaiian T ~ ~ Of Conditions and Cycle Arrangement on Marine Trans. Power-P1ant Performance as ~eterminedby the Elec68 James J. Henry and Henry J. Kamch, ,,Container tronic Computer," Trans. SNAME, 1961. 52 H- M. Cheng and C. E. Dart, "Cycle and Ships, " Trans. SNAME, 1966, 69 5'. G- EbelJ "An Analysis of Shipboard cargo Economic Studies for a 25,000-Maximum-S~pSteam Power Plant for Singlescrew Tanker InstsllationJ Cranes, " Trans. SNAME, 1958. Trans. SNAME, 1958. 70 E. Scott Dillon, Francis G. Ebel, and Andrew R. 53 M. L. Ireland, Jr., H. W. &marJ and N. L. Goobeck, "Ship Design for Improved Cargo Handling, Trans. SNAMEJ 1962. Mochel, "Higher Steam Conditions for ShipsJ 0hiner3'JJJ paper presented to the International Con71 John F. Meissner, "World Development and foreace of Naval Architects and Marine Engineers, 1951. Movement of Iron Ore, Trans. SNAME, 1962. 54 W. L. Coventry, "Fundamentah of Steam 72 -Harry Benfod, Kent C. Thorntan, and E. B. Turbine The-odynami~s,' Trans. Institute of Ma* Williams, "Current Trends in the Design of rron-ore Bngineers, 1962. Ships, " Trans. SNAME, 1962. JJ
JJ
JJ
JJ
Trans. SNAME, 1965.
*
THERMODYNAMICS AND HEAT ENGINEERING
C H A P T E R II
-O,,,(HEAT
TRANSFERREDI
1.1 Basic Equations. The applied thermodynamics problems of marine engineering depend on the conserve tion of mass and the conservation of energy. The first of these is conveniently expressed by the sional steady-flow continuity equation
h?
+ 9+
Q1.2
=
4
h t i-iwt1,2
Typical applications of the general energy equation occur where the working floid is being heated without work being done (a heat exchanger), where work is being done under adihbatic conditions (turbi~le wheel), W = AC/v ('1 or +here mechanical energy is being degraded under adiabatic conditions and without work being done (flow against friction). The equations that apply in tbese situations are esaily found by eliminating the inapproA = flow area, sq ft priate terms from equation (4). An application is C = flow velocity, fps illustrated by Fig. 1. v = specific volume of the fluid, cu ft/lb Evaluation of the general energy equation usually W s flow rate, lb/sec requires assistance from other equations. The conThe second is conveniently expressed for the usual shady tinuity equation is one. Equations of state for the one-dimensional situationby the general energy equation fluid involved are also frequently needed. The simplest form is the familiar perfect gas equation
1
+
Enemy equation as applied la a single-stage turbine
Typical values of R are 53.34 for dry air; 53.5 for wet air (40 percent humidity, 100 F); and 50.3 for flue gm (15 percent excess wet air and standard fuel oil)For perfect gases, the following state relations also hold :
-
I Numbem
in brackete designate References at end of ckpter.
(10)
Carbon. . . . . . . . . . . . . . Hydrogen. . . . . . . . . . . sulfllr... . . . . . . . . . . . . Oxygen. . . . . . . . . . . . . Nitrogen ... . . .. . . .. . Free moisture. . . . . . . .
0.8775 0.1050 0.0120 0.0040 0.0015
-
Charts PI, must be used. An alternative, particularly adaptable to turbine design work when calculations are 1.0000 made by Computer, is to use the equhons from which Other properties of flue w, such as its viacasity and these tabulations are made. thermal are also needed, and are given in Special relations for steam that are useful in nozzle Fig. 4. conductivity, values for steamand air can be found in the design are the equation of state Steam Tables [ll and Gas Tables [2], respectively. pv = 1.222 (h - 823)' 1.2 Heat Transfer. An investigation of the & * (') term in equations (2) or (4) entails a consideration of the and the equation for isentropic expansion principles of heat transfer. The transfer takes place by molecular diffusion between bodies in contact, or by pl.s = constant electromagnetic radiation between separated bodies. The following two are the corresponding relations for Diffusion between solids is c d e d menone the wet region or both of the bodies are fluids, conduction is nearly
T = absolute temperature, deg R R = a constant characteristic of a particular gsa P = pressure, psf J = mechanical equivalent of heat = 778 ft-lb/Btu g = gravitational constant = 32.17 ft/sec2 z = height above an arbitrary datum, ft Q = heat transferred, Btu/lb W r = external work done, Btu/lb
p0.07v= 0.467 (h - 366) p ~ . ~= s constant
(11) Units are psi for p, cu ft/lb for v, and Btu/lb for h. These equations are for use only in the vicinity of normal turbine state h e s , and not for use at high superheat with low Pressure, with very wet steam, or in the reheat region. In boiler design work, the sensible heat, and specific heat of the flue gas must be known. These are presented in Figs. 2 and 3 for a standard grade 6 or residual fuel oil of the composition (by weight) tabulated below when burned in air with a 40 percent relative humidity at a temperature of 100 F.
J. B. Woodwad, III
Review of Fundamentals
39
Re. 2 Selulbk heat of gases
MARINE ENGINEERING
THERMODYNAMICS AND HEAT ENGINEERING
where the subscripts o and i designate the outside and inside surfaces of the tube. Heat transfer problems frequently involve conduction through successive layers of distinctly different conductivity. Formulas for this type of problem are readily derived, as are formulas for the transfer of heat through cylindric composite walls. b. Convection. The convective heat transfer between a fluid at a largely constant bulk tempe~atu~e TB and a surface at temperature T is expressed by
Q = h j # ( T ~- Ts)
property of the material conducting the heat. It is generally a function of temperature, particularly for liquids snd gases, but the effect of temperature is sufficiently weak that conductivity can be treated as a constant in most problems. Fourier's Law can be expressed for one-dimensional problem as
(16)
where hj is the film coeficient of convective heat transfer. major practical problem in applying equation (16) in the evaluation of the fdm coefficient for the several distinct mechanisms of flow and thermal behavior possible in the fluid. Single-phase convection occurs when the fluid involved uoither boils nor condenses at the solid surface. Familiar axamples abound aboard ship; for instance the water side of condenser tubes, both sides of the tubes in liquid-toliquid heat exchangers such as lube oil coolers, and the gtM side of convective heating surface in boilers are typical locations where this mechanism is prominent. Tho value of h, is generally a function of fluid properties, of the fluid' velocity, and of its degree of turbulence. Under conditions existing in a typical condenser tube, far example, the value of h is likely to be in the neighborhood of 1000 Btu/hr-sq ft-deg F, while on the gas side of r boiler tube, the value of h j can be 10 Btu/hrmsq ft-
tho situation is described aa forced convection. When
k
= conductivity, Btu-ft/hr-sq ft-deg F' = conducting area, sq f t
either dropwise or film condensation. The names are quite descriptive of the processes. The rate of heat transfer is much higher for dropwise condensation, and is comparable to that for nucleate boiling, since the drops quickly fall off as they form and thereby expose the surface to more vapor. In film condensation, the condensed film tends to cling evenly to the surface, and so forms a barrier between the surface and the vapor. C. Radiation. All matter emits radiation of one or more kinds. The thermal radiation of practical concern requires only that the matter be at a temperature above absolute zero, and so is characteristic of all bodies. Thp radiation is electromagnetic, and at industrial temperatures lies within the infrared part of the electromagnetic spectrum; but the wavelength is a function of temperature, and at higher temperatures it falls within the range of visible light. The radiation is not, however, monochromatic. A curve of its intensity, IA, against wavelength, A, shows a considerable spread with a peak intensity at a wavelength that is a function of temperature. The total energy emitted is thus the integral of IAover all wavelengths. For a black-body radiator, i.e., one that emits at the maximum intensity at all wavelengths, the integration produces the Stefan-Boltzmann relation (17) for T in degrees Rankine and Eb in Btu/sq ft-hr. But actual bodies are not black-body radiators, and their degree of imperfection must be accounted for by equation (I7) becomes
[A] 4
E
= 1730.
Bodies for which this equation holds are said to be g ~ e g
k is constant, this equation can be integrated for a slab of thickness x, having a temperature difference between faces of TI - Tal to obtain
~f
If the conducting body is circular, as when heat is transferred through tube walls, equation (12)is modified
always -ly modified by the transport of heat by where r is the radius dimension, and fluid in motion; this phenomenon called convection. tube. Integration of equation a. Conduction. Conduction follows Fourier's Law1 which states that heat is diffused at a rate proportional to the temperature gradient; the factor of proportiond t y is known the the~malconductwity, and is a
density differencescaused by expansion or contraction of the fluid near the surface are the principal source of the driving force, the situation is described as natu~alcon-
41
is the length of the
ges
MARINE ENGINEERING
-TI
I n pract.ice, heat exchange by radiation occurs between bodies of different temperatures and different emissivities. The situation is complex because the geometrical arrangements and sizes of the bodies are significant. For an elementary case of two parallel infinite planes, and of respective temperatures and emissivities Ti, €1, T2, e2, the net energy exchange rate is
I
THERMODYNAMICS AND HEAT ENGINEERING
--
-11-
C
COUNTERFLOW
-1
1
-
l"b
PARALLEL FLOW
For a sphere or cylinder, enclosing a smaller sphere or cylinder, the equation is Fig. 5
Simple counterflow and parallel-flow heat exchangers
to the heat that it receives or rejects. If the fluids on both the hot and cold sides of the heat exchanger undergo a change of state, their respective temperatures are constant, and equation (21) applies without change if S is understood to mean the 'total heat transfer area. On the other hand, if there are temperature changes, the temperature difference in equation (21) is not constant throughout the heat exchanger, and in consequence this equation must be integrated for application to the entire apparatus. The case where there is no change of state is illustrated by a simple concentric-pipe heat exchanger, Fig. 5, in which the two fluids flow either in the same direction (parallel flow) or in opposite directions (counterflow). The temperatures of the two fluids are plotted as a function of position for both exchangers. Such a temperature differenceintegrated over the length of the heat exchanger produces a mean temperature difference;because of its logarithmic term it is familiarly known as the log mean temperature difference. I n the
43
general case, the log mean temperature difference can be written as AT,
=
- ATmin AT, log. ATmin
AT,,,,
(23)
Equation (23) is the general expression for AT, for both simple counterflow and parallel-flow exchangers. I n condensers, boilers, and feed heaters, to list several prominent examples, where a change of 'state rather than a temperature change occurs on one side of the tube wall, a derivation of the log mean temperature difference again produces equation (23). If the heat exchanger is multipass, equation (23) must be modified (see Section 2.1 of Chapter 14). I n any case, equation (21)) when applied to the heat exchanger as a whole, is written as
9 = USAT,
(24)
contributions will be additive. Thus, starting at the I n boiler tube banks where the heat transfer fluid is a radiating gas, heat transfer simultaneously Occurs by both radiation and convection. Under these conditions (in order for the two heat transfer coefficients to be directly additive) it is often convenient to express the radiation heat transfer in the form of the artificial heat transfer coefficient
Section 2 Heat Transfer in Boilers where hrl, hj2 = convective surface coefficients at tube out-
T,, ti = metal temperatures at tube outside and inside surfaces respectively k = conductivity of the tube wall X, = equivalent thickness for the circular tube
where the subscript G refers to the radiating gas and 8 refers to the tube surface (see Subsection 2.3 for further discussion on this subject). d. Overall Heat Transfer Coefficient. The typical occurrence of heat transfer in power plant apparatus is cold fluid through an inter- Addition eliminates intermediate temperatures, do-g between a hot fluid and a vening tube wall. Convection and radiation are in- assessment of U as volved at the inner and outer surfaces, and conduction is involved within the tube metal. The rate of heat flow is summarized succinctly by $ = US(T - t) (21) BYa similar pmoess, U can be written for any number of layers. The practical pmcess of heat exchanger design is where T and t are the bulk temperatures of the two fluids, S is the surface area, and U is the ooerd heat transfer often aided by ern~irioalformulas for U which $ve coefident. U is the net effect of the conduction, con- results of suffcient accuracy for industrial purposes. vection, and radiation contributions. To illustrate the They are usually ~ r o m u l ~ a t e dby man~fa~turer's ~1 makeup of U, consider the transfer of heat from a hot associations to standardize methods of ~ a l ~ u l a & and are found in publications such as references [Bland [71. clean tube to a second fluid inside the tube. gss outside As a preliminary, note that the artificial radiation e. Log Mean Temperame Difference. The fluid coefficient ic, of the same dimensions as the convection flowing through a heat. exchanger undergoes either a coefficient hj, is used so that the radiation and convection change in temperature or a change in state in response
[(&)( (&)I + + [ + ++
2.1 Types of Heat Transfer in Boilers. A boiler may be divided functionally into four parts: first, a ~~~OSCFEFA USw(T8 - Tc) chemical reaction chamber where the chemical heat of LHV q~ (to - to)CpR fuel combustion k released and the reaction controlled; = WF(R 1) 1 R second, a steam generating section where heat is transferred to the tubes by radiation, convection, and con(25) duction; th,ird, a superheater, where the steam is super- where heated to the desired degree; and fourth, a heat recovery U = convection heat transfer coefficient section, employing air heaters and/or econombers T c = furnice surface temperature where some of the remaining heat in the flue gas is exTB = furnace exit temperature T F = effective flame radiating temperature S w = convection surface area
where an overall U is estimated and an exit temperature from each bank of tubes is calcdated. The designer must h t estimate the performance of the furnace and 2-2 Heat Tmnsfqr in Boiler Furnaces. Furnace heat transfer is principally radiation, and it is possible to b p t the basic methods of Hottel in reference [5] to evaluate a tohl emissivity in terms of furnace conditions. The problem consists of equating the heat given up by the omb bust ion gases to the heat transferred by radiaflon and convection to the f b a c e surfaces. The
q F = sensible heat of fuel above to sensible heat of gas above to C, = average specific heat of combustion air R = air-fuel ratio FA= arrangement factor FE= emissivity factor
Q T = ~
The heat given up by the gas is evduated by ordinary s t o i c ~ i o m e t ~ means c and the use of a set of sensible heat cumes (fig. 2). The shape emissivity factor, FEFA,has been treated by Hottel, and if the flame fills the furnace, it has been demonstrated that
,
MARINE ENGINEERING
44
0
FLAME EMISSIVITY, EF Fig. 6 Shape emiuivity factor versus Aame emiuivify fw various valuer of cooled surface to cooled surface plus refractory surface ratio (Sc/Srl
THERMODYNAMICS
by the chemical breakdown of the fuel to basic constituents. The flame mass then consists of a cloud of flaming fuel, carbon, some ash particles, and molecules of carbon dioxide, water vapor, sulfur dioxide, oxygen, and nitrogen. Of these constituents, the fuel, carbon, and ash particles and the carbon dioxide, water vapor, and sulfur dioxide molecules radiate. The gas molecules radiate only in certain wavelengths, that &, they are not grey. The solid particles radiate in all wavelengths. These radiations are superimposed upon each other, resulting in an overall radiation which is essentiay grey in character, and the resulting emissivity is independent of temperatufe. Combustion of oil is not instanta~eous,especially when residual oils are fired. The oil droplet first ignites, then burns and breaks down into carbon and hydrogen. The carbon appears aa minute flecks. These small particles make up most of the radiation. Their concentration is a function of burning time, and of the rate of flow of the gases through'the furnace. An expression derived for cp by applying probability theory is
"
TRANSVERSE TUBE DIP,PITCH 0
2
4
6
8
10
12
14
18
18
20
22
24
FIRING DENSITY, WFIPF VF. LB/FT~-HR-ATM
Fig. 8 Wectiveneo factors f a water walls bared on ma1projected arm
Fig. 7 Furnace concentration factor
(26) where fa
= emissivity of a cloud of i n f i ~ t ethickness,
assumed to be 0.95 (27)
where ec = emissivity of the heat absorbing surface e p = flame emissivity
S B = refractory surface area S c = cooled surface area F R C= a geometric factor, dependent on the extent of cooled surface
An ?ppmximation of FRCto a reasonable degree of accuracy is SR when 0 < -- < 0.5 (28) FRC = g Sc
'
+
where ST = S R SC Faired intermediate values may be taken between the two sets of limits quoted, as illustrated by Fig. 6. The radiating temperature T p may be approximated by T p= ( T A ' T E ) ~ ~ ~ (30) where TA' = adiabatic flame temperature with 100 percent theoretical air. Evaluation of the flame total emissivity presents a complex problem. The flame cloud consists of droplets of fuel from the burner nosde which in turn are reduced to smaller fragments by various air and gas currents and
P = furnam pressure, atmospheres L = mean radiating path length; for ordinary marine furnaces, L = 0 . 6 m furnace volume, cu f t an empirical concentration factor, a function of a time parameter W p/PpVp qith WP representing the pounds of fuel burned per hour. Wp/P;Vp is a crude measure of article life but better data on the flame path is lacking. The concentration factor, K, is evaluated from test results on various boilers and plotted against the firing density WP/PPVF, as on Fig. 7. This plot was calculated from the test results on five different boilers, all burning residual fuels. The curve shown represents an average of the test results with 10 to 20 percent excess air. It is necessary dso to consider the question of effective cooled surface. A water wall consisting of tangent tubes may be treated as a surface having an area equal to the projected area of the surface. If the tubes are widely spaced, exposing the refractory surface behind the tubes, the simple projected area of the tubes is not sufficient since the refractory receives some of the direct radiation from the surface and returns only a portion of this heat to the furnace; the remainder goes to the tubes. The effective radiant heat absorbing surface (RHAS) may be calculated by multiplying the projected area of the walls, including backing refractory, by an arrangement factor from Fig. 8, for each area making up the furnace envelope.
Solution of equation (25) is best accomplished by trial-and-error methods by brealdog up the equation into three simultaneous equations, as follows:
+ U%(TB
- Tc)
+ -
(ta t*)CpR In equation (33), the term LHV.+ q~ a t 1 is the total sensible heat released to the furnace per pound of combustion products and may be replaced by qpA, the adiabatic sensible heat. q ~ may , be read from Fig. 2 at any assumed vdue of T g . With these simplifications, equation (33) reduces to ,-..A
(32)
The solution may then be achieved by assuming vdues of T B and plotting solutions for equations (35) and (36). The point of intersection of the two equations is the solution. T A may ~ be evaluated by cdculating the adiabatic sensible heat
The term U s- w ( T E - T c ) in aquation (32) is Sc generally negligible except for rear waIl impingement effects. It is convenient to drop the term at this point and correct for the effect later. For most marine boilers the temperature of the radiant heat absorbing surface (RHAs~is close to 1000 R (540 F ) , so the term Tc/1000 is approximately unity. Since the value of Tp/1000 is between 3 and 4, the relative value of ( T ~ / 1 0 0 0is) ~so much higher than 1 that the term Tc/1000 can be taken as equal to 1 with little error. Further noting that T P = (TA.TE)112,equation (32) becomes
then Tnf may be read from Fig. 2. Usually, the convective term in equation (25) is negligible; but when a rear wall ie fitted, especially in a shallow furnace, the convection effect of the flame blasting against the rear wdl may be significant. An equation for the surface heat transfer coefficient WRW, based on the actual surface exposed to the gas, is
Where = Prandtl number k GD - = Reynolds number F
THERMODYNAMICS AND HEAT ENGINEERING
MARINE ENGINEERING
invariably inside the tubes, with some type of extended surface outside. There is a large variety of extendedsurface types, ranging from cast iron fins shrunk on steel tubes, to stud fins, aluminum fins, and spiralwelded steel fins. Such elements are proprietary in nature and performancedata must be obtained from their
49
manufacturers. ExtendedeUrface perfomance data are usually acquired by tests of the particular geometric design. See Subsection 3.12 of Chapter 3 for an example calculation which illustrates the considerations involved in designing boiler heating surfaces.
Section 3 internal Thern~odyna~nicsof the Steam Turbine 3.1 Nozzle Flow. A nozzle is a short flow passage of converging or converging-diverging flow area whose function is to convert thermal or pressure energy into kinetic energy. It thus forms an essential feature of both steam turbines and gas turbines. As the fluid passes through a nozzle, no external work is performed, and no heat is transferred, so that the general energy equation reduces to
LOSSOFPRESSUREB~D~GRADAT~ON OF ENERGY AT INLET
Fig. 13 Temperature f ador
Table 1
Tube Bank Depth Correction Factor, FD
EXIT PRESSURE
upstream. This degrylation is evident as a loss of stagnation pressure and, hence, of the pressure difference available to cause flow through the nozzle. ~t is indicated on the enthalpy-entropy plot of the nozzle process shown by Fig. 15. There is degradation of energy within the nozzle itself, so that the exit velocity is not as high as ideally ENTROPY, s possible. The total degree of degradation is expressed Fig. 15 The nozzle flow procer by the nozzle efficiency, which is thus the ratio of the energy actually converted to kinetic energy to that theoretically possible. In equation form, the definition of the nozzle efficiency q N is such as,the angle through which the fluid is turned, nozzle dimensions, and the ratio of approach kinetic q N = C?/(&~J) (50) energy to the total kinetic energy developed. Empirical h00 - hl' curves, such as Fig. 16, give nozzle efficienciesfor bladeThe meaning of hl' and the derivation of equation (50) type turbine nozzles for dry or superheated steam. T~ are evident in Fig. 15. determine the efficiency of a nozzle (either fixed or An alternative designation of the degree of energy moving), the basic nozzle efficiency h-2 and height tor~ ~ ~ ~ e r is s i given o n by the velocity coefficient kN, rection factor f~ are read from Fig. 16 and the nozzle which is the ratio of nozzle exit velocity to that ideally efficiencyis computed as
A, calculate L from equation (47) and multiply L by
ST= tramverse pitch, inches
sL= longitudiial pitch, inches
pressure of CO2 and HIO- Enter pR,the total ~ i 14~at the . average gas bulk t e m p e r b e , and at the proper p R value, ~ read € Q on the left scale. Then reenter at the gas bulk temperatme and read h?/rQ On the right scale a t the appropriate tube surface temperatme. hr is equal to the product E Q X h r / r ~ iincluded in the result is a tube surface emissivity of 0.g5. the 2.4 Heat Recovery Equipment That portion hest transfer equipment that absorbs heat at ternperatures below the saturation temperature of the generated steam is considered to be heat recovery equipmenta the Generally, such equipment absorbs the hest
d = tube diameter, inches l-he e ~ s s i v i t y of flue gas is a function of its temperature, the mem radiating length L, and the pressure Pa of its rdiating constituents (pfiwatervapor and carbon dioxide). The flue gas fuel oil in 15 percent produced when burning p ~ t i a pressure l of 0.114 excess air has a water vapor atm/atmand a carbon dioxide partial pressure of 0.125 combustion air (gas air heater) or into the incoming atm/&tm, a total of 0.239 atm/atm. Values of r~ for feedwater (economiser). Combustion air heaters are generally of either the this mixture are plotted in Fig. 14 for a range of gks bulk temperatures and a range of P R Lvalues, where PRis in rotary regenerative type or are tubular, with air atm/atm' and L is in feet. The curves are usable from through tubes heated by combustion gM p a s a d mound the outside of the tubes in Cross flow- The mtar3' re10 to 20 pementexcess air without appreciable error. and performance data is generative type is Plotted on the same figure is the value manufacturer. Tubdm best from the proposed air heaters can be readily evaluated by the methods reviewed in the foregoing for tube banbe ~ ~ t ~ ~ d ~e~onomizers d - ~ ~ are ~ femployed a ~ e of equation (20). TO determbe to the exclusion of bare-tube units- The feedwater which is another
(51) '
It may be seen that k~ is simply the square root of qN. The nozzle velocity coefficient is a function of factors
kN2 = fLk2 (52) For wet steam, a correction is necessary to,account for the impingement of the slower moving droplets of water on the back of the blades. l-his correctionis taken by some authorities to be
1 111
MARINE ENGINEERING
"
'CHERMODYNAMICS AND HEAT ENGINEERNG NOZZLE PARTITIONS
-
NOZZLE HEIGHT
' *' ''
8 RADIAL CLEARANCE, IN. D-CLEARANCE DIAMETER, IN. C O N T R A C T lFACTOR ~~ DUE TO THROTTLING OVERALL LABYRINTH PRESSURE RATIO FACTOR
a
NOZZLE OPENING
Fig. 20
N o d e nomenclature
Fig. 19 Blade and nozzle partilion nomenclature far a typical converging
'k
nozzle, section taken at mean diameter
C1, = tangential component of steam velocity leaving and blades is a relative one, nozdes being considered as fked and blades (or buckets) as moving. Fok nozzles or blades below the critical pressure.ratio, the area at exit is of primary importance. From Figs. 19
in the plane of the turbine wheel. Let m =
nozzles d = diameter T denotes the tip of nozzle M denotes the diameter of nozzle R denotes the root of nozzle
- -- -C=l r-T W 8
As
A = L d ~ a m Esin al
..
(64)
,where A is the area at exit from the nozzle, in square inches; L is the blade height in inches, and d~ is the mean wheel diameter. al is the angle of the steam to the plane of the wheel. Usually there is a small difference between the actual steam angle, all and the geometric angle a!. This angle arl - a; = 8 is known tis the deviation angle and is a function,of both the angle through which the steam is turned and the Mach number, and approaches zero as the Mach number approaches 1.0. In equation (64), al should be used when its value is known.
VlnT
A C C E L E R A ~DUE ~ ~ TO ~ GRAVITY- 32.2 F T , / s ~ ~ >
PRESSURE BEFORE LABYRINTH, LBS/FT2 PRESSURE AFTER LABYRINTH, ~~~.~~2 STRIP THICKNESS ATTIP, IN.
.
PITCH OF STRIPS, IN.
CIZM= C z v l n ~
VlnR
'
yln 7 specific volume.of steam leaving nozzles A, = axial flow area in plane a t wheel
W~ =< total weight of steam flow
't
"
CARRY-OVERCORRECTION FACTOR FOR STRAIGHTTHROUGH LABYRINTH; UNITY FOR STAGGERED
,
.
The result of these requirements is a warped blade, with generally pure impulse at the root and with a large degree of reaction at the tip. Normally, the laat few stages of the LP turbine are based on the free vortex condition, with the other stages having reasonable approximations of this flow. Obviously, as the turbine size incremes, the blade lengths increase and the free vortex design may be extended into higher-pressure
MARINE ENGINEERING
54
THERMODYNAMICS AND HEAT ENGINEERING
q-he rnmbined windage and friction losses may be estimated by Kerr's equation as follow:
A h
z
hoO- h*'
I,
--
(72)
1 f hllo pressures Po for each stage are known, a they
bo in the design process, then a stage-by-stage I ) I ~ ~ of ( J the conditions for each stage can be made on the Mllllitrr chart (h-8 plane; see Fig. 18). When completed, 11ll)t is known the state line, or condition line, for 111. bllrbine, shown by Fig. 22. It is I ' * I ~ I ~ in the andysia of extraction point conditions, "L'l"r requirement is knowledge of enthdpiea @A flitlotions of Pre8sures. However, the end of this f1111fl(r
d VI
w
hp = horsepower 10% d M = mean blade ring diameter, ind a = b k diameter to root of blades, inE = peripheral admission fnx?tion L = blade height, in. iy = blade speed, fps
55
h e , known the state h e end point (SLEP), represents the static enthalpy only, whereas the stagnation e n t h d ~of~the exhausting steam must be known for e such Uses 88 condenser design. Thus the C -- component 2gJ is u s u d ~ added, and the resulting stagnation enthalpy at exhaust is also plotted on the state line diagram at exhaust Presswe. This component is that which "presents the approach-velocity h p u t to a followkg stage, in equatioq (62), but which, in the last stage, must be wasted. ~t thus forms the major part of the t u b h e leaving or loss.
Sectio~~ 4 External Thenodynanrics of the Steam Tflrbine ''I
IWe Line for the *@' a hrbine unnll the
a
unit. The wheel horse- power output, thmttle pressure and temperature,and is given by equation (68). exhaust pressure are sufficientfor this task. of Wa, WL, hoop hsO,and hpj A preliminary step is to express steamflow in unitfom @@Qmowllt. The total Power delivered into the turbine a a steam rate (or waterrate) thusly: @h@fb i h must ~ be expressed zw a summation of the
compounded stages
who01 horsepowers by
ENTROPY
fig.22
where
propulsi~nturbine state line
SR
=
W,SHP
2544 UEtW= SRVM
UEw = heel Used energy (see Fig. 22), Btu/lb 9~ leakage efficiency
- hr4 - hpf
(68)
= total steam flow entering stage, lb/hr
wL = leakage
flow, i.e., flow that byk'asses the or moving blades, lb/hr hp = power absorbed by windage The wheel work per pound, based on total flow, is
enthalpies. stage efficiencyis then the ratio of ~h~ work delivered to the shaft to the available stage, or
, =
or, as a close approximation
per
(74)
(75)
MARINE ENGINEERING
59
be read from the intersections of the shifted pressure lines with the nonextraction state line. The approximation outlined here is not quite adequate for reduced-power conditions, or at unusually large extraction flows. Other techniques, such as discussed next, must be used. 4.4 Lambda Ratio. For large variations in flow such as occur when reducing to 80 percent power Or less, it is necessary to account for the change in efficiency because 'peed of the change in the ratio of blade speed to in SucCeS~ivestages. This can, of murse~be done by returning to the original design and applying the theory in Section 3 again. However, prodiscussed for a computer this is a tedious operation and normally carnot be accomplished in a timely manner, especially for preliminary work. The designer must (gq) therefore resort to other techniques based on external
that most desigoers and turbine builders prefer to use t6e s t r ~ g h tLine. since the state line is ~ m primarily d for heat balance work, an error of 6 ~ t inuestimating the e m in enthdpy at a given point resultsin a extraction flow. T~ arriveat a satisfmtory state line for fdl power, it is then only neoeasary to comect, on a Mollier diagram, the point of idtial PreBme and enthdpy, ho, with the enthdpy a t 90 per cent of throttle point of pressme, and the state line end point (see Fig. 22). ntraction ent.alpies csn then be read a t the appropriate shell or stage pressure. The steamleaves the tmbine at a total enh, = hi
+ EL + (RL)(SHP)(2544)
I
since the pressure a t the condenser is very low, the
Every turbine stage has a value of U/Cf for which its
preame ratio from the point of interest to the condenser efficiency is a maximum (this is shown by Fig- 18 of is typically supercritical, and maximum flow exists Chapter 5). For example, it is 0.5 for an ideal impulse
for the pressure a t that point. Under such conditions, stage. And although a propulsion turbine Consists of a flow theory predicts that the flow parameter number of stages for each of which the ratio may be W 6is and in f m t hm a value of approxi- different, as when impulse and reaction stages are used in the same machine,' there is always some vdue of mately 0.40 for superheated steam when To is in degrees d ~ Nfor whi& the efficiency of the entire turbine R, po in psi, W is in Ib/sec, and A is in squme inches( ~L ~ Z~ ~ 1 Further, for modest changes in conditions at a point in is a maximum (Z implies sumation Over all compared to the the the in is his parameter is known as Lambda, and is convenchange in Po, and A is fked, SO that the relation r pofouom. m e premure a t a point should thus tiO*lly expressed to the flow from that point to the be pediction is found to be essentially condenser, and true in practice; it is further theinpressure is wherein the constants, kcluding 0.5 for U/C', are pmportiond to the flow pastfound the that point question. included in the numerical mefficient- The efficiency at This additional distinction is necessary because some of function of the Lambda a point may be extracted downstream, ofi-design points is a ateam ratio X/Xo, i.e. and therefore does not reach the condenser. The principle stated in the foregoing is used to find shell pessures a t extraction points, and from them the extraction enthalpies following small changes in flowThe state line does not shift significantly because of Fig. 26 is a plot of the relative efficiency ofofa impdse rsaeonable ortraction flows, 80 that the enthalpies can s g h t h/lo- This plot is
THERMODYNAMICS AND HEAT ENGINEERING
61
MARINE ENGINEERING
ICS AND HEAT ENGINEERING
BOILER 88% EFF
000 SHP ABS MAXIMUM
Eb ~ 0 . 8 5 2 7(SEE FIG.23)
a = FLOW, L W H R h = ENTHALPY, BTUILB. Fig. 34
Simple steam cycle
EXHAUST ANNULUS AREA = 25 FTP
f r ' I.O125(SEE FIG.24)
A € = 1481.2-907.5 = 573.7
R L = 0 . 3 5 ~1 . 5 ~ 0 . 5 LEAKAGEaMECHANICAL LOSSES = 3.5% ITERATION, STEAM FLOW = W PbxA
I
1163 .56 ~ 0~02 5=.4360 ~
BYadn per hour, divided
1
-
(I.o¶)(-)
= 163,600
Le/HR
**. E L 5 10.8 (SEE FIG.25)
SECOND ITERATION, STEAM FLOW = STEAM RATE =
E x s ~,Xf,xAEn0.8527x 1.0l.25 x 573.7-495.3 h,- .,E, 1481.2 495,3 = 85.9
hi
(I .04) (245i,:
!:60p)=
[63,800 LBIHR RL X 2 5 4 4 hc. hi + EZ+ STEAM RATE = 999'0 Ag. 35 S t a h one and steam rate for cycle cafculationr
~
~ =, 5.46 ' LB/SHP-HR ~ ~ ~
the m~chanicdequivalent
t# b a d and the efficiency. Thus
divided by the net or 15,180~30,000= O e 5 0 6 ' lb/sh~-hr. The heat rate is the quantity of *heat t o produce one horsepower per hour and 144APvfQ~ t u / h r ~ ~ ~ u l aby t edividing d the net heat added to the plant, Per hour, by the horsepower produced.
MARINE ENGINEERING
'I'llo quantity of fuel required is determined by dividing l d I ~fr~el ~ j heat output by the boiler efficiency of 0.88 and
IIIIII higher heating vdue of standard fuel of 18,500 llbl~/lb,to which is added 46 Btu/lb to account for the ti~rlnibloheat added by the fuel oil heaters (100 deg F rim tdt 0.46 specific heat) :
Ipuel required = 2509846,193 = 15,370 ,b/hr (0.88)(18,546) I)ividing by the 30,000 shp output, the specific fuel ~~t~ll~nrnption is found to be 0.512 lb/shp-hr. Sa2 The Regenerative Cycle. The power cycle shown 111 Itig. 36 is complete, but certain problems would arise If oh a cycle were used. The feed temperature is unkr('melylow; a result, the economizer in the boiler wclllltl condense sulfur ~roductsfrom the flue gas, which wnrlld cause corrosion. Further, the feedwater would IIELVI! high oxygen content (no deaeration is provided), wal,trr wlriuh ~ides. would cause corrosion and pitting in the boiler
1
200 F Q = 1050
Q=
---~ g 36 .
FLOW. LB/HR
h = ENTHALPY, BTU/LB
Simple steam cycle with parasitic loads
ta&. ~ ~ ~ j + b l ~ - ~ and system leakage are bssed on reference [gl. soot,blowing requires 760 lb/hr, and system l e h g e losses are taken as 1/2 percent flow, or about 900 lb/hr. Both of of the mustbe replaced by makeup these items lost feedwater introduced to the condenser. ~h~ t h e r d enerw added to each pound of water by the feed pump is the same in the example, so that the thermal energy added to 177,920 lb/h. is 1,060,384 Btu/hr.
Uowance must be made for Pressnue and temperature drops in the main steam line. It is customm to allow up to the nearest about 2.5 percent on Pressure, 5 psi, and 5 deg I? for temperature. Thus, the heater outkt conditions are taken to be 875 psig and 955 F. NO dlowance for loss is made in the deSWerheated system, so the desuperheater outlet enthalpy is is 1250 Btu/lb- The total is 2300 lbhr' 175,620 lb/hr and the demperheated The flow of heat in the system illustrated by fig. 36 may be tabulated follows:
1l)rltrainedoxygen and air can be released by bringing tlrn foodwater to a boil. By using steam, bled from the t\~rl~iaas, the feedwater can be raised to the boiling I@lll~lorature and held there in a deaerating feedwater I r e ~ b in r an efficient manner as the bled steam has al~entlydone useful work in the high-pressure turbine befo~~o being used for feed heating. Illtrod feed heating may also be done after the dercrr~~l~ing feed heater and feed pump. For highest rflinioncy, there should be N-1 feed heaters, where N (r lllro number of turbine stages, since this leads to lncxirnum regeneration, but such an array of heaters &r&dbleed points is not justified in marine service. flbltbionary practice employs an extensive number of Iie&llrrr8,but such plants are not restricted by the space llmill~~tions of a ship's engine spaces, and they develop
generally used in cycles where the feedwater is heated to a maximum of about 285 F, so that economizers may be used for heat recovery in the boilem. Where feed temperatures are higher, gas air heaters ape used. Consider a simple single-heater cycle, using a deaerating feed heater, and otherwise identical with the cycle of Fig. 36. Steam could be bled at the crossover pipe between the H P and LP turbines at 60 psis and 1243 Btu/lb, and led through a pressure set at 46 psia, to a deaerating heater. The bled steam and the incoming feedwater could be sprayed together resulting in a saturation temperature of 276 F, 245 Btu/lb enthdpy. Bleeding steam from the crossover would reduce the horsepower developed since less passes through the unit. To compensate for this, the throttle flow must be increased. Let QI = the quantity of steam bled in lb/hr A&, be the increase in main throttle flow required. m e n , using the figures developed in Fig. 35, the reducis tion in heat available to the turbine lost heat = Qr(1243 - 996.7) and the flow needed to replace this heat is AQt(l481.2
AQt
.............a.
.................. ~ o t a l................................ .
175,620 X 1483.5 = 260,53%270Btu/hr 2,300 x 1260 2,875,000 263,407,270 B t u b 177,920 1 b b
r:
0.5084 QI
Then, leaving the exhaust of the LP turbine, the steam flow would be 163,550
+ AQt - QI
=
163,550 - 0.4916 QI.
Employing the procedure used in Section 5.1, the heat entering the gland exhaust condenser is
I b b
Btu/lb
Btu/hr
Leaving the main condenser 176,325 - 0.4916Ql 58.7 10,350,278 - 28.8691 Air ejector intercondenser. . 245 1250 - 93) 283,465 Air ejector after condenser. 245 1250 - 168 265,090 Gland exhaust condenser. ., 300 (1281 - 1681 333,900 Total.. .............. 176,325 - 0.4916Q1 11,232,733 - 28.8691
.
Fuel heating at 200 F . . 1,050 ~ b d dram a t 200 F................ 300 Air ejector after condenser drain a t 200 F....... 245 ~ ~leaving t dsurge t and entering feed pump8 177,920 Feed -p ........................... 177,920 .fotd to boiler.. ......................... 177,920
- 996.7)
Equating these gives
I
The boiler output is t b n .
65
THERMODYNAMICS AND HEAT ENGINEERING
--
THERMODYNAMICS AND HEAT ENGINEERING
MARINE ENGINEERING Ib/hr Btu/lb Leaving gland exhaust ,denser. ............. 176,325 - 0*4916&1 168 After condenser drain ..... 245 ~ h condenser d drain .... 300 168 +&I 1243 Bleed flow.. .............. F.0.hater drain. ........ 1050 168 ~ ~ t r r .............. l.. 177,820 4- 0.50&1Q1
Since the DFT enthalpy is 245 Btu/lb, (177,920 + 0.5084~~)245 = 11,500,693 + 1214.14~1 QI = 29451 lb/hr The total flow to the boiler is
+ 6) = 48,416,143 B t u / b
to the boiler, and the boiler output is (192,893 - 2300) (1483.5) (2300) (1250) Total boiler output less feed input Net heat input to boiler
= 282,744,716 Btuihr
2,875,000 = 285,619,716 48,416,143 = 237,203,573 Btu/hr
After condenser.. ....... 245 G h d condenser drain. .. 300 Bleed flow. .....-...... NXo fie1 oil heater drain.. ... DFT outlet flow. ....... 177,560
+ lW1 11,500,693+ 1214.1+@1
1.04
+
=
41,:1!J0,787 Btu/hr.
The boiler output is ,
(192,537 - 230011483.5 = 282,216,590 ~ t u / h r (2300)(1250) = 2,875,000 Total output = 285,091,590 ~ t ~ / h r input = 48,326,787 Net heat from boiler 236,764,803 ~ t u / h ~
firbogenerator Throttle steam conditions.. .850 pig, 950 F Exhaust. .................
13,963
and the horsepower developed by the HP turbine
+
-1- (1481.2 - 1243)(149,322 29,451)/2544 = 16jog5 1.04 and the total is 30,058 shp~ hthe reduced ~ ~ steam , flow in the LP turbine slightly increases the turbine efficiency, and the calculation be repeated with a new ratio of AQr/Qz and a new nonbleed flow. Since the exhaust enthalpy has been changed, the equivaJent nonbleed water rate is :
=
Dividing the net boiler heat input by a fuel heating value of 18,546 Btu/lb and a boiler efficiency of 88 gi~a fuel oil requirements of 14,5341b/hr. Dividing by the 30,000 shp o ~ t ~ u ~ r e s uinl tas specific fuel rate of 0.4845 lb/shp-hr. Tbis is a saving of 5.4 percent over the simple cycle, and in addition the boiler is protected from corrosion. A further gain in efficiencyaccrues in this cycle. In the high-pressure turbine, more steam Passes than in the nonbleed condition, and less passes through the lowannulus is the same prerrnve tw~lI1e.Since the before, the volume flow is m-hced, and the h3aving velocity be less also. The apparent exhaust flow is l63,30 - 0.4916 x 2 ~ 5 = 1 149,322 lb/hr (apparentflow is t h t t l e flow less any bleed but including the gland leakoff steam). The exhaust annul- is
(177~600+ 0.5095Q~)245= 11,4799561
176,400
1 (1243 - 995.6)(149,322)/2544 -
Assuming the same pump efficiency, the feed P U P per pound of wateris unchanged (iVe.,6 Btu/lb), so that the generator load should have increased somewhat, but this can be balanced by the decrease in boiler forced-draft blower power requirementsThe total enthalpy of the feedwater is 192,893(245
11,232,733 - 28.86Q1 41,160 50,400
25 sq ft, SO 149,322/(1.5)(25) = 3982, and from Fig- 25 the exhaust loss = 9.7 B ~ and ~ hw/ = g85*g ~ ~+ 9.7 = 995.6 Btu/lb, vice the 996.7 Btu/lb &own by Fig. 35. Then, the horsepower developed by the LP turbine is
177,920 + (0.5084)(29,451) = 192,893 1b/hr
Fig. 37. Figure 35 outlines the calculations for the propulsion turbine. This calculation, and those sum1214.21Q~ marked below for units,use the metho& and data
'I'I"' DPT outlet e n t h a l ~ is ~245 Btu/lb; therefore
Btu/hr
Iso4
W R = 1481.2
2544 = 5.448 Ib/shp-hr
- 995.6
Then, the throttle flow is 5.448 x 30,000 AQt = 163,440 + AQt and A Q ~= (247.4/485.6)&1 = 0.5095Q1
+
and the exhaust flow is 163,440 - 250
+ AQ, - QI
=
163,190
0.4905Qr
where the 250 lb/hr is the gland leakoff steam. Leaving the condeqser: Main turbine exhaust. 163,190 - 0'4905Qz Turbogenerator exhaust. Makeup feed. ................. lS6O Air ejector drain. ............... ,175,965 245 - 0.4905~1lb/hr ~ ~ t ............... d.. ......a
....a*.
'I'll~~ preparation of a heat balance is usually the first ?*P in initiating the design of a steam propulsion plant. results of the prelimioary heat balance are the ki~llmentalinput to purchase inquiries, and also for lllfill Plant desifP tasks rui sizing of piping. The first !wJ balance may be done from the approximate data in ~ ~ f c m n c[gl, e but subsequently, data supplied by the @f)tllponent vendors is used to update the calculation. In the last example in 5.2 a direct solution of the heat balhaoe problem was presented. Obviously this problem Would be more difficult if several bleed points are needed nlwl rrlore heaters employed, especially if the bleed presU l l t ~vary with flow. The problem becomes even more @~l~ll~lOx if ships's service steam is added to the balance. Wlliln a direct solution of the heat balance is possible by b i @ i l la~ series of simultaneous eq~ati008,it i s generally Illor0 ~impleto use an indirect trial-and-error solution b,Y na~uminga condensate flow leaving the main con-
rated capacity.. ........ .........13,600 lb/hr 1135 Btu/lb Exhaust enthalpy.. consumption at 480 gpm and 1200 psig. . 12700 lb/hr Exhaust Bnthalpy at 480 gpm and 1200 psig.. . .1139 Btu/lb Main air ejector steam consumption 1st stage.. ............. .245 Ib/hr 2nd stage. ............. .245 lb/hr Intercondenser drain temperature.. .......... .I25 F After condenser drain temperature.. .......... .200 F . Steam supply at 150 psia. . .I250 Btu/Ib ~ i ~ t i plant l l i ~ Water production.. ....... .11,400 gpd steamconsump~onfrom low-pressure bleed at
Setting up in tabular form:
Total makeup feed.
+ O.M)gWr
-
...
.3330 lb/br
consumption) ...........lo00 l b / b
1
69
MARINE ENGINEERING
-------
TURBO-OENERATOR LOAD
L, , ----- -
The f i s t step is to estimate optimum bleed points. The feed temperature leadng the deaerating feed tank has been set at 280 F to prevent the condensation of sulfur products from the flue ges in the boiler economisers. To achieve this temperature, a pressure of 49 p$a must be available from the auxiliary exhauatlintermediate pressure bleed system. This pressure is controlled by a pressure regulator installed in the bleed systems. To provide heating steam when no bleed steam is available (as for example, when going astern), makeup steam is supplied from the desuperheated system through a pressure regulator set at 45 psia. At certain times, there may be too much auxiliary exhaust regulator, re set at 53 psia, will steam, so a b a ~ k - ~ r e s s ~ - dischmge excess steam to the main condenser. Thus the auxiliary exhaust system can fluctuate only between 53 psis and 45 psis. This limit should be sufficient to prevent the feed suction water from the DF'T from flashing during maneuvering, especially if the DFT is placed well above the pump (40 to 75 ft). Allowing a 7 percent pressure loss through the bleed/exhaust system, the bleed steam at the turbine must be at least 7 percent hi&er than the desired 49 psia, or 53 psia. operation to be It is further desirable for this able to continue bleeding to at least 65 percent flow in
the main turbine. Since the bleed pressure at any stage is dependent on the flow through that stage, it is desirable to select a bleed point at (53/0.65) = 81 psia. Note that since this is a direcbcontact heater, there is no terminal temperature The optimum bleed difference. point for the low-pressure stage is then selected so that the temperature rise in the condensate is evenly distributed between heaters. ~ e a v i n ~ the condenser a t 1.5 in. Hg abs pressure and 90.7 F, the condensate will be heated by the air ejector intercondenser and after condenser and the dand condenser to a temperature of about 100 F. The temperature rise to the DFT is 180 deg F, approximately half of which should be achieved in the LP feed heater. ~ h u s ,a proper condensate temperature leaving the LP feed heater is 190 F. Since a 10 deg F' terminal difference is usually needed between the heating steam and heated water in shell-and-tube-type heaters, the steam entering the heater must have a saturation temperature of 200 F, corresponding to 11.5 psis at the heater shell. With a 10 percent pressure drop in the piping, the turbine bleed point pressure must be 12.8 ~ s i a . For the preliminary heat balance then, the LP bleed point pressure is 12.8 psia at an enthalpy of 1138 ~ t u / l b and the I P bleed point pressure is 81 psia at an enthalpy
lbb
Btu/lb
Btub
154,865
58.7
9,090,576
283,465
265,090
154,865 lb/hr
Note that in this calculation, the weight of the drain is not added to the total, since the intercondenser &&n goes to the condenser and the after condenser drain to the freshwater drain collecting tank shown on the
THERMODYNAMICS AND HEAT ENGINEERING
MARINE ENGINEERING
Entering the system lb/hr Chde-te from LP htr.. .......... 154,865 Drains from drain tank. ............ 30,175 Exhaust from feed pump. .......... 12,700 Feed pump recirculation............ ~lOO0 Bleed steam from turbine. .......... Q Total. ...................... 217,740 Qa system .................. (217,740 + Qe)
+
(217,740 4- Qc)249 = 49,471,470 Q, = 4,650 lb/hr 14340 X 16.5 X 0.2445(278 1270 - 277 = 10,370 lb/hr
+ 1270Qc
- 100)
Btu/lb
Btu/hr
158 24,468,670 180 5,431,150 14,465,300 255.3 5,106,000
1139
1,270 Qa 49,471,470 + 1,270Qe (217,740 + Qc)249
1270 249
'"-
-
(ho h ~ = ) 17,030(1481.2 1138) 2544 2544 = 2,297 whp and the wheel horsepower of the IP bleed flow is:
= 1,247 whp Total = 31,242 whp
EnteFing the subsydm
lb/hr Condensste from mitt air ejedor after condenser.. .................. 154,865 300 G h d steam from turbines. .......... 245 M . e air ejector after condenser drain. . fistder mr ejector dram. ............ Makeup feed, taken at 75 F . ......... 3,330 Air heater drain.. ................... 10,370 Miec. heating drains.. 1,100 Low-pressure feed hater.
Btu/lb
1281 168
9,639,130 384,300 41,160
............... ............
+
h v i the ~ subs &em From gain t a d at 212 F............ (15,595 QL) 180 2,807,1°0 From LP heater at 190 F . 154,865 158 24,468,670 Total. ..............................................27,275,770
...........
+ I80 QL + 180QL
Equating incoming to outgoing total enthalpies gives the air heater drain in the air heater coils. It should also be noted that, in many cases, the low-pressure feed heater is drained, via a drain cooler, to the main condenser. Also, sometimes, the entire drain tank is also drained via a drain cooler to the main condenser. The steam supply to the steam air heater is bled from the 81 psia stage at 1270 Btu/lb. Allowing 7 percent pressure drop in the piping, the pressure at the air heater is 75 psia when rounded off. At 75 psia the saturation temperature of the steam is 308 F and the condensed drain enthalpy is 277 Btu/lb. There must be a temperature difference between heating steam and heated air leaving the air heater; this terminal difference should be between 25 and 35 deg below the steam temperature. Choose, for example, 30 deg F aa a terminal difference,so
+ 1138Q~ = 27,275,770 + 180&~ QL = 14,580 lb/hr Then, the drain tank flow is 15,595 + 14,580 = 30,175 lb/hr at 212 Fj and the condensate flow is 154,865 13,307,070
lb/hr. The next part of the system is the deaerating feed heater (or DFT). This unit receives condensate from the LP heater and drains from the drain tank via the drain transfer pump. I t also receives heating steam from the feed pump turbine exhaust and bleed steam he bleeder from the intermediate pressure bleed. steam is controlled by a pressure regulator set to maintain 280 F. A weight flow and heat flow balance for the DF'T gives
-1139)(12,7m) = 6.3 Btu/lb, 222,390
-
W
145 045
-=-L=
P6.A
3868
1.5.25
= 27,698 whp
is made for liquid compression. The fuel rate is now calculated from the heating value
18,546 \
n@ wll001 horsepower of the LP bleed flow is:
+ L16.5 X 0.2445 X (278-loo)] =
19,264 Btu/lb of fuel
The heat added to the superheated steam is
MARINE ENGINEERING
185,520(1483.5
'THERMODYNAMICS AND HEAT ENGINEERING
- 255.3) = 227,855,660 Btu/hr
1 Ill1 rl~l1t111,ted steam,
II I ' r
The enthalpy of the desuperheated steam is 1250 Btu/lb, and the heat added to it is
and a low-pressure turbine. The
1 I' unit8 are mounted on the same shaft in the
tlg, with their high-temperature ends back to 'I'h~sthere are only two input pinions, as with Ill0 crc,l~vc!ntional nonreheat turbine. I1rflflit:Cionsof a reheat turbine state line during E b = 0.865 lflulll1lillllJ'.Y design studies are more difficult than for llod the temperature correction may be computed aa l l ~ ~ ~ l t ~ ~ Iturbines, 1 ~ ~ 1 1 ~ 0 since there is a wider range of 17l1tl 1 nt\(,crri~tics that affect the line. The marine engineer T 4100 ft = r l l l l ( tolll! burbine designer must cooperate more closely to ~ l l ~ l l l l ~ c ! e11r design acceptable to both than is necessary 5000 (96) f t = 1.01 11 nil 11l)lor propulsion plants. Nonetheless, as a first R a l l ~ ~ ~ ~ of l ( la ( ' suitable state line, the high-pressure turbine The state line energy for the low-pressure turbine is ~ ; ~ l ~ r r1)rossure ~ ~ a b can be selected at about 20 percent llr I 141~rottle pressure, and a state line can be conE E L= EbftAE (97) l l ~ c l t ~ f l twith l an efficiency of 70 percent (excluding E B L= 449 Btu/lb ll~fll.11lg lblld gear losses). A 10 percent loss of pressure With an astern turbine loss of 0.5 percent [see equation i l l ( 110 rfrI~oater may be assumed. The balance of the (78)1, a first estimate of the steam flow is determined to be 138,400 Ib/hr from equation (83); therefore c l l ~ l l lc ~ !tl~i
Ijlllll{,
16,870(1250 - 255.3) = 16,780,590 Btu/hr For a boiler fitted with a steam air heater, a boiler efficiency of 88.5 percent can be expected; therefore, the fuel burned per hour is determined to be 227,8555660 16,780,590 = 14,349 ib/hr (19,264)(0.885) and the fuel rate is 14,349/30,000 = 0.478 Ib/shp-hr. The results of the foregoing calculations are entered on the heat balance diagram shown in Fig. 37. Since the shp check was close and the f i s t estimates of the steam consumption by the feed pump turbine and the air heater were well confirmed, these results can be considered final. If any of the checks had failed, the process would be repeated with revised estimates based on these results.
73
+
ENTROPY
RU. 38
&timote of state line for propulsion turbine with reheat
Section 6 Waste Heat from Diesel and Qas Turbine Engines I
from Diesel Waste Heat. A large fraction illput to an internal combustion engine is r @ a l e ls* ~l!ll*ibleand latent heat in the exhaust gases. *h+rrllpr l l u 1 ~fii~llificantfractions are lost via cooling of )fi@lvl, nlLtcr1 lube oil, and inlet air (turbocharged nlllle@ "'ill1 ~bftercoolers only). For example, the f ~ ~ ~ ~Ilt!atin~ut l l ~ l going l ~ ' into ~ ~the waste heat streams hf H k\~l'boal~larged two-stroke engine might be (1
#f tl!fl
Ilfldl
0.35 to exhaust 0.15 to jacket water 0.05 to lube oil 0.05 to aftercooler I I!r ~llnnt!loxlraust gas temperature is a t least 600 F
*&11111 111bf1,
it is feasible to extract part of its sensible ~lro(lllc~ usable steam. As the cooling water I E + f l l l ~ l ' i l l J l l ' ( i~ ! 1 ~than ~ s200 F, there are few uses for this *' I ' One use of practical importance, how-
ever, is the operation of a vacuum distillb for freshwater production. Steam can be produced in a heat exchanger (waste heat boiler) in the exhaust duct. ~h~ maximum steam pressure 9 b t ~ n a b l eis limited of course by the exhaust gas temperature, but othelqrise the premure is set by considerations regarding the use of the steamand the quantity needed. If steam is to be used solely for heating purposes, a relatively low pressure, say 15 psig, may be adequate, but usually the heat available is far in excess of lowpressure heating needs. Often, the ship service electrical needs at sea can be met by waste-heat steam applied to a turbine-generator. The higher the steam pressure, the lower will be the turbine steam consumption, but also the lower will be the quantity of steam that can be produced. Figure 39 illustrates alternative steam production at 50, 100, and 150 psig, showing that with
MARINE ENGINEERING
THERMODYNAMICS AND HEAT ENGINEERING
MINIMUM TEMPERATURE DIFFERENCE("PINCH POINT*).
o
b~ 4s
EXHAUST GAS FLOW 119.000 LBlHR
Z 4 6 8 1 0 1 BACK PRESSURE, INCHES H,O
2
IRect of back preoure and intercooler outlet tcwnperctture on exhaurt tmpbroturq Sulser RD-type engine
taken by the external cooling devices. The evaporator must not change the temperature of'return cooling water from its specified range under any condition of operation. 6.3 Use of Gas Turbine Waste Heat. Exhaust gas heat from gas turbines can produce steam in the same manner as for diesel engines, and for the same purposes. Since g&8turbines are generally less efficient than diesels, the heat available tends to be greater than with diesels. In fact, there is sufficient energy available to suggest use of the steam in a propulsion steam turbine geared to the propulsion shaft in parallel with the gas turbine. Perhaps 20 to 35 percent of the total power can be produced by the steam turbine, with a consequent major improvement in the fuel rate obtained with the total system. The design objective in a combined gas turbine and
0.25 0.50 0 75 FRACTIONAL LOAD
pa. 40 Exhaust ROW and temperature, Sulzer RD engine
the same inlet temperatures more steam is produced at progressively lower pressures. The minimum temperature difference, or "pinch point," as indicated, is the governing consideration in the steam quantity that can be produced. However, additional &earn is sometimes obtained at a lower pressure in a second boiler downstream of the fist. The minimum temperature to which the exhaust gas 2 cooled is also a limitation, since the temperature should not be allowed to drop below the dew point in order to avoid corrosion in the cold end of the boiler. Wade heat steam systems are designed in a variety of forms, but generally contain the components expected in a self-contained system. The designer, in making a heat balance, will apply the same techniques outlined earlier in this chapter. He must allow for the fact that
0
ZOO
400
wo
800
EXHAUSTGASTEMPERATURE,.F
1000
76
THERMODYNAMICS AND HEAT ENGINEERING
MARINE ENGINEERING
1 D. Q. Kern, Process Heat Transfer, McGraw-Hill Iloolc Co., Inc., New York, 1950. fi W. H. McAdams, Heat Transmission, McGrawllill Book Co., Inc., New York, 1942.
LEGEND
=
- ---------
AIRORGAS SUPERHEATED STEAM LOW PRESSURE STEAM FEED AND CONDENSATE DRAIN , GLAND LEAKOFF AND VENTW STOP V. NCHECK V. 4 : ORIFICE ~
--
--
&coNTRoLv. P-OR-T
A BACK PRESS.V.
PG=PSIG PA= PSIA P = LBIHR FLOW h = BTUILB F = TEMP., DEG. FAHRENHEIT W = GAS FLOW, LBlHR
STEAM AND FEED CONDITIONS SUPERHEATER OUTLET 2 8 5 PSlG MAlN TURBINE THROTTLE 2 8 0 PSlG MAlN CONDENSER VACUUM AT 108.7.F FEED WATER TEMP. TO BOILER CALCULATED FUEL RATE
617.F 612.F I,NJ. 27.5 HG 260.F
(I "Standards for Steam Surface Condensers," Heat Il)xcitiungeInstitute. 'I "Standards of the Tubular Exchanger Manufarilurcrs Association," Tubular Manufacturers Associa-
,399 LBISHP HR
BASED ON
lioll.
MAlN TURBINE NON-EXTR ST. RATE 0.BLBISHP HR BOILER EFFICIENCY 46% HHV OF STANDARD FUEL OIL 19,650 BTUILB
H A. Egli, "The Leakage of Steam through Labyrinth
PIRJLIH," Tram. ASME, i935. FUEL F W , = 10.700
LOST 9-47
74I0F
(1 "Recommended Practices for Preparing Marine Htnrcm Power Plant Heat Balances," Technical and #e~tinrchPublication No. 3-11, SNAME. 10 A. Norris, "Developments in Waste Heat Systems klr Motor Tankers," Trans. Institute of Marine En&lrlnatwj, 1964. I I R. M. Marwmd and C. A. Bassilab, T h e l'lirrtnodynarnic Design of a Combined Steam and Gas
T'lrtdna fl7-[IT-16, 1967.
VACUUM PUMP
(5-
MAlN CONDENSER 2 . 5 " ~ABS. ~
MAlN FEED DEAERATING
Q =5 8 9 7 6
MAIN
CONDENSING PUMP
Fig. 45
Design-point,heat-balancediagram for a combined gas turbine and steam turbine cycle
steam turbine is fired by the exhaust gas. Observe also that the gas leaves the boiler a t 440 F, and thus still has considerable thermal energy available for the production of additional steam at a lower pressure. A second, low-pressure boiler is provided to make steam for the deaerating feed heater. The heat balance shown is for design power. It is also of interest to see how the important parameters change as the load is reduced. Figure 46, also from reference [ll], illustrates this. Actually, the effect on the system parameters is influenced by the manner in which the plant load is controlled. For the example given, the fuel flow to the gas turbine is controlled to
77
maintain a governed gas-generator rpm. The steam turbine is uncontrolled, with the output being determined solely by the energy available in the gas turbino exhaust. References
1 Thermodynamic and Transport Properties of Steam, ASME, 1967. 2 Joseph H. Keenan and Joseph Kaye, Gas Tables, John Wiley & Sons, Inc., New York, 1948. 3 Frank 0. Ellenwood and Charles 0. Mackey, Thermodynamic Charts, John Wiley & Sons, Inc., New York, 1944.
ASME
3001 20
-Fig. 46
I
40
00 6 0 \" 100 PERCENT OF PLANT RATING
120
hdanannof a ,..rhed gas turKne and steamNrKne cycle at fractional power
BOILERS AND COMBUSTION
C H A P T E R Ill
Sedion 1 Classifisation of Marine Steam Generators
tained during the record-breaking runs at about 30 ~ s i g , which was about the upper limit of pressure during the Civil War era.
1
Numbers in brackets designate References at end of chapter.
1 BOILERS AND COMBUSTlON
MARINE ENGINEERING
type of boiler. ~t is with this type that attention is focused primarily since it has been used most frequently since World War 11. while there have been many variations of the foregoing boiler types employed throughout the world, the typesdiscussed are fairly representative and provide an steam adequate backgmund for an understanding generator types and characteristics. 1.2 current lypes of Mer&."+ and Naval Boilers. the Past100 years steam pressures and temperatures have increased fmm 30 psig saturated to 870 psig-950 F in mostmerchant vessels, and 1200 psig-lOOO F maximum (950 F in mostpostWorld war 11 naval vessels. A trend is apparentin large, high-power installations where steamat 1500 psig-g50 F, and in some instances reheat to 950 F, appears desirable. ~h~~~ installations will be used in increasing numbers where economically feasible. For the mostpart, widespread use of water cooling in the furnaces is employed to reduce refractory mainheaters singly, or in tenance. ~~~~~~i~~~~and are used to obtain the desired overall steamgenerator efficiency. Attemperators are employed in most new construction to control the steam temper* operating range and thereby improve ture over a turbine performance. Desuperheaters are installed to provide low-temperature steam for audiary purposes throughout the ship. ~~~k~~ c residual oil is the most widely used fuel,
although in some instances diesel or other light fuel ofis are used. Steam-atomizing oil burners, first used aboard ships in the late 18001s,have returned to favor with the advent of high-capacity low-cost evaporators to supply the necessary water' This type of atomizer' while providing an extremely wide range of operation, results in a 1088 Of distfiled water which was, until recently, too big a penalty to pay for its advantages' However, improvements to reduce the consumption Of steam, coupled with abundant distil1ed water, have led to its widespread use, particularly in automated boilers. Two-drum integral-fumac0 a Two-Drum boilers, or D - ~ Y boilers P~ ss they are Often Of many and called1 are made steam drum and water drum connected by water and boiler bank tubes. Superheaters are instal1ed between the water screen and the boiler bank and may have tubes arranged either vertically Or horizontallyr depending in part on which arrangement best fits machinery arrangement- Where required, the temperature may be controlled by means of a control desuperheater or attemperator located i n either the location Of the Oil or water drum. The hing front burners is frequently dependent On the machinery (mart 'Onarrangement and may be in the Figures and indicate ventional), roof, or sidewall. some of these variations. In most i~tallationssome form of air heater is used with an economizer. The type and pmportions Of these auxiliw heat exchangers depend On the 'yd0 arrangement. If two stages of feed heating are selected, a steam air heater and an economizer are Often used'
'
81
BOILERS AND COMBUSTION
MARINE ENGINEERING
83
fig. 7 Two-drum, top-fired boiler with verfical superheater and ecanamizer
Ilg. 6 Tw-drum,
single-furnace bolkr with horizontal superheater
*I~lnlr would be incompatible with natural circulation. i'tco greatest disadvantage is the circulating pump Ib~lf,which is a potential source of trouble and mainCFII~IIOO. Fig. 5
Twdurnace, single-vptoke cantrolled superheat boiler
other furnace supplies heat to the superheater. Some designs incorporate a part of the superheater (called a primary superheater) in the reheater zone to provide additional protection for the reheater and to obtain the desired steam temperature characteristics [3]. The gas flowing from both the reheater and superheater combine in the main generating tube bank, and a single gas flow path is maintained through the auxiliary heat exchangers as in the single-furnace design. c. Forced-Circulation Boilers. Ever since the first boiler was used aboard ship, marine boiler designers have investigated and experimented with various means
to reduce the size and weight of boilers. A boiler arranged for natural circulation of the water and steam requires low waterside pressure drops which can only be obtained by installing sufficient downcomers and risers. This adversely affects size and weight. By supplying a pump to either augment or supplant natural circulation, a smaller and lighter boiler can be designed for a givengteam output (41. The circulation in such a boiler is said to be controlled or forced. The chief advantages of this are that very small-diameter tube^ with a high resistance to flow can be used in arrangements of heating surfaces and steam drum location^
'rlln LaMont boiler, shown schematically in Fig. 9, la rr typical example of the forced-circulation type. Wlllln wed abroad, it haa not found wide application in Ill@~rrrtrinefield in the U. S. The LaMont boiler uses a sltrgle clrurn into which the heating surface discharges a
wlato~ilatr of steam and water. The circulating pump e~teklnr~ ia supplied by gravity from this drum and forces refiller lllrough the generating tube surface, which is ~ I ~ ~ I ~ ! I Jof U Ia) number ~ of tube circuits arranged between r! hlslllbibutingheader and the steam drum. The inlet HI P~I!II tube is fitted with an orifice to balance the flow ~ralrrhnnoowithin the various circuits. This is necessary III 11l8bdnun adequate flow of water in each tube dependllrl ik oxpected heat input. The furnace, oil burners, sl~l~arl~nr~tor, and economizer are similar to those of $ 4 I ~111,ttl-airculation boilers.
Fig. 8
Slngle-furnoce, gas-bypass reheat boiler
BOILERS AND COMBUSTION
MARINE ENGINEERING
WATER INLET4
CONVECTION BANK RADIANT HEAT ABSORBINGSECT\ON
~ g 9.
Schematic of LaMant forced-circulafion boiler with economizer and superheater
of 150 to 300 psig. Boilers of this type are usually built only in small sizes and supply up to 7500 lb of saturated steam per hour. Because of the difficulties in maintaining feedwater chemistry, adequate water flow through parallel tube circuits which would be required for higher capacity boilers of this type, and the control of superheated steam temperatures, the once-through boiler is not well-suited for marine propulsion purposes. e. Supercharged Boilers. The superchased boiler has the characteristic of using combustion pressures higher than one atmosphere in the furnace to take advantage of higher gas densities and higher gas velocities than are available in the usual marine boiler. Figure 11 is a typical supercharged boiler. This unit is an outgrowth of the Velox boiler which has been used in a few stationary power plants for a number of years. In
I (11orlomisers of either the bare-tube or extended-surface t ~ v l ~ [we c r used to increase the temperature of the incoming
Iwlwuter by cooling the flue gases leaving the boiler. Ail- I~oatersare used to increase the temperature of the a~)~t~l)ustion air so as to promote better combustion of Iilln fuel. In the case of gas-to-air heat exchangers, air I ~ ~ n ~also c l r improve ~ the boiler efficiency by reducing 1,110 tomperature of the flue gases. By using low~ l r n ~ ~ u low-temperature re, exhaust or turbine bleed 3bnr1,rn to heat combustion air, as in the case of the steam dlu I~oeter, the overall cycle efficiency is improved. 'I'llcmo various types of heat exchangers may be used rrl~~yly or in combination with each other. a. Economizers. An economizer is a simple heat u~c\l~nt~ger consisting of a bank of tubes connecting an I ~ ~ l nr~nd b outlet header located in a relatively cool gas Iel111 mrature zone beyond the boiler main generating I~alllt. Supplied with water at a temperature near that Iuavil~gthe last feedwater heater, the economizer supplies ~rlrlihionalheat to the feedwater by cooling the flue gas. Irr lrlnrly installations the economizer is the final heat cttallttnger in the exhaust gas path. I t may, however, ko followed by an air heater where a higher efficiency is
typo# me forced circulated by the main feed pump. In $~l~nt*rkl, they are designed to heat the incoming feedwater CII willliinabout 35 deg of saturation temperature. They rre r~~rrangedfor counterflow of the water and the
The work of compression shows up, in good measure, as an increased temperature of the combustion air. As
A- FAN B-OIL BURNER WITH IGNITOR AND FLAME SCANNER C- FURNACE D-GENERATING COILS E- STEAM SEPARATOR F-STEAM TEMPERATURE LIMIT CONTROL
Fig. 10
Once-through boiler
d. Once-Through Boilers. The boiler in Fig. 10 is an example of once-through boilers used for auxiliary steam. Water is passed through the heating surface in one continuous circuit by the feed pump. The boiler is basically one long spiral tube arrangement composed of a economizer and a transition zone, where evaporation is completed, which surrounds the furnace. The feed pump pressure determines the outlet steam pressure, which may be 1200 to 1800 psig, dthough for the usual marine installation the pressure is in the range
tive naval vessels. The original Velox boiler, from which supercharged boilers evolved, was a forced-circulation boiler. However, subsequent supercharged units have employed circulation to avoid the extra complication of the circulation pump. f. .Waste-Heat Boilers. In vessels powered by diesel or gas turbine engines, the exhaust gases contain considerable available heat. Boilers placed in the stack to reclaim this otherwise wasted heat are called wasteheat boilers. Usually they generate low-pressure saturated steam which can be used for purposes such as tank heating, galley, and space heatingIf desired, they may be designed to bum oil when the main unit is shut down. Basically, waste-heat boilert3 consist of a bank of generating tubes that are either bare Or or of the extended-surface ~ Y P - Either forced circulation may be used. 1.3 Auxiliar), Heat Exchangen. In addition to the steam generator, several forms of a d i a ~ heat exchangers are inwrporated in boilers to impr0ve the efficiency and the overall operation of the plant
' h o nimplest economizer arrangement is the bare-tube !,up0 ~ ~ this n d was the form the first economizers took. 8flwuvcr, it was recognized that the use of extended ~ ~ l ~ ' f #tou oincrease the total heat-transfer surface for a PII lorlgth of tube would provide significant increases performance without penalizing weight and space rullwidorutions adversely. Figure 12(a)shows anefficient b41tn of extended surface in which flat studs are spaced rb dlimclogangles around the circumference and at %-in. IiikerfvaInalong the tube. 1h(tulldedsurface can also take the form of spiral fins #@ldpd on the hlbes or of cast iron Or alu~linumgill rings ~~~~1111~d 01 shrunk onto steel tubes as shown by Fig. 12(b). ba Air Heaters. The cooling of hot flue gases by the iilPo1rlillb! combustion air is one of the oldest of concepts Iily)r()ve boiler efficiency. In addition, heated air ~ I F U V ~ ~ ~an O B additional beneficial effect by promoting @@illd lblld complete combustion of the fuel. This can irn~ortancein the relatively small furnaces used III lr\~tl*ino boilers. Alr htraters fall into two broad classifications, the ke~~ll~~~!l'r~tive and the regenerative. In the recuperative bvlle, II(!IL~ from the products of combustion passes
1:
through a partition which separates the products from the air. Tubular and plate-type air heaters are examples of recuperative air heaters. In the tubular heater (Fig. 13) the walls of the tubes transfer the heat from the gas to the air. The plate-type heater is not c o m m o ~ yused in the U. S. in marine service. In it the air and gas are separated by plates through which the heat flows. In the regenerative air heater, heat is first stored in the structure of the heater itself .as it passes through the hot gas stream. The heat is then givenup to the air as the structure turns through the airstream. The air preheat& shown in Fig. 14 is an example of this type [7]. I t consists of closely spaced heating elements packed into a revolving frame. The frame speed is constant and is controlled by a small electric motor. The frame speed is selected such that the elements will absorb heat from the gas with a good temperature merential and, at the same time, 'the elements will heat the incoming combustion air to the highest possible extent. The upper section of the air heater is in the cold-air zone and also "sees" the coolest gas. It is usually arranged SO that the heat-transfer surface can be conveniently removed in easily handled s e ~ t i o ~ s - ~ a l l ~"baskets'-since d corrosion and fouling may occur there. These baskets
MARINE .ENGINEERING
is removed by steam traps. The latent heat of this steam which would otherwise be rejected in a condenser is returned to the boiler via the hot air. 1.4 Boiler Terms and Definitions. The. location of some of the more important boiler elements are shown in Fig. 5. For an understanding of marine boiler technology, a review of the applicable terms and definitions of various essential boiler parts may be helpful. The following terms and definitions are based on the standmds of the American Boiler Manufacturers Association
of a superheated vapor. Boiler hand.. . . . . . . . . . . . . . Boiler arrangement is described by reference to the location of the uptake gas outlet with respect to the designated front of the
Fig. 13 Tubular air heahn
Heating surface. . . . . . . . . . .
close all or e portion of a steam generator unit.
may, in addition, be provided with a ceramic coating similar' to porcelain enamel for protection against the
for treating the boiler water are introduced. Circulation ratio. . . . . . . . . . . The ratio of water entering a
are used as supply hbes to supply water to a drum or header. fercrad circulation. . . . . . . . . Circulation in a boiler by mechanical means external to the- boiler. Pllrnaoe screen. . . . . . . . . . . . One or more rows of tubes arranged across the furnace gas outlet. Pursl~cevolume. . . . . . . . . . . The volume contents of
may flow from the steam drum to the water drum or header. That surface which is exposed 'to the heating medium for absorption and transfer of heat to the heated medium, including any fins, gills, studs, etc. attached to the outside of the tube for the purpose of increasing 'the heating surface per unit length of tube, '
'
steam, usually expressed as the percentage by weight.
well system through which fluid flows downward. or box inside the steam
atural circulation. . . . . . . .
watertube boiler convection bank which is normally provided with a blowoff valve for periodic removal of sediment collecting in the bottom of the drum. Circulation of water in a boiler caused by the difference in density between. the water in the down. comers and the watersteam mixture in the gen-
BOILERS AND COMBUSTION
MARINE ENGINEERING
AIR IN
GAS OUT
1
I
t AIR OUT
89
I ROTOR
SEALS
I G A S IN
(a) Assembly
(a) Assembly of typical section
(c) Crimped spiral fln Fig. 15
(b) Replaceablebaskets Fig. 14
Rotary regenerative air heater with replaceable cold-end baskets
. . . . . . . . . . . The plates, centrifugal sepaRadiant heat absorbing.. . . . The projected area of tubes Steam baffling.. rators, or baffles arranged surand extended metallic surface (RHAS) to remove entrained watcr faces as viewed from the from the steam. furnace. Included are the --walls, floor, roof, and partition walls in the plane of the furnace exit screen. Steam or steam-and-. . . . . . . A pressure chamber located at, the upper extremity of II water drum Heat-transfer apparatus for Reheater. . . . . boiler circulatory system i t 1 heating steam after it has which the steam generateti given up some of its original in the boiler is separated heat in doing work. from the water and fro111 A tube through which steam Riser. . . . . . . which steam is discharged and water passes from an a t a position above a watcr upper waterwall header to level maintained therein. the steam drum. ~
Steam air heater
R1111~rlv)ater.. . . . . . . . . . . . . . A group of tubes which absorbs heat from the products of combustion to raise the temperature of the vapor passing through the tubes above the saturation temperature corresponding to its pressure. 'I'rr11yr311l;-tube wall. . . . . . . . . A waterwall. , in which the tubes are substantially tangent to each other with practically no space between the tubes. I'llllr I I I I I ~ ~.. .. . . . . . . . . . . . . A group of two or more rows of tubes forming part of a watertube boiler circulatory system and to which heat is transferred from the products of combustion mainly by convection.
Tube sheet. . . . . . . . . . . . . . . The part of the drum or header which the ends of the tubes penetrate. Unheated downcomer. . . . . . A tube not exposed to the products of combustion in which water may flow from the steam drum to the water drum or header. Watertube.. . . . . . . . . . . . . . . A tube in a boiler having the water and steam on the inside and the products of combustion on the outside. Water-cooled furnace. . . . . . . A furnace wall containing watertubes arranged to form a waterwall. Welded, mono-wall, or. . . . . A waterwall in which the membrane wall tubes are welded together (or to filler bars between them) to form a continuous furnace wall.
90
BOILERS AND COMBUSTION
MARINE ENGINEERING
Section 2 Consideratiofls in the Selection of a Boiler 2.1 General. Many factors influence the design and selection of steam generating equipment to produce the required quantities of steam at the design pressure and temperatures for a particular installation. Efficient operation when burning the various fuels available throughout the world is a requirement. The boiler also must fit easily and conveniently within a minimum of engine room space, yet be accessible for operation, inspection, and maintenance. Although light in weight, it has to be sufficiently rugged to operate dependably under adverse sea conditions. Operation over a wide load range, with a minimum of attention, and operating characteristics compatible with a high degree of automation are also required. The factors used in both the thermal and structural design must be conservative to provide assurance that continuous operation over extended periods of time will be provided with minimum maintenance. Finally, the boiler must meet the rules and regulations of the regulatory bodies. 2.2 Cycle Requirements. The design of a marine boiler is directly affected by the heat cycle selected by the ship's designer. Over the years steam pressures and temperatures for marine power plants have advanced by a series of broad jumps. After each jump there has been a pause to consolidate the gaina, review the operating results, and plan the next jump. In general, marine steam conditions have not advanced as rapidly as those in use ashore. In part this has been due to the relatively small horsepowers involved and in part ta the demands of the ocean environment. As the safety of the vessel and its personnel is dependent upon a reliable power plant, each new advance is made only after adequate experience is accumulated with the last. High steam pressures and temperatures may make reductions in the size and weight of a given propulsion plant possible, or permit a higher horsepower installation in the same space. During World War 11,most combat naval vesgels operated at 600 psig-850 F while steam to 450 psig-750 F was widely used in merchant ships. In the postwar era the Navy advanced to 1200 psig-950 F (nominal) for its combat vessel construction. In the late 1940,s and 1950's a significant number of merchant vessels appeared using steam at 600 psig-850 F and 850 psig-850 F. By the 1960's almost all new construction used 850 psig-950 F steam; several large vessels used steam (in some cases with reheat) a t 1500 paig-950 F. Machinery plants utilizing steam st pressures of 850 to 1500 psig and temperatures from 950 F to 1000 F are characteristic of most commercial steamships built during the 1970's. The quantity of steam produced by a marine boiler can range from as little as 1500 lb/hr in small auxiliary boilers to over 400,000 1b/hr in large main propulsion boilera. Steam outputs of 750,000 lb/hr or more per boiler are practical for high-power installations.
2.3 Heat Balances. The fuel cost per shaft horsepower is one of the deciding factors in establishing the characteristics of the boiler installation and whether or not the installation is economically sound. The fuel rate can be decreased by the use of higher steam pressures and temperatures or a more sophisticated cycle can be employed by the use of reheating, economizers, and/or air heaters, more stages of feed heating, etc. The designer must analyze these factors in light of initial cost, maintenance, weight, and space requirements versus the savings resulting from increased thermal efficiency. As steam pressures increase, it is essential to use additional heat-reclaiming equipment in the boiler unit. This is because of the corresponding increase in saturated ' steam temperature which results in a higher gas temperature leaving the boiler bank and thereby reduces the boiler efficiencyat a given firing rate. Reheating the steam improves thermal efficiency but requires larger boilers and special provisions to protect the reheater during astern operation. High steam pressures and temperatures, along with reheating, are more likely to be used in installations of 30,000 shp and up, where the value of the fuel saved may well justify greater initial cost and cycle complication. In addition, the utilization factor or load factor in such vessels is apt to be much higher, giving added impetus to the establishment of more efficient designs [9]. It is from the detailed heat balances prepared by the marine engineer that the quantities of steam and feedwater flow are determined. In the usual plant from two to four stages of feedwater heating are used to supply water to the boiler at temperatures from 270 to 400 F. Boiler efficiencies of over 90 percent are possible. However, to minimize corrosion and maintenance in the cold-end heat exchangers and uptakes, it may prove advantageous to limit the boiler efficiency to 88.5-90 percent with some fuels. Fuel oils vary widely in quality and often contain significant amounts of sulfur which can form sulfuric acid if there is condensation in the exhaust gas path. Corrosion and maintenance costs should be balanced against the possible savings in fuel costs derived from a higher boiler efficiency. 2.4 Fuels and Methods of Firing. The characteristics of the fuels which will be available to the ship in its usual trade should be established early in the design process. This will permit the optimum selection of equipment for burning the fuel and cleaning the boiler. In addition, a suitable selection of uptake temperatures and materials can be made for the entire boiler plant so as to reduce corrosion and maintenance problems. Most marine boilers are oil-fired, with wood, gas, and coal-fired boilers less common. Wood firing is generally confined to riverboats operating on streams with an abundance of nearby timber and is not an important
i11nl oxcept perhaps in some remote parts of the world. ( I~r~-fired boilers are used primarily on power or drill Imrgtrs which are fixed in location and can be supplied ~ I I I I I I~hore. At sea, tankers designed to carry liquefied ~iul,~rrr~l gas may use the natural boil-off from their cargo &#a lllulks as a supplemental fuel. This cargo gas I~~~iI-off is collected and pumped to the boilers where it is I~~rrncrtl in conjunction with oil. The oil burners serve BWpilots to provide ignition stability and also to augment tire l l t r l ~ available t from the gas. The quantity of boil-off ~vrilt~ble from the liquefied natural gas is a function of r t r ~ hiont r sea and air temperatures, the ship's motion, and It10 trnrgo loading, among other things, and may vary I r c ~ nduy ~ to day. C !old-lired boilers have persisted chiefly in older vessels trljer~~l~irlg on lakes and rivers, and in ferries, colliers, tti&dI t~ndtowboats operating in coastal services. Their t n ~ ~ l l l ~have o r ~ decreased steadily year by year as labor r114tw rino and air pollution control is expanded. M o ~ lcoal-fired marine boilers used hand or stoker n r i ~ ~ aThe . use of stokers, particularly the spreader Bylre, gormitted firing rates per square foot of grate l ~ ~ l r r u~tpproximately o 40 to 50 percent in excess of those tor ha~idfiring. This resulted in boilers which were far nlura aompact and lighter than those designed for hand I/glrrp; but even they were much larger and heavier than u(i4rsd boilers designed for comparable steam outputs. !3rllv~rizedcoal firing, widely used ashore, has seldom Rri~usud a t sea since the,furnace volume necessary for d@iii~m,Lmvel, low heat release rate, and satisfactory kmbuatian requires a tall boiler. The high fly ash kdllrg of the flue gas aggravates tube erosion, slagging, dtaak emission problems. Qilwwore used as boiler fuels as early as the 1870's but f#d nos aohieve widespread use until the automobile age fgqulrecl a world-wide petroleum industry. Compared &$ ei,har fuels, oil is easily loaded aboard ship, stored, lnbroduced into the furnace; and the firing equiplVequireslittle costly maintenance. The small l$#i@unCof ash and contaminants it contains does not mdre t,ha extensive ash handling facilities required for
ma6
@&jl flrlng,
1) ~lrelrldbe recognized that fuel oils from different WIFOPH, while similar in heating value, have varying r n ~ u n b aof contaminants which may be harmful in
mpiew ways. The major contaminants consist of @@a of vfinadium and sodium. As a class, they are ~ l e A"a~h"and their presence must be fully taken into l@@@~irt by the designer. Likewise, the sulphur content wry over a range from almost none to as much as &f psroallt in "sour" crudes; sulphur has a decided en the cycle efficiency which can be obtained @t!t&rb tierious corrosion in the economizer, air heater,
a&&
~ptrtlees.
Tkr oompounds of vanadium and sodium affect the 11
af the superheater. If oils to be burned in a
trtde are especially rich in these constituents, r sriperlla~tttorcan be designed with tube metal temCinltricrr lower than normal to avoid the possibility
91
of severe slagging and tube metal corrosion problems. Cold-end heat exchangers designed with full recognition of the sulphur content present in the fuel will experience a minimum of corrosion and expensive maintenance. A boiler designed to take advantage of low-cost residual fuel oils can always burn lighter fuels if the situation justifies it. However, a boiler with tightly packed heating surfaces designed for light oils such as diesel or aviation turbine fuels would not perform satisfactorily on residual fuels for very long. Gas-side fouling and oil burner and combustion problems in the furnace could be anticipated. 2.5 Effect of Ship Delign and Other Machinery on Boiler Design. Factors such as space, weight, and the
requirements of the regulatory bodies are major considerations in the design of a boiler. In addition, however, the prospective vessel owner or his naval architect may have preferences regarding the boiler design and specific design requirements. These preferences may include the number of boilers, types of boilers and their arrangement, locations of major connections, the use of economizers and/or air heaters, fining, and evaporative ratinga, and the type and method of firing. Life-cycle costs can have a bearing on the preference likewise, since the total cost and labor involved in maintaining a previous design or construction may be reflected in the owner's specifications and result in the selection of an improved design and construction. a. Space. The space provided for the machinery is held to a minimum by the naval architect because the space occupied by the machinery produces no revenue. The boiler designer is usually required to adapt the boiler design to the available space. The boiler height may be limited by deck or machinery casing locations. The fore-and-aft or depth dimension of the boilers may be controlled by bulkhead locations, access, or tube renewal space requirements as well as the location of control consoles, main engines, etc. To a large extent the aviilable space determines the economy of the design. A height restriction is particularly serious, since it usually necessitates increased boiler width or length to obtain the required heating surface. This generally results in a marked increase in boiler cost, weight, and the base area occupied. b. Weight. With drum-type boilers, the minimum ~ efficiency is obtained with rninimuql weight f o maximum furnace depth, maximum tube length, and the maximum number of tube rows. Limiting the height may restrict capacity because of reduced circulation. It may also result in tube slopes and in burner clearances less than the minimum necessary for a good design. In header-type boilers the width is changed by increasing or decreasing the number of header sections, and the height is varied by changing the number of tube clusters in a header. Because of reductions in the number of boiler sections and the length of the steam drum, it is readily evident that long, narrow, and high boilers lead to minimum weight. Further, since the maximum efficiency for a given heating surface is obtained with the
MARINE ENGINEERING veatest numb& of tube rows in height, header-type these limits may be modified in the special specifications boilers always should be arranged with the maximum issued for a particular class of vessel. m he Maritime height, rnmimum length, and minimum width which are Administration follows a somewhat similar procedure and usually establishes evaporative and furnace heat compatible with the design conditions. The minimum weight of any type of boiler will vary release rates for each design. considerably with desi@ conditions;increases in evapora- . 2.6 Boiler Design Criteria. heo ore tical and practical tive rating, burner capacity, or air pressure decrease the considerations have led to the establishment of boiler weight of a boiler design@ for a specified steam output. design criteria in a number of areas not directly associated With a fixed evaporative rate per square foot of heat- with the regulatory bodies' rules, which concern mainly absorbing surfaoe, the weight of a boiler per pound of pressure-part scantlings and construction techniques. generated will be less for boilers with greater steam The design criteria are most important in the areas of output, since certah boiler parts remain fixed in size and combustion, heat absorption rates, circulation, and pressure drops through the boiler system. They provide weight over a reasonable range in capacity. Weight is greatly dependent on space also. Generally the yardstick by which various boiler designs Can be the larger the physical dimensions of a boiler for a given compared for their suitability for specific applications. a. Combustion. At the heart of a successful boiler output, the greater its weight. is a properly designed furnace and fuel burning systemThe ocean environment is no place Regulations. to test unproven principles. This became evident in the If the fuel supplied to the furnace is not burned cleanly construction when it was and completely within the furnace throughout the range early days of recognized that some rules and regulations were necessary of operation, it will not be possible to accurately predict to protect life and property. These rules were not the performance of the evaporator-superheater comintended to inhibit the designer or innovator but rather bination. For example, the total steam generated may for comparison of be insufficient, the steam temperature may be incorrect, to provide a sound basis and or the efficiencymay be lowered by incomplete combusnew designs with older successful designs. tion Or improper excess air. Disastrous boiler explosions, common to both marine A number of criteria by which combustion in furnaces and stationary boilers, resulted in the establishment of a boiler inspection senice and strict regulations can be gaged and by which different furnaces Can be care, and operation of compared have been developed. In"general1 with the governing the steam boilers. In the design of marine boilers the exception of the furnace heat absorption rate which is applicable regulatory rules and standards must be rigor- derived from the actual heat transfer calculations ously followed. Most units built for American-flag developed for the furnace, they are empirical relationthe requirements of the United States C o ~ t ships with little theoretical value; however, they can be ships used to compare similar boiler designs provided their Guard and the American Bureau of Shipping. Boilers for naval combatant ships are built in strict limitations are recognized. The criteria most fI'equent1~ used for these comparisons accordance with Navy specifications, although for are: auiiliary naval vessels the use of the United States Coast Guard or the American Society of Mechanical rate per cubic foot of furnace volume. Heat Engineers codes often is permissible. For foreign-flag ~ i rrate i per ~ ~square footof radiant heat absorbing ships, the rules and regulations of other midatory surface. bodies would apply. In addition, many shipyards and Heat absorption rate per square foot of radiant heat operators of large fleets have established their own absorbing surface. supplementary rules and regulations. Since the requirements of the various regulating and A brief review of these factors will sewe to indicate their inspection groups differ, specifications must be clearly importance and usefulness. The heat release rate per cubic foot of furnace volume defined to assure fabrication and installation of boilers which will be approved by the boiler inspectors. Fur- is useful in comparing geometrically similar furnaces, ther, it is important that all competitive designs be to the but while widely used because of its simplicity, it is not A design difference caused by the an important criterion. The heat released is the product same me of inapplicable specifications could be the deciding of the hourly fuel rate and its higher heating value, factor in final cost or wei&t evaluations, particularly ignoring any heat above 100 F in the combustion air. If radiant heat absorption rates, furnace gss temperaon high-pressure unito where a difference in pressurepart thickness might involve not only price and weight, tures, and furnace tube metal temperatures are satisfactory, the only limitation on the heat release rate Per but also design and fabrication changes. ~~~t rules pertain to const~ctionand the inspection cubic foot of furnace volume should be that imposed of materials, and establish very few by the ability of the firing equipment to maintain good and The use of a high1 yet satisperformance limitations. ~ l t h o u g hNavy specifications combustion conditions. furnace volume heat r f ? l ~ ~late e peat1y factory, rates per cubic foot of furnace limit the heat installation of high-capa~ib~ lightweight facilitates the volume, per square foot of radiant heat absorbing surface, and per square foot of total heating surface, boilers in a minimum of space.
BOILERS AND 'I'll() temperature within a boiler furnace can be ~llilll~rolled to a large extent by the effective radiant heat r~lno~~bing surface (RHAS) present in the furnace [lo]. 1 IPIIII is radiated from the flame envelope to thee heat t~lluorhingsurfaces with the uncooled refractory surfaces n i \ l ~ i as r ~ ~an intermediary, receiving heat from the flame ru~dl111cnre-radiating most of the received heat back to 1 II* ll~uneand cold surfaces. For a given heat input or ~ ~ " rate, I I K the heat absorbed per unit area decreases wlIllr ILILincrease in total RHAS. The greater the RHAS ~ I I * ~t'aaterwill be the total amount of heat absorbed by (Itn Fllrnace. Therefore, the temperature of the gases \
COMBUSTION
93
boiler has more demands placed on i i than a comesponding shoreside boiler. In addition, the heat input and the steam output of the marine boiler are probably higher than for a comparable application ashore. It is customary to consider a momentary roll of 30 deg from the horizontal and a momentary pitch of f5 deg when computing static and dynamic loads. In establishing circulation, boilers are u s p d y designed for a permanent list of 15 deg and a permanent trim by the bow or stern of 5 deg. The latter, when coupled with the momentary pitch of 5 deg, means that in the fore-and-aft direction, the boiler may be as much as 10 deg from the horizontal. The arrangement of the tubes and steam-
,
.. .
n
f
MARINE ENGINEERING boiler must likewise be capable of prolonged periods of steady operation a t its design rating. Also, in port it may be subjected to long periods of operation at low or minimum outputs. Cleaning, with the exception of the daily use of the mot blowers or occasional attention to the atomizers in the oil burners, is normally deferred to the annual or biannual period when the vessel is in a shipyard for other maintenance. This must be fully taken into account by properly locating soot blowers so they are effective; by using the optimum burner combinations for the range of fuel-oil types anticipated to be bunkered; and by using the best possible arrangements of economizer, air heater, boiler furnace, and generating surfaces to pinimize fouling. must also include margins in the scantlings The ---- desim of tubes, supports, casings, and other parts exposed to corrosive flue gases or waterside contaminants. Simple and easily accomplished maintenance procedures can also do much to assure that the boiler will be available to meet the ship's requirements. The duty cycle may also have a pronounced effect on the number of boilers selected. A single boiler may be employed in ships of up to about 90,000 shp. Two or more boilers may be selected for higher power levels or where redundancy is desired or required. Single-boiler vessels have proven reliable in service and should continue to do so. This is in part due to the fact that a boiler kept continuously in service reaches thermal equilibrium and can have the waterside chemistry optimized. In general, from a boiler performance point of view, the least number of boilers which can deliver
the required steam will prove to be the best selection for any particular vessel. e. Automation. Widespread use of automatic controls @ndmonitoring equipment has made bridge control of the power plant possible and has permitted a reduction in the number of watch-standers in the machinery space. These desirable improvements have added additional -considerations .to the problem of designing a suitable boiler. Of prime importance is a fuel burning system that can respond rapidly throughout the range of operation from standby to maximum power without a fireman's attention. It must do so to prevent excursions in steam pressure and reduce water level fluctuations (shrink and swell due to changes of the volume of steam present in the boiler), which might result in water carry-over into the superheater [12]. Burners can be designed to operate over the full boiler range with all burners in service, or other burner types with less range can be sequenced, that is, placed in or out of service on command by the control system. Suitable flame-monitoring safeguards and purge interlocks are necessary in varying degrees of complexity depending on the extent of manual supervision desired. Feedwater regulators, steam temperature controls, d a t a logging equipment for flows, pressures, temperatures, levels, etc. are all available from the simple to the ultrasophisticated. The owner and his naval architect usually select the scope of equipment and advise the boiler designer so that the boiler and burner combination can be made compatible with it. See Chapter 21 for additional discussion regarding automation and controls. -
atttl in part on the space available for the installation ant l its operating requirements. 'I'ho quantity -of fuel required is determined from the ~ltwirod steam generator efficiency, the given steam prtrnHure, temperature, and flow, the feedwater temperaI,II~'o, and the heating value of the fuel. 'I'ho fuel characteristics and quantities establish the' I~lrlburning equipment to be employed. This in turn ICI~H the excess air requirements. Combustion calculal l l r l r l ~are next made to determine the hourly quantities rlf llue gas flowing through the unit. The exit or stack baa tomperature to which the flue gas must be cooled b nohieve the desired efficiency is determined (Fig. 16); R I I ~if experience indicates that it is attainable or otherw l ~ t r natisfactory, the design can proceed. If not, a~rr~t~hor selection of efficiency must be made and the ealaulations repeated. 'I'ba furnace exit gas temperature is next calculated. Ell@ value is dependent on the radiant and convection 11ewt-transfer surface installed in the waterwalls, floor, tr~nf,t~ndscreen (radiant only) as well as the extent of refractory present. Next, the gas temperature drops &acl tho heat absorbed by the screen and superheater are dsbarmined. The size and spacing of tubes and the &mount of surface are assumed initially. These are lhrn modified to provide the desired steam temperature rrild cronservative tube metal temperatures as necessary. V~uallyseveral screen and superheater combinations are Invemtigated to determine the most economical solution. r heater surfaces ke gas temperast outlined, initial aterials for tubes,
the heating surfaces established, the draft loss all components is calculated. If the draft loss the capability of the fan desired, the heat drafts previously calculated are adjusted he tube spacing, number of rows crossed or height of the boiler components.' A ers may be necessary ce of draft require-
Section 3 3.1
General. The fundamental boiler design prob-
lem is to determine the proper proportions of the various heatrabsorbing surfaces to use the maximum heat available in the products of combustion. A proper design will accomplish this at the lowest cost on a lifecycle basis. Each component must be integrated with the other elements of the unit to provide a balanced design in which the first costs and fuel, maintenance, and operational costs will be a minimum over the useful life of the ship. In no way must safety or reliability be compromised by these cost considerations. For the steam generator system, the following must be considered :
1. Fuel burning equipment 2. Furnace 3. Boiler generating surface 4. Superheater (and reheater if used) 5. Economizer and air heater
6. Attemperator (or control) and auxiliary desuperheaters 7. Circulatory and steam separator system 8. Casing and setting 9. Cleaning equipment 10. Safety valves and other mountings 11. Feedwater and treatment 12. Foundations and supports 13. Combustion air supply system 14. Uptake gas duct system and stack These considerations require many interrelated steps. In most cases, a number of assumptions must be made in order to initiate the design. ks the design calculations proceed, the assumptions are refined to achieve the desired accuracy in the final analysis. The first step is the selection of the basic type of boiler, superheater, and economizer or air heater (or both) to be used. This selection is based in part on preference
95
BOILERS AND COMBUSTION
I
drops of water and steam through all comm the economizer feedwater inlet to the superuted. They, in turn, estabeconomizer design pressures tho safety valve settings. A circulation analysis @aprepared using the heat absorptions determined . From this, the bes are adjusted as for each design. er can make very ntially reduce the
@,P
Fuel Combustion. The basic function of a
ilrp frirnace is to generate the maximum amount of rrb Imm a given quantity of a specific fuel. A useful
RAOlATlON AN REFERENCE 0
FOR UNITS WITHOUT STEAM AIR HEATER
STbCK GAS TEMPERATURE, F
Rg. 16
Efficiency v* stack gar temperature
secondary function is to generate steam in the furnace wall tube circuits. The theoretical aspects of combustion have been well known for many years. However, the achievement of good combustion within the furnace of a relatively small marine boiler requires practical knowledge and experience. Complete combustion can be obtained provided there is sufficient time (a function of furnace volume), turbulence (provided by the geometry of the burner assembly), and a temperature high enough to provide ignition. Combustion may be defined as the chemical combination of oxygen with the combustible elements in the fuel. The common fuels have only three elemental constituents which unite with oxygen to produce heat. The elements and their compounds, as well as their molecular weights and combustion constants, including heating values, are given in Table 1. Oxygen combines with the combustible elements and their comgounds in accordance with the laws of chemistry. Typical reactions for the combustible conatituents of fuel oil, based on the assumption that the reaction is completed with the exact amount of oxygen required, are : for Carbon (to COa) for Hydrogen (to HzO) for Sulfur (to SOa)
+ + +
++ +
C 0 2 = COZ AQ 2H2 0 2 = 2Hz0 AQ 2s 302 = 2508 A Q where A Q is the heat evolved by the reaction. The heat evolved or heat of combustion is commonly called the "fuel heating value" and is the sum of the heats of reaction of the various constituents for one pound of the fuel considered. The heating value of a fuel may
96
BOILERS AND COMBUSTION
MARINE ENGINEERING
1113 calculated from theoretical considerations or may be clt!l,ormined, for an actual oil, by burning a sample in a I)olnb calorimeter (see Chapter 23 for additional discusi4o11in this regard). 111 testing fuels by a bomb calorimeter to determine the l l t r ~ b tgiven up, two values may be reported: the higher ([)I' Kr088 Or upper) heating value and the lower Or net Il~~~ltling value. For the higher heating value, it is nafl''med that any water vapor by burning the I1,Vtlrogen constituent is d l condensed and cooled to the l11ll~i1~1 temperature in the calorimeter at the end of the tsrl,. The heat of vaporization, about 970 Btu/lb oil, is inoluded in the reported heating value. For the lower ~isrtl1iug value, it is assumed that none of the water vapor mnclo~~sesand that all the products of combustion vermin in a gaseous state. In the United States higher I ~ e ~ t ~ vdues i n g are used as they are available directly fl'c~lllthe calorimeter determinations and because of the @stnll>li~hed practice of buying fuel on a higher heating vnlue basis. The lower heating values are generally ~irreclia European practice. Fuel Analysis. For design and comparative IrlitlptrNos, the standard reference fuel oil is #6 fuel oil [@uelrur C) having the following characteristics [13]:
CHEMICAL COMPOSITION (percent by weight) Carbon 87.75 Hydrogen 10.50
Total
100.00
tho following expression :
By weight By volume
%OXYGEN%NITROGEN 23.15 76.85 21.00 79.00
The rare gases are included as part of the nitrogen constituent. Air is assumed to be supplied to the forced-draft fan at a temperature of 100 F, a rklative humidity of 40 percent, and a barometric pressure of 29.92 in. Hg. Under conditions air has the following physical prope*ies: Dry-air density, lb/cu ft MoistureJ lb/lb of dry air Mixture density, lb/cu ft Specific heat
0.0709 0.0165 0.0701 See Fig. 3 of Chapter 2
Based on the foregoing fuel and air standardsJ analysis will show that the s~ic-,iometrical or theoretical quantity of dry air to burn one pound of fuel is 13-75Ib. From this, the following quantities of air for various excess percentages are determined : Excess air, percent Dry air, It, Moisture, Ib Moist air, lb Volume, cu f t (at 100 F, 29.92 in. Hg) dry air moist air (40% RH)
0 5 10 15 20 13.75 14.44 15.13 15.81 16.50 0.23 0.24 0.25 0.26 0.27 13.98 14.68 15.38 16.07 16.77 194 200
204 210
213 220
223 230
233 240
The ultimate analysis of the fuels actually encountered in service varies from that of the standard reference fuel. Figure 17 shows the effect of these variations on
98
MARINE ENGINEERING
BOILERS AND COMBUSrlON
MARINE ENGINEERING Table 2
Oil Burner Clearances
PARTIAL STUD TUBES
FULL STUD TUBES TYPICAL STUD-TUBE WALLS
wider angle is employed to shorten the flame length and produce a wide bushy flame while a narrower angle increases flame length and decreases width. The burner manufacturer should always be given the opportunity to review the projected furnace design so the best possible installation can be obtained. Generally suito' able burner clearances are shown in Table 2. When firing Bunker C oil, it is customary to use the minimum clearances established by experience. These may be Fig. 19 Change in efficiency vs. load decreased perhaps by six inches, if distillate oils are fired. Furnace depths of watertube boilers which are front-fired are usually limited to a minimum of six feet boilers the large amount of fuel and air to be introduced although there are highly rated boilers in service with into the furnace necessitates a multiple burner instal- furnace depths of only five feet. The selection of the oil burner must also include the Each size burner has a minimum rate of operation type of atomizer to be used. There is a wide variety of below which it becomes unstable and there is risk of atomizers from which a selection can be made. The losing ignition. In part this is a characteristic of the alternatives include: steam atomization (internal mix), burner, but the forced-draft, fuel, and control systems steam mechanical (external mix), straight mechanical, also have an influence. The minimum rate is of great return flow, rotary cup, and others. Of these types, the a much simpler plant results when all internal mix steam atomizer has the greatest turndown importance burners can be left in service at all times. When in and provides the smallest and most uniform particle port or during rnanuevering conditions, the minimum size over its wide range of operation. Development0 oil flow capability must be less than that required by have materially reduced the quantity of steam required the plant demand, if frequent safety valve popping or (80 to 120 lb/hr-burner depending on the maximum oil steam dumping is to be avoided. Both of these actions capacity) so that earlier objections to the loss of evaporated water have been more than offset by the other waste steam and lead to increased maintenance, Burner sequencing can be used effectively to follow advantages. The uniform and finer article size has the load demand where burners with limited range or provided more surface area for combustion of the fuel lower higher-than-desired minimum flows are used. Solid- droplets. This has permitted less excess air and necesstate, computer-controlled logic systems are often used draft losses since the high air velocity to sequence burners; hovbever, this equipment canincrease sary to provide the turbulence to burn larger droplets i~ no longer required. costs considerably [15]. The number of burners selected usually results in a Care must be taken in arranging the burners to provide for even air distribution to each burner within burner draft loss equivalent to about 35 to 50 percent of combustion with a minimum of the total draft loss of the boiler unit. The burner draft the windbox to varies with the volumetric flow of air through it. excess air. The clearances between the burners and the loss At any given air flow, a change in the temperature of tho to prevent interference furnace walls must be air will increase or decrease the draft loss in the ratio of The furnace volume must be large and impingement. enough to provide the time necessary for complete the change of absolute temperatures. In desiping a to take place before the gases enter the super- boiler with an air heater, it is standard ~racticeto limit heater screen. Satisfactory combustion has been ob- the air temperature leaving the air heater and enter in^ tained at furnace release rates of up to 1,500,000 the burners to no more than 600 F and refer ably l e s ~ to assure long life and prevent overheating of the burher Btu/cu ft in marine boilers. If the preliminary design ~ i e l d san excessive air parts. Each burner manufacturer has his own recommended temperature, the designer must reapportion the surface*, clearances and the shape of the flame can be adjusted to possibly adding a small economizer, to reduce the air This is some extent to modify them when necessary, done by changing the spray angle of the atomizer. A heater air outlet temperature to an acceptable value.
TANGENT TUBES
MEMBRANE WELDED TUBES TYPICAL BARE-TUBE WALLS
Fig. 20
TUBE AND TILE
A
Furnace wall construction
1 02
BOILERS AND COMBUSTION
MARINE EN
estimates of furnace exit gas temperatures were not necessary because of conservative firing rates and the use of saturated steam. Those units which generated superheated steam usually had several rows of boiler tubes between the superheater and the furnace. Consequently, a large error in the calculated furnace exit gas temperature had very little effect upon superheater performance. In units with superheaters located dose to the furnace, however, the furnace exit gas temperature must be determined accurately to assure a satisfactory superheater design. In addition, an accurate determination of the heat absorption in the various furnace waterwall areas is necessary to provide adequate water circulation with a practical number of supply and riser tubes. When estimating the furnace gas temperature, most designers use formulas based upon the Stefan-Boltzmam law, which states that the heat absorbed by radiation is proportional to the difference between the fourth powers of the absolute temperatures of the radiating bodies and receiving surfaces (see Chapter 2). However, in a boiler furnace the exact determination of radiant heat transfer, or heat absorption, is extremely complex and depends upon: the furnace size and shape; the radiant beam (mean distance from the radiating gas mass to the absorbing and the re-radiating surfaces); the partial pressure of the products of combustion; the amount, type, and effectiveness of the heat absorbing surfaces; the ratio of the heat absorbing to the refractory surfaces; the type, quantity, and heat content of the fuel; the amount of excess air; the temperature of the combustion air; the latent heat losses; the emissivity of the various surfaces and the radiating mass of gas; and the flame luminosity. Designers usually calculate furnace exit gas temperatures and heat absorptions by rational methods and then, as a check, plot the calculated values against empirical data derived from boiler tests 121. b. Radiant Heat Absorbing Surface. In evaluating the radiant heat absorbing surface, the flat projected areas of the walls and tube banks are used. The spacing of the tubes in the boiler bank adjacent to the furnace has no effect upon the furnace temperature; but with widely pitched boiler tubes, a large percentage of the radiant heat is absorbed in the tube rows behind the furnace row. Furnace waterwalls and roofs usually consist of bare or covered tubes (Fig. 20) and, with the exception of bare tangent tubes or welded walls, the effectiveness of the absorbing surfaces is less than the black-body coefficient of 1.0 considered for the furnace rows of boiler tubes. The furnace gas temperatures usually are not accurately estimated in preliminary analyses since the general design characteristics are of primary interest, and an approximate estimate of furnace gas temperatures and heat absorption rates can be made with knowledge of the boiler and the firing conditions. Thus, with the assumed excess air, the heat content of the products of combustion and the adiabatic temperature can be determined. Further, the approximate furnace size
provides an indication of the water-cooled surface8 and estimates can be made of the surface absorption effectiveness and the expected furnace gas temperature. In approximations of this nature it is usually desirable to estimate both the furnace temperature and the heat absorbing surface on the low side when firing oil. This increases the estimated furnace heat absorption and assures a margin of reserve in the final design. However, with coal firing it is more important to estimate the furnace gas temperature on the high side to preclude the possibility of operating with furnace temperatures above the initial ash deformation temperature. In a boiler furnace, both the furnace exit gas temperature and the heat absorption can be changed appreciably, for a given firing rate, by varying the amount of radiant heat absorbing surface. The furnace gas temperature and heat absorption also can be lowered, at any firing rate, by increasing the excess air (Fig. 21), except when operating with a deficiency of air. The additional air increases the weight of the products of combustion per pound of fuel fired. This decreases the adiabatic temperature since there is less heat available per pound of products of combustion; and, as indicated by the Stefan-Boltzmann law, lowering the radiating temperature reduces the heat absorption rate. Generally, the radiating temperature is assumed equal to one third of the adiabatic temperature plus two thirds of the furnace exit gas temperature. c. Heat Absorption Rates. The furnace heat absorption rate per square foot of radiant heat absorbing surface increases with larger heat release rates. However, the percentage of the total heat released which is absorbed in the boiler by radiation decreases with an increase in firing rate, and varies from as much as 50 percent, or more, at the lower firing rates to about 15 percent at the higher firing rates; see Fig. 22. This results from the fact that the adiabatic temperature remains practically constant, except for changes due to variations in excess air and combustion air temperatures, over the entire range of boiler operation, while the temperature of the gases leaving the furnace and entering the tube bank increases with the firing rate. Even though the furnace heat absorption rates may be conservative, the furnace exit gas temperatures may be excessive with respect to ash fusion temperatures and slagging. This is true particularly in coal-fired boilers where the gas temperatures entering the tube bank should be less than the initial ash deformation temperature. Because of the lower ash fusion temperatures of oil slags, they pass out of the furnace in a gaseous or molten state and are not amenable to control by reducing the furnace exit gas temperature. They must be considered in the design of the superheater. d. Tube Metal Temperatures. In boilers, the heattransfer rate across the boiling water fdm on the inside of the tubes may be as high as 20,000 Btu/ft2-hr-F; however, when estimating tube metal temperatures, a transfer rate of only 2000 Btu/ft2-hr-F is usually assumed in order to provide a margin against the resis-
103
EXIT-WITH 15%
FIRING RATE,PER CENT OF FULL OUTPUT FIRING RATE,PER CENT OF FULL OUTPUT
.
I I l k c t of excess air on odlobotic tind furnace gar temperature
Flp. 22 Relotianhip of rodlon) heat absorption ond Aring rote
ratings, including port loadings. However, at t.he same time they should not be so high as to cause high casing temperatures or excessive furnace maintenance. Because of the requirements for exceedingly lightweight and compact units for naval installations, evapntly, with a steam pressure of 600 psig orative ratings in naval boilers are 3 to 4 times greater steam temperature) and a heat input than those common to most merchant installations. Consequently, the furnace exit gas temperatures in the full-power to overload range are about 2800 to 3050 F when firing oil with approximately 15 percent excess air. Adiabatic, or theoretical, flame temperatures are about 3450 to 3500 F with oil firing, 15 percent excess air, and 100 F combustion air. With combustion air temperaappreciable and it is good design practice t o tures of 300 to 350F, the adiabatic temperatures L tolerance for variations in the quality of the increase to approximately 3650 to 3700 F. Although furnace heat release rates vary considerably, practically all oil-fired merchant boilers are designed for heat release rates of 65,000 to 125,000 Btu per cubic foot of furnace volume per hour at normal rating-approximately 15 to 20 percent of the corresponding full-power heat release rates on naval boilers. The heat release rate per square foot of radiant heat a b s o r b i surface is generally in the range of 200,000 to 250,090 Btu per horn on merchant boiler designs.
104
-.
MARINE ENGINEERING the minimum longitudinal tube pitch (direction parallel to the drum and perpendicular to the gas flow) consistent with good manufacturing practice and acceptable drum design, unless the draft requirement or the type of fuel fired dictates the use of a greater pitch. Manufacturing and fabricating practices permit the use of +-in. metal ligaments between 1-in. or la-in.-OD tubes. The circumferential, or back, pitch (direction parallel to the gas flow) of the tube usually is set to maintain circumferential or diagonal ligament efficiencies2 equal to, or better than, the longitudinal ligament efficiency in the drums. Tube arrangements utilizing a minimum back pitch reduce the drum periphery required for a given number of tube rows and allow the use of smallerdiameter steam drums provided the steam drum release rates are satisfactory. With such arrangements, the size and weight of the boiler can be reduced. When designing for high steam pressures, it is often necessary to increase the tube spacing in order to improve the ligament efficiency and reduce the thickness of the drum tube sheet [l6]. If this is not done, large thermal stresses may be set up in the tube sheet. It also i~ possible to maintain close tube spacing and yet reduce the drum tube sheet thickness by using tubes with the ends swaged to a smaller diameter. The number of tube rows installed should be limited so that an impractically large steam drum diameter i~ not required and so that heat absorption in the last tube rows is adequate to maintain good circulation. The tube length should be such that the total absorption per tube does not result in too high a proportion of steam it1 the water-steam mixture leaving the upper end of tho tubes. b. Header-Type Boilers. Single-pass header-typo boilers (Fig. 3) generally have two rows of 2-in. t u b e ~ above the furnace and if-in. or 1-in. tubes in tho remainder of the bank. In these boilers a group or cluster of fourteen 1-in. tubes can be substituted for ono of nine la-in. tubes. Thus, in boilers having the samo width, length, and number of tube clusters in height, 25 percent more heat absorbing surface can be installed by substituting 1-in. for la-in. tubes. However, tho advantages resulting from the compactness of the 1-in.tube boiler must be balanced against the greater tolerance provided by the la-in.-tube boiler for poorer feedwatcr quality. For the new header-type boilers that arc1 installed, chiefly in motor vessels for auxiliary steam purposes and in drill barges and dredges, the feedwator quality is apt to be such that the selection of larger tubo sizes will offer more reliability. c. Boilers Delivering Superheated Steam. Practically all marine boilers built recently deliver superheated steam from convection-type superheaters. In these boilers, the generating tube bank is arranged in two
Naval boilers are designed for ratings four to five times greater than those used for merchant marine boilers. Radiant heat absorption rates vary greatly depending upon the firing rate and the amount of cold (watercooled) surface in the furnace. Generally, a radiant heat absorption of 120,000 Btu per square foot of cold surface per hour is considered satisfactory for continuous overload operation of merchant boilers with treated evaporated feedwater. This results in an absorption of about 100,000 Btu per square foot of cold surface per hour at the full-load rating. There are merchant boilers in continuous service with radiant heat absorptions of approximately 150,000 Btu per square foot of cold surface per hour; and most naval boilers have been designed for radiant heat absorption rates of 150,000 to 200,000 Btu per square foot of cold surface per hour at overload rating, but operation a t this rating is infrequent. 3.4 Boiler Tube Bank. The arrangement of the boiler tbbe banks is established after development of the preliminary furnace size. The simplest type of tube bank is that of a boiler delivering saturated steam. Usually two sizes of tubes are used in such banks. The tubes in the rows adjacent to the furnace absorb considerably more heat than those in the other rows and, therefore, should be of larger diameter to increase the water flow. The total heat input to the furnace row tubes is the sum of the radiant and convection heat transfers; in general, the convection heat transfer is approximately 5 to 20 percent of the radiant heat transfer. This relatively wide range in convection heat transfer results from variations in tube diameter, tube pitch, gas mass flow rate, and the temperature difference between the products of combustion and the tube surface. The number of tube rows installed is primarily dependent upon the circulatory system and the desired gas temperature leaving the tube bank. The gas temperature leaving the boiler tube bank varies with changes in steam pressure, firing rate, and tube size and arrangement (the tube arrangement may be either staggered or in-line). However, sufficient boiler heating surface must be installed to obtain exit gas temperatures which result in economical operating efficiencies and do not require excessive stack and breeching insulation. Generally, the exit gas temperatures should not exceed 750 F unless economizers or air heaters are used. The resistance to gas flow can be varied appreciably in drum-type boilers by changing the pitch of the tubes in a direction perpendicular to the gas flow. This change is not possible on header-type boilers because of the fixed tube pitch and, therefore, variations in resistance to gas flow must result from changes in boiler width, tube length, and the number of tube rows. a. Drum-Type Boilers. Mbderately rated drumtype boilers usually have 13-in. tubes in the furnace roes, but these are increased to 2 in. in boilers of higher rating. One-inch and I&-in. tubes are common in the 2 Ligament efficiency is the relative strength of the ligamenln There is no standard pitch for tubes between main tube adjacent tube holes in a drum or header as compared with in drum-type boilers. However, it is customary to use a drum or header having no holes.
I
BOILERS AND suc~(~ior~~. The section between the furnace and the +!~~l~n~~ltoater is known as the "waterscreen" and the other F~UI~~~IO installed II, beyond the superheater, is called the " l ~ ~ ~ i l rbank" ir. or "generating bank. " 'I'l~ti~ i z eand arrangement of the waterscreen greatly r1l;fecrln the design of the superheater. A superheater I~~c~~iiCtsl d.oser to the furnace behind a few rows of widely j~ll.irl~n(l tubes in the waterscreen provides a relatively ili~l, nl,nl~mtemperature characteristic over a wide range ~ r l1-abi11g since the radiant and convection heat-transfer titten tmd to .complement each other. However, a i ? ~ i l l ~ r I ~ ~located t ~ t e r farther away from the furnace i.uiiat,inn behind a deeper waterscreen has a steam i r ~ ~ ~ l ~ n ~ characteristic ature which rises steeply with ~IIIIIQ@B~MO~ rating, due to the greater effect of convection 11111LIIN reduction in radiation heat-transfer rates. Navril boilers usually have waterscreens consisting of t l l i ; ~or :four rows of tubes and merchant marine boilers ,
FIRING RATE. PER CENT OF FULL OUTPUT
Fig. 23
de a relatively constant steam de range of rating. Superheaters. The superheater must deliver the ed ateam temperature during the operating life t during the initial trials or test cted performance must be mainvariations in firing d excess air. The necessity of unscheduled~oqtaiges rder to maintain performance.
Temperature characteristics of radiant and convection superheaten
these two factors and the surface. Increasing the temperature differential takes advantage of the available temperature potential, while an increase of the heattransfer coefficient necessitates a larger resistance to gas flow. Full advantage should be taken of a high temperature difference, but the entering gas temperature should not be so high as to result in excessive tube metal temperatures or high-temperature fuel ash corrosion (these are primarily a matter of location). The change in steam temperature with firing rate should be a minimum in order to prevent excessive temperatures during maneuvering and, again, this depends upon location. Steam velocities should provide for good distribution of steam, minimum tube metal temperatures, and acceptable steam pressure drops; all of which require correlating the effects of size, location, and the arrangeure dictates the thickness of the super- ment of the steam passes. which in turn is an important factor in the a. Types and Characteristics. The radiant and of superheater pressure drop and tube convection-type superheaters are the two basic types. They are, as their names imply, superheaters which receive heat by radiant or convection heat transfer and they may be arranged horizontaily or vertically. In the radiant type the steam temperature decreases with increased rating since the quantity of heat absorbed by radiation does not increase proportionally with steam flow; see Fig. -23. In the convection type, the steam temperature generally increases with increased rating are designed to have a because the heat absorption, due to greater heattransfer coefficients and higher inlet gas temperatures, urface can be obtained by increases a t a faster rate than the steam output. hcnt-transfer coefficient and the CemperaMost superheaters are a combination of the two basic oducts of combustion and types in which the designer builds in a radiant combsorbed is the product of ponent to achieve a flatter temperature characteristic.
II 1
MARINE ENGINEERING
106
(a) Three-pass hairpin loop type
BOILERS AND COMBUSTION
1 07
(b) Two-pass continuous loop type
~ i24~ Schematic . arrangement of hairpin and continuous-loop superheaters
GAS TEMPERATURE- F
,
108
MARINE ENGINEERING
Economizer elements (particularly the extended-surface type) are more expensive than boiler tubes. In air heaters, part of the advantage resulting from the improved temperature difference is offset by the high resistance to heat flow across the air flm [17]. Therefore, the proportions of component surfaces must be studied carefully to obtain the most economical overall arrangement. The minimum temperature of the feedwater to most merchant marine economizers vaxies between 270 and 280 F. The standard feedwater temperature for most naval installatior~is 246 F. This lower temperature is satisfactory because a premium fuel with a low sulfur content is used. Since the gas temperature leaving the economizer cannot be less than the inlet water temperature, .it follows that high feedwater temperatures limit the obtainable efficiency. Consequently, with high feedwater temperatures, economizers are not often used unless they are installed in conjunction with air heaters. I n an air heater, the minimum uptake gas temperature is dependent on the entering air temperature. Therefore, the attractiveness of air heater installations is due to the possibility of operating with a high boiler efficiency when using feedwater temperatures in the range of 300 to 450 F. When steam turbines are bled for regenerative feed heating,'the plant efficiency is increased about 1 percent for each 100 deg F rise in feed temperature due to the reduced heat loss in the condensers. Whether this improvement in efficiency warrants the expenditure required for additional feed heating and other equipment should be carefully weighed for each application. The use of an air heater necessitates an increased air pressure to the boiler unit because of the additional resistance to air flow through the air heater. Air pressures also must be increased when using economizers because of the relatively high resistance to the gas flow across the economizer, but, for boilers of the same size operating at comparable firing rates, an air heater installation will usually require a higher total air pressure than will a unit fitted with an economizer. w:1R rlg. IU. Air heaters are not pressure vessels, so the tubes can be If the uptake gas could be cooled to a temperature fabricated from mechanical tubing (less expensive than equal to the steam's saturation temperature by the use of pressure tubing) that is lightly expanded into the tube an infinite amount of heat absorbing surface, the improved efficiency would only be 83.75 percent. There- sheets. However, economizers are part of the pressure fore, air heaters or economizers must be installed to system and must be designed to withstand the main increase full-load efficiencies to the 88-90 percent range feed pump discharge pressure, to operate without leakage, usually desired. Further, the use of high evaporative and to withstand thermal shock. a. Air Heaters. Increased efficiency and reduced ratings a t any given steam pressure increases the need boiler maintenance can be obtained by improving comfor additional heat reclaiming equipment. When air heaters or economizers are installed, the bustion. Preheated air can improve combustion, reduco proportions of the boiler, air heater, and economizer boiler sooting, and reduce the possibility of ignition loss surfaces must be balanced. Usually, the temperature particularly at the extreme low end of the firing range. Practically all of the older marine air heaters were of differential between the products of combustion and the the tubular type; the regenerative types were not often heat absorbing fluids in the economizer and air heater is used. However, in recent years, particularly for highgreater than that in the last section of the boiler tube bank. This is advantageous in reducing the heat powered installations, the rotary regenerative air preheater has found wide application. A typical absorbing surface required for a given heat recovery.
Experience has shown that the diligent use of sootblowing equipment (particularly mass-action retractable units) usually can keep superheater surfaces satisfactorily clean for a year, or more, of opelation and that manual cleaning and washing of the external heat absorbing surfaces are required only during scheduled overhauls. h. Reheaters. The design of reheaters involves the same procedures and considerations that are pertinent to superheater design. However, the steam distribution and tube metal temperature problems are more critical since reheaters must be designed for exceptionally low steam pressure losses if a high cycle efficiency is to be obtained. Steam or combustion gas can be used as the heating medium in reheaters. When steam heating is used, the temperature of the reheated steam usually is limited to 550 to 600 F, since it is customary to use condensing rather than superheated steam as the heating medium because of the much higher rate of heat transfer. The use of gas reheaters is necessary if high reheat steam temperatures and cycle efficiencies are required. Such reheaters may be fired separately or installed in the boiler proper. Separately fired reheaters are not common because they require an individual firing aide and renewal clearances. as well as additional piping, -. controls, breechings, firing equipment, fans, etc. 3.6 Air Heaters and Economizers. Air heaters and/or economizers are necessary to obtain high boiler efficiencies. Preference alone should not arbitrarily influence the selection of either since the design of the power plant and it? performance characteristics greatly affect the choice. The temperature of saturated steam at a pressure of 850 psig is 528 F and the temperature of the products of combustion leaving the boiler tube bank would be, for a conservative boiler design, approximately 150 deg F above this value, or about 675 F. When firing oil, and operating with 14.0 percent COs in the products of combustion (approximately 15 percent excess air), this uptake gas temperature would result in an operating efficiency of only about 80 percent as can be see" from
BOILERS AND COMBUSTION
nxt~rnpleof a regenerative air heater is shown in Fig. 14. I l,n gastight casing forms part of the boiler forced-draft rir nnd uptake gas ducts. The heater is separately ~llourltedabove the boiler and suitable expansion joints mBoused in the ducts joining the two [7]. 'I'l~eessential component of the heater is the rotor in wllitill the heat-transfer plate elements are packed. The aila for combustion is passed axially through one side of Llln lutor while the flue gas is passed through the other aliltr in the opposite direction. As the rotor turns, heat I. nitltinuously transferred from the gas to the heating r~l~*lrbco; heat is also continuously given up to the air as the lioated plates traverse the air side. Counterflow I I tl~e ~ gas and air insures efficient heat transfer. 'I'l~n heat-transfer elements are made of corrugated and flnl alloets which are alternately packed in the main secI11111 of the heater and in the cold-end baskets. The coldel111basket is designed to be readily removable for cleanilly or replacement when conditions warrant. For daily elon~irig,a cleaning device consisting of a mass-action ar~ol~ blower is installed. Air and gas bypass dampers fiiw ctn integral part of the preheater and are useful in rature sion at imize soot ers can be made air heaters are of the horizontal type vertical type is no%often used since it is stall considerably more surface for a eat absorption than would be needed for the e, it is customary he gas across the In the vertical type the gas usually passes the tubes and the air crosses the tubes. ontal tubular air heaters generally utilize in-line &rrangements. These facilitate cleaning of the far more r heat transfer
isite tubes aximum the heatacross and y as about and, thus, ses with a s, both the tube size satisfactory, in most bes with *-in. tube de of the length of ow, the number of
109
tube rows, and the number of gas and air passes. This facilitates determination of the heat-transfer rates and the heating surface. The preliminary assumptions are then adjusted, if necessary, so that the surface arrangement and heat transfer provide the required heat absorption. Gas and air flow patterns also must be analyzed since maldistribution could reduce heat absorption, increase fan power, reduce or elevate tube metal temperatures, or restrict the capacity of the boiler unit. Air heater designs are usually predicated upon inlet air temperatures of 100 F, and exit air temperatures ranging from 300 to 450 F at the normal full-load operating rate. Design exit gas temperatures of 290 to 320 F are common for tubular air heaters and result in boiler efficiencies of 88.5 to 88 percent. Regenerative air heaters can be designed for lower uptake gas temperatures for a given risk of corrosion since for the same air and gas temperatures the heating surface metal temperature is somewhat higher than that of the tubular heater. Gas temperatures from 240 to 260 F are common for regenerative air heaters with boiler efficiencies of 90 to 89.5 percent respectively. Both the weight of the gas produced and the specific heat of the flue gas are greater than that of combustion air. Therefore, when firing oil with about 15 percent excess air, the reduction in the temperature of the products of combustion passing through the air heater is about 13 percent less than the rise in air temperature. In air heaters the heat-transfer coefficients across the gas and air films are of about the same magnitude, and high resistance to heat flow is encountered in the gas film on both sides of the tube. b. Economizers. Marine economizers can be grouped into two general classifications, the "bare-tube" and the Uextended-surface"types. They are generally nonsteaming and are usually arranged for counterflow of the water and the products of combustion. This results in larger temperature differentials, and greater heat absorption can be obtained. The counterflow nt permits a higher boiler efficiencybecause the temperature can approach that of the inlet omizers use tubes ranging in size from to 2 in. arranged in the form of either hairpins or continuous loops. The hairpin type consists of U-bend tubes that are welded, or expanded, into headers. Single or multiple rows of loops can be used as well as two or more headers. I n the continuous-loop type, each tube element consists of a length of tubing bent back and forth to form the desired number of rows; the ends of the tube are attached to the inlet and outlet headers, usually by welding. Since only two headers a& required, the number of tube joints is greatly reduced as may be noted from Fig. 26. There are many types of extended-surface economizers. The most prominent are those having steel studs or circumferential fins of aluminum, steel, or cast iron (see Fig. 12). Features common to all extended-surface
110
MARINE ENGINEERING
cient varies as the 0.65 to 0.70 power of the gas mass flow rate. Usually, if the economizer width is increased, a reduction can be made in the height of the economizer. Most marine economizers use counterflow arrangements with up-flow gas and down-flow water. The water pressure drop at about 25 percent of the normal full-load operating rate should be equal to, or greater than, the static water head in order to prevent recirculation. This minimum pressure drop requirement is not necessary if parallel-flow, up-flow gas and water, nonsteaming economizers are used, since the water pressure in the outlet header always will be less than that in the inlet Multiple water passes are often used in hairpin-type economizers to obtain satisfactory water velocities and pressure drops. These arrangements have both counterand parallel-flow relations between the water and the products of combustion, and the calculated heat transfer should be based on the average of the flow arrangements. Most continuous-loop and extended-surface type economizers have a single water pass arranged for flow counter
I
111
BOILERS AND COMBUSTION
bare-tube economizers the temperature drop across I ~ I I tube I ~ wall is small and, for all practical purposes, the Iltll)(j rmtal temperature can be considered the same ( J I I L ~ of 411'
the water it carries. Tube metal temperatures
extended-surface elements also are about the same as
Idltl ldjacent water, although the tip temperatures of the @xl,trrrdedsurface are considerably higher. I't'udence, and often regulations, requires a check valve 11) Illlo connecting piping between the economizer and the fltflfirndrum to prevent the loss of steam pressure in the ~Vnlltof an economizer casualty. Further, the valve fanilitates filling the economizer, particularly since a wnh@rhead of several feet is required to lift the check if k l i troonomizer ~ is located above the normal water level, h bypass line around the economizer will allow rrl8arrition of the boiler with an economizer outage. Ihbwover, few economizers are fitted with bypass lines k~aarlaeof their cost and the piping r ~ o m p l i c a t ini~~
Fig- 27
Drum-type desuperheater
BOILERS AND COMBUSTION
MARINE ENGINEERING
rlrr~rr sections and there is a definite transition zone I~ebwoonthe heated downcomers and the riser tubes, the 111t~alion of which varies considerably with changes in I IIH boiler firing rate. 111 the U-tube analogy, there is initially a vertical ja"mure plane a t the bottom on which the pressures ~ ~ n r l , oby d the hot and cold water legs are equal. As
NOZZLE
u(111111pressure plane in the lower water drums, or Ir@a(lurs,the pressure corresponding to the flow of water Ilrrough the downcomers is equal to the product of the I~oatlof water and its density minus the resistance to flow. 'I'lrk pressure must balance the product of the head of wnt1trrin the risers and its density plus the resistance to
+
N~TE IN ATYPICAL BOILER (SEE FIG 44) THE SIDEWALL AND ROOF CIRCUIT A,AND THE SCREEN AN0 FLOOR CIRCUIT B ARE SUPPLIED BY TWO DOWNCOMERS C. FURNACE FRONT AND REAR WALLS D AND GENERATING BANK E ARE SUPPLIED BY HEATED BANK DOWNCOMERS F. A HEAD TO WATER-STEAM MIXTURE F L W CURVE IS REQUIRED FOR EACH INDEPENDENT CIRCULATING SYSTEM. AND WOULD BE SIMILAR TO
TF
= '6
VENTURI-MIXING AND THERMALSLEEVE SECTION
Fig. 29
Fig. 28
Uncontrolled and controlled steam temperature
External-spray desuperheater
ture to the design value. The temperature of superheated steam is a function of rating and for the usual marine boiler rises as shown by the "uncontrolled curve" in Fig. 29. To make the most effective use of the materials in the superheater and main steam piping, the final steam temperature can be controlled so as not to exceed the design value. This can be accomplished by passing a portion of the superheated steam through a desuperheater in the drum. The location of the outlet and inlet connections is usually "interpass"; a typical arrangement is illustrated in Fig. 30. The desuperheated steam is returned to the last passes of the superheater where it mixes with the main flow to deliver the design temperature [2, 3, 61. A manually operated valve or an automatically controlled valve is used to regulate the temperature at all rates above the "control point9' (that point on the uncontrolled steam temperature characteristic curve which crosses the desired controlled temperature line). 3.8 Circulation and Steam Baffler. The natural circulation characteristics of the boiler and the type of steam drum b a a n g are determined after the arrangemerit of the heat absorbing surfaces has been established. Generally, because of the effect of the steam drum baffles upon the circulatory system, simultaneous analyses are made. circulation calculation procedures are in part empiricaland in part theoretical. The purpose of the is to establish a system of downcomers, riaers,
(a) Interpass, across restricted pass ~ g 30 .
[b) External bypass, three-way valve system
Interpass control desuperheater arrangements
and generating tubes which will insure that each tub0 receives an adequate supply of water in relation to the maximum heat absorbed. a. Circulation: Boiler Tube Banks and Furnace Waterwalls. The circulation characteristics of f u r n ~ o waterwalls and boiler tube banks are determined by tho same procedure and, since the water-steam ratio decreases with increased rating, the characteristics must be established for the maximum contemplated rating. In analyzing boiler cirqulation, it can be assumed that each circulating system is, in effect, a U-tube [6, 181. The riser section of the U-tube is that portion of the tubn bank in which the flow of steam and water is upward a* heat is applied. The downcomer section consists un" heated tubes or those ~ o r t i o n sof the tube banks ill which the heat absorption is considerably lower than "l the riser section. Because of the difference in fluid densities, heated tubes can act as downcomers for thf'
~ B @ Cdensities,
minus the riser friction 1oss-a quantity as the net available circulation head [3]. 111 most circulation analyses the steam geaerated in the rlmr tubes is calculated and the water-steam flow, as well r ~ t lthe net available head, is then determined for V L P ~ O ~ water-steam ~R ratios. In analyzing circulatory @hrrl.noteristics,it is customary to graphically plot both llro downcomer friction losses and the net available sirolllntion heads for the assumed water-steam mixture %ewu, As shown by Fig. 31, the flow tit which the ~ltr~dlt~ble head minus the resistance to flow through the ~ C e ~baffles nl equals the resistance to downcomer flow is that required to balance the circulatory system. From F ~ flows P at the balance point the percentages of steam by wlnnlo at the top of the riser tubes can be calculated. Tho percentage of steam by volume a t the top of the @@el'llubes must be such as to preclude overheating of ih@tlibes. If the quantity is excessive, the circulatory 6YPb111 must be redesigned to provide additional downkmflrrr, or the size and contour of the downcomers must )I atlonged to reduce the resistance to flow. It also CW&y ho necessary to change the location, size, and &llllt~llr of the boiler tubes to redistribute the heat ~Brrerptionand reduce flow resistance. f 11 a satisfactory circulatory system, an adequate @moullCof water must be supplied for each pound of ~ C ~ a gonerated. nl Therefore, if the percentage of steam b,Y vol~lmeat the exit of the riser tubes is used as a design @rlk@rlori, it is necessary to vary the allowable percentage MI @Irapressure changes since the percentage of steam by vtllulno will increase as the pressure is reduced because @f tlla irlcreased specific volume of the steam. Naval Btrllera nro usually designed for water-steam ratios (i.e., k@l#irll of water/weight of steam passing through the #elrornLiag tubes) ranging between 5.0 and 10.0, and hlel'bll~ttltunits usually fall in the range of 15.0 to 20.0 IC blra overload rates of operation. Lower water-steam p ~ b l onro ~ used on naval boilers in order to reduce the Crjlltlr ~ i a o and weight by minimizing downcomer bt~trwtl
I
Fie. 31
Characteristic head venus water-steam mixture flow for circulation calculations
b. Heated Downcomers. If evaporative ratings are conservative and the gas temperatures leaving the boiler do not exceed about 750 F, the first several rows of tubes will function as risen with the remainder serving as heated downcorners. As the firing rate increases, the high-temperature zone moves farther back into the tube bank and additional tubes become risers while a corresponding lesser number act as downcorners. If the firing rate is further increased, the number of downcomers becomes inadequate, circulation is impeded, and tube casualties may occur; when design analyses indicate such circumstances, external or unheated internal downcomers must be installed. c. External and Unheated Internal Downcorners. With conservative evaporative ratings, external downcomers 'are required for only those portions of the boiler in which the tubes cannot act as downcomers (i.e., a single tube row forming a furnace boundary, a shallow tube bank installed between two furnaces, or tube banks shielding a superheater from two furnaces). If downcomers are required for the main tube bank, they usually are located external to the tube bank even though the arrangement requires longer boiler drums. The use of unheated internal downcomers minimizes the drum length and eliminates tubes in the main boiler bank; however, unheated internal downcomers usually enter the steam drum at high water levels and they may lose water during heavy rolls or inadvertent reductions
STEAM OUTLET
upon the natural separation of steam and water. For higher boiler ratings a positive means of steam separation is required and compartmenbtype baffles, Fig. 32(b), are frequently used. Centrifugal steam separators are used primarily in highly rated merchant and Navy drum-type boilers; they are particularly desirable for boilers subjected to rapid maneuvering, fluctuating water levels, or high solids concentrations in the boiler water. Centrifugal steam separators may be arranged either horizontally or vertically in the steam drum as in Fig. 3 2 ( ~ ) . The resistance to flow through centrifugal separators is
(a) TRIPLE PERFORATED PLATE BAFFLE
115
BOILERS AND COMBUSTION
MARINE ENGINEERING
llloat merchant units having 48 to 54-in. drums and most llnval boilers using 46 to 60-in. drums. As power levels Ill(:rease,60 to 72-in.-dia drums are used more frequently 101) provide the necessary room for steam baffles and to lw()videthe capability of accommodating the shrink and n w d l that occurs when maneuvering. b- Headers. Headers for water walls or economizers ~ 1 . ousually fabricated from pipe stock. &llow forgings 1lltU' also be used especially for superheaters. They may b~round or forged to a rectangular or other cross @fl(ltion to facilitate tube installation. Tubes are lll~lulledby expanding or by welding. Htandard boiler and economizer tubes are fabricated from either electric resistance welded or seamless stock. t1:lo0tric resistance welded tubes are less expensive and lllbvo been proven to be as dependable as seamless tubes 111 boilers and economizers. Superheater tubes are made
ULATING FIREBRICK INSULATING BLOCK ASBESTOS CEMENT
!TEAM OUTLET
supplied to the downcomers is greater than that of the "frothy" water-steam mixture discharged from perforated-plate and compartment-type baffles. e. Effect of Drum B d e s on the Circulatory System. The steam-water flow through the steam drum baffles is in series with all of the flow circuits in the circulatory system. Thus, if the flow through one of the circuits is increased, for example, by the installation of additional downcorners, the flow through the steam baffles also is increased. This imposes an additional resistance in the circulatory system with the result that the flow in downcomers will not increase in direct propodion the additions made.
emperature to 130 F or less. Local areas, for where superheater inlet or outlet nozl;les
( C) CENTRIFUGAL SEPARATORS
FIBERGLASS
ng. 32 Typlcal steam reparation equipment
st boilers are of double-casing comtruction. An
boilers, combustion gases are discharged into ry space in the event of a leak. ( 0 ) BANK AND UPTAKES
materials vary. to suit the application; those for a particular unit can be readily determined. ral or strength members of the casing are used art some of the loads of the pressure parts. The ing bank and screen and furnace walls are eolf-supporting; however, the casing may lend these pads during rolling and pitching of the it is not on an even keel- It is U S U ~ 00 h o s u ~ ~ o r t tsuperheater he headers and the super' t'lbfis (wholly or in pad) , as well as the economizer l'lrlbtcr Or tubular), on the casing structLlre.
'
Fig. 33
Typical sections of boiler casings
Suitable access and inspection doors are required and their location is an important practical aspect of casing design. Provisions must also be made for differential ~expansion between the pressure parts and the casing and between the casing and the boiler foundation and surrounding decks, platforms, piping, etc. In large boilers where welded walls are used, another
>
"
,
>
rF
%
'h
T L
.*.t. \ *
MARINE ENGINEERING
BOILERS AND COMBUSTION
117
MARINE ENGINEERING
BOILERS AND COMBUSTION
MARINE ENGINEERING
give the operator a direct view of a light source which shines through the boiler uptake and the combustion gases. Another type employs a photoelectric cell and provides a readout on a meter scale calibrated in smoke density units; it may also be fitted to sound an alarm when a certain smoke density is reached. d. Instrumentation and Controls. The need for operating instruments and manual and/or automatic controls varies with the size and type of equipment, the method of firing, the proficiency of the operating personnel, and the desired degree of automation. Chapter 21 covers the application of control equipment to ship's propulsion plants. For safe operation and efficient performance, information is required relative to the water level in the boiler drum; burner performance; pressures of the steam and the feedwater; temperature of the superheated (md reheated) steam; pressures of the gas and air entering and the leaving principal components; feedwater and boiler water chemical conditions and particle carry-over; operation of feed pumps, fans, fuel burning, equipment; relationship of the and fuel actual combustion air passing through the furnace to that theoretically required for the fuel fired; temperatures of the water, gas, fuel, and air entering and leaving the principal component parts of the unit; and feedwater, steam, fuel, and air flows. Icor many years, marine boilers have been equipped with control equipment permitting steady operation at sea with little operator participation except while maneuvering. However, the trend is toward complete automation of the boilers so that, with the exception of starting up, they can be operated throughout the full range from standby to full load without manual adjustTo attain fully automatic operation, the development of adequate control components and system designs is essential. The operating characteristics of the principal and auxiliary items of steam-generating equipment must be fully known since these characteristics affect the degree of controllability, the scope of the controls required, and the response obtained. These in turn affect the safety of the installation and establish its economic justification. As an example, where the burners have a range of oper~tionor turndown capability equal to or greater than that required by the boiler, the necessity to sequence burners (or take them out of service) is eliminated. This, in turn, eliminates many decisions and functions that would otherwise be required of an automatic burner management system, and a simpler system may be selected. The degrees of control which can be achieved, in ascending order of sophistication, are manual, local supervised manual, remote supervised manual, automatic (nonrecycling), and automatic (recycling). These various types of control can best be delineated by relating their functions to burner operation. With the manual type of control, Fig. 40, a burner is manually purged and ignited. I t may be automatically
modulated but it is stopped manually. Although no operator function is ~erformed automatically, widerange burners can be used with automatic comb us ti or^ controls to facilitate dock-to-dock operation without manual participation. However, without boiler and burner monitoring devices, the operator must remain in close proximity to the boiler to provide the necessary surveillance. In the local supervised manual system, Fig. 41, a burner is purged and manually ignited, but certain ~ r o c e d u r e ~ and conditions are supervised by safety interlocks. ~ l l manual functions are performed and checked by tho operator a t the burner station during normal operation, and if the demand for steam is within the capability of the burners, unattended boiler operation is attained. Monitoring and safety interlocks are ~rovidedto alter the operation if an unsafe condition develops, and to trip the burner and/or the boiler, if necessary. After 11 trip-out, the operator must take the necessary correctivn action to clear the interlocks and recycle the burner and/or the boiler. The remote supervised manual system, Fig. 42, allow^ a burner to be purged and ignited by a ~ushbuttonor selector switch, modulated automatically, and securcd by a remote manually actuated pushbutton or selector switch. I t also provides supervision of procedures by safety interlocks. The burner is mechanized and all operating functioris are ~erformedby mechanical device^ initiated from a remote control station which indicaton whether or not each function has been performed correctly. This system of control does not relieve thtr operator of burner manipulation. He must devote hin undivided attention to the step-by-step procedures folstarting and securing burners, which is a time-consuminlr, process. This control system can only be justified i l l installations where the turndown capabilities of tho burners do not match the turndown requirements of tho boiler, and, the burners must be manipulated to covclr. the operating range. Its application will not meet thtr USCG requirements for an automatic boiler. The automatic (nonrecycling) control system, Fig. 43, involves a burner which, when actuated manually by 11 pushbutton, is purged, ignited, and modulated automatically; and although secured either automatically or remote-manually, the burner does not recycle automatically. When start and stop sequences are manually initiated from a remote control station, each function i l l the start-up and stop sequence is performed and checltatl automatically and all ~roceduresand conditions arc' supervised by safety interlocks. Since the operator may be required to initiate the start-up and securing of n burner to meet load requirements, this control systenl does not meet the USCG definition of an automatic! boiler. With an automatic (recycling) type of control systenr, a burner is purged, ignited, modulated, and stopp(-tl automatically, and the burner recycles within a prescribnd load range. 3.12 Sample Design Problem. The steps followcxl
BOILERS AND COMBUSTION
Fig. 40
121
Burner operation-local
manual control
Hlltrthm pressure, drum, approx.. . . . . . Hll(rfbmpressure, superheater outlet. . .
nuperheated 185,520 Ib/hr (losuperheated 16,870 lb/hr Pntdwater temperature. . . . . . . . . . . . . 41:Hiciency (based on 13% radiation auld unaccounted for losses and 15% flxcess air) . . . . . . . . .'. . . . . . . . . . . . . P'ud total heating value (standard h n k e r C 4- added heat in air). . . . 19,264 ~ t u / l b Pll(!l required. . . . . . . . . . . . . . . . . . . . . 14,349 lb/hr Alr temperature, leaving steam air houter . . . . . . . . . . . . . . . . . . . . . . . . . 811'flow (16.07 Ib/lb oil at 15% excess
nir) . . . . . . . . . . . . . . . . . . , , . . . . . . . . Ylue gas flow = 244,937, say. . . . . . . .
245,000 lb/hr
fpol' the example, only one rate of operation will be lalsul~btedalthough for an actual boiler design it is not ilfitl@llfil to calculate three or more rates to establish @l*kl'fi()~Ori~ti~ Curves of performance. Rated power will orl(lulated since this establishes the design meeting khr~ np(>eifiedefficiency and steam temperature. The h m l - ~ ~ s fdata e r are derived from the cumes and pro@@tlrlr.ao of Chapter 2. Boiler Layout. Two oil burners will be used to @MPP~,Ythe total oil flow of 14,349 lb/hr at rated power Mia ahout 8000 lb/hr each at overload. The necessary @!@#r~lces for burners of this capacity are obtained from the ~(rlocted burner manufacturer. Based on this !ltfl1rllll~tionand experience, an approximate furnace and bll@l' l h ~ o u is t prepared (see Fig. 44) from which the hlfllfitf(fvolume and heating surfaces can be estimated. k~. Furnace Calculations. The furnace volume, cold
*
@ul.f#fltr,and
radiant heat abs~rbingsurface @HAS) are
Fig. 41
Burner operatiolr-local rupewhed manual conko]
MARINE ENGINEERING
BOILERS AND COMBUSTION
To determine the shape emissivity factor, FBFA,the following data are required: VF = 2655 fta ST = 1200 fta S, = 1175 ft' PF = 1 atmosphere
X. = tube equivalent thickness =
k
Thotefore the firing density is [see equation (31) of
Fig. 7 of Chapter 2 the concentration factor -K la 0,086. The mean radiating length is L = 0.6q2655 = ft. Equation (31) of Chapter 2 can now be evaluated &odotermine the flame emissivity
123
=21 Dolog, O-D Di
2 -22 log. = 0.182 1.67
= tube conductivity = 310 Btu/hr-ft-F
Next, by estimating the corrected furnace exit ternperature, T E ~to, be 2200 F the tube film temperature can be approlrimated as
R I I ~from
EF =
~
~
~of furnace ~ exit~ temperature i ~ and furnace ~ tabrorpriar i
~
for an &/ST value of 0.98, FEFAis determined to be bSd4 from Fig. 6 of Chapter 2. In order for the calculation to proceed, it necessary w u m e several values of the furnace exit temperature. bibking this assumption
* ture and heat absorption can be calculated (See Sectio" 2 of Chapter 2)) based on the following furnace surfacO
The temperature coefficientf~ [equation (40) of Chapter 21 then becomes
f~ = 0.00003875T,r
0.95(1 - e- (o.o6a)(i)(s.a) ) = 0.353
FURNACE TEMPERATURE TE,
Ag. 45
= 1653 F
+ 0.1035 = 0.1675
With a flue gas flow W nof 245,000 lb/hr and two burners having 2-ft thmat diameters, the flue gas weight flow rate
G is 39,000 1b/ft2-hr. Since the furnace depth D is 14 ft, the surface heat-transfer coefficient hRw can be computed from equation (39) of Chapter as ~0.6a
,/&)a; equation (36) of Ohapter 2; Btu/ft2-hr
2,200 2,660
2,300 2,760
2,400 2,860
95,100
88,000
81,300
haw =
f r = 13.2 ~tu/hr-ft2-F
e246,000~o.ai4 TB~ = 2239 F
The adiabatic sensible heat in the combustion can be computed from equation (37) of chapter a fuel lower heating value of 17,500 Btu/lb and a fuOl Ensible heat of 46 Btu/lb (100 deg F rise at 0.46 heat), for pedect combustion the sensible heat bemmo* determined by the methods of Section 2 of Chapter T & R Bulletin 3-14 [lo] to be: Furnace volume = 2655 fta Projected surface = 1200 fta RHAS = 1175fta
Or
With a fuel higher rating value of 18,500 B t u m the furnace ratings at rated power are: Release rate =
LHV
QTAI
-
+ q~ 4- (ta - ~o)CPR
17,500
+ 46 + (278 - 80) (0.2445) (13.98) 13.98 4- 1
TUBE TRANSGAB VEEBE BACK FLOW
DIAM-
mture in the furnace is From Fig. 2 of chapter 2 the adiabatic flame tempertLture, TAt,is found to be 3990 F or 4450 RWith 15 percent excess air
3.5
2.5
100
850
\
641 F at drum saturation pressure of 975 psia
The screen, superheater, and generating bank performance calculations may be conducted as follows:
126
MARINE ENGINEERING
This practice is expected to becO1'lr' used after proper treatment (19, 201. In essence, this from corrosion. common, particularly at higher steam Pressures an(( entails: the removal from the raw water of those con1 stituents which are known to be harmful; supplementary single-boiler installations. Filming amines introdl~(~~sl treatment (within the boiler or connected system) of into the feedwater or steam lines also provide ~rotec1,l~)ll impurities to convert them into harmless forms; against corrosion, but by forming a coating on the mrlftll and systematic removal, by blowdown of boiler water surfaces rather than by changing the PH of the watts. of 4.3 Boiler Water. Boiler water is treated within I ~ I I I . concentrates, to prevent excessive boiler to prevent corrosion, the fouling of heat-absorblll# solids within the unit. surfaces, and the mntamination of steam. T h i S r e q ~ i ~ " ~ ~ The ultimate purpose of feedwater and boiler water the injection of chemicals into the steam drum W I I ( ~ I ~ treatment is to keep the internal surfaces free from deposits of scale or sludge and to prevent the corrosion they react with the residual impurities in the feedwi~ln-I of these surfaces. Hard-scale formations, formed by Properly controlled, internal treatment can mai111.nlll certain constituents in zones of high heat input, retard boiler water conditions within satisfactory limits [6, 1x1 an al1c:~li111~ Corrosion is minimized by maintaini~l~ the flow of heat rnd raise the metal to higher-than111 boiler water and this condition is usually expressed temperatures. This can cause overheating and The PH of w;lt1'l '
BOllERS AND COMBUSTION I I I * ~ I I l~lr~ny factors, and a feedwater specialist should be p~tlrwtll~crtlto establish specific procedures. However, iltr* I.P@(II~R obtained will depend upon the diligence and ttrtygrl(~y of the routine sampling and the control measures B + # I I l11trl~od ~ by the operating personnel.
rmal wn; tiny operation; inspection and maintenance; and rlb8o. In all phases the handling of the equipment or, but the overall tions and operation
stic
127
govern the time required for start-up and also, to some extent, for cooling after shutdown. c. Boiler Cleaning. For mtisfactory and efficient operation, a boiler must be kept clean on both the waterside and fieside. With adequate attention to the and by maintaining the boiler prescribed limits, there the waterside. The fireside, on the other hand, requires daily attention if the steam temperature and boiler efficiency are to be maintained a t their optimum values. Only distilled and deaerated water should be used for feeding the boiler and for feed makeup. Total solids in the boiler water should not exceed a of 500 Suspended solids should five percent of the total lower than 2 ppm and range of 10 to 25 ppm. The o d d be in range of 10.2 to 11.5. r (sodium sulfite) should range maintained within these limits, will not form scale or
a steaming boiler should be given a good surface blow each to maintain about 50 percent of the normal day. A test for total dissolved solids made before and ~bhrgpressure. This procedure facilitates the desired after the blow will indicate if additional attention is g. required. The water drum bottom blowoff connection frrCa
ing suspended or total solids in the
ce to reduce
shutdown, nspect the of normal
eheater) flow. POS-
allowed to deteriorate to the point bakedqn sludges are found during waterside inspections, chemical analysis of the deposits will indicate the cleaning method best suited for their s and scale cutters through each tube, and a high-pressure water hose. The entire eaned as a unit more quickly and efficiently ng. A specialist should be consulted to procedure, which entails the use of acid rinsing agents. The acid strength, the temperature at which they are used . are of vital importance if the cleaning process is to be kept within safe limits. Excessive acid strength or unneutralizpd acid remaining after cleaning will pit and attack the metal possibly to the point that replacement to facilitate the cleaning surfaces of the superheater well as the economizer and air anged in in-line patterns which rough which inspection and cleaning hed. Staggered patterns are slightly more efficient from a heat-transfer standpoint but are more difficult to inspect and clean. In extreme conditions, hand lancing or watermu.ashing
MARINE ENGINEERING
boiler and its cleaning equipment have all but eliminated the need for hand cleaning. Soot blowers are used to clean the fireside at regular intervals. The frequency depends on the fuel ash characteristics, combustion efficiency, and the rates of operation. Air or steam can be used as the blowing medium; however, oil-fired boilers almost universally use steam. Steam is available in large quantities and at a low cost. Air, often used in coal-fired units, is "puffed" intermittently to permit repressurization of the air receivers by the air compressor. Superheated or desuperheated steam can be used with good effect. The steam should be supplied in a dry state, and the supply system must have adequate traps or be fitted with orificed drains to remove condensate so as to prevent it from reaching the blower elements. Three basic types of steam soot blowers are used. The long retracting mass-action type used in superheaters; the rotary valve-in head line blower used in boiler banks, economizers, and tubular air heaters; and the stctionary-type unit which is used in hoppers and where fixed directions of blowing are desired to remove localized deposits, such as those forming on top of the water drum. The soot blowing system can be manually operated or sequential pushbutton controls can be employed to automatically program the cleaning process. Once initiated, the automatic sequencing control opens the steam supply valve, warms the lines, blows the soot blowers in sequence, and then shuts dourn the steam supply. 4.5
Boiler Storage
a. Dry Storage. When a boiler will be idle for a considerable length of time and there will be ample time available to prepare for its return to sewice, the drystorage method is recommended. To accomplish this, the unit is emptied, thoroughly cleaned internally and externally, dried, and then closed tightly to exclude both moisture and air. Trays of lime, silica gel, or other moisture absorbents, are placed in the drums to collect the moisture trapped in the air when closing the boiler. To insure against a possible overflow of corrosive liquid after the moisture has been absorbed, not more than 75 percent of the tray capacity should be filled with the dry absorbent. Care must be taken to prevent water, steam or air leakage into the unit, and periodic inspections should be made to make sure that there is no corrosive action' The absorbent should be replenished as required. b. Wet Storage. If boilers are to be placed in standby service but must be available for immediate operation, before shutting d ~ w nthey should be steamed to stabilize the boiler water conditions and to remove oxygen bubbles from the internal surfaces. The boiler firing rate should then be decreased slowly and the steam drum water level should be raised as high in the gage glass as is consistent with safe operation while still passing steam to the line. The hydrate alkalinity in the
boiler water should be increased to a minimum of 400 ppm, and, with the addition of sodium sulfite in thv amount of 100 ppm, oxygen corrosion can be prevented. During storage, boiler connections should be checkctl for leakage and frequent samples of boiler water shoultl be taken and analyzed. If analyses indicate that t h hydrate alkalinity is less than 250 ppm, the water in thth steam drum should be lowered to the normal operatirig level and chemicals should be injected to bring tlic~ hydrate alkalinity back to 400 ppm. The boiler shoultl then be steamed sufficiently to circulate the addctl chemicals, following which the process of wet storago should be completed in the usual manner. c. Steam Blanket. The steam blanket method provides excelle~lt protection for short-time idle storagct, but requires a continuous source of low-pressure steal11 (in order of 150 psig) and connections for maintainirr~ this steam pressure in the stored boiler. All vents arrtl drains should normally be closed to allow the boiler arrtl superheater to fill with condensate but the boiler can ht, drained periodically if desired. d. Nitrogen Blanket. The oxygen-free nitrogc!~~ storage method is one in which nitrogen gas at a presstlro of 10 to 15 psig is maintained in the unit at all tirntbn during its idle status. It can be used with very satinfactory results if the boiler, terminal valves, and fittirlga are tight under normal hydrostatic pressure. The boiler can either be emptied or a normal wahr level maintained in the steam drum. The nitroger1 i# admitted when the boiler pressure has dropped below the gas pressure which will be maintained in the u~ril,, Satisfactory protection against corrosion depends uport system checks and the renewal of nitrogen, as necessary. To ready a boiler for sewice after storage, the nitrogtl~~ supply is secured and the water level in the steam dr~llri is raised to that required for lighting-off. Any nitrogt~ll in the steam drum and superheater will be displaced I)y the steam generated during the customary venting 01' the steam drum and the superheater as steam pressure is increased. References
1 George W. Melville, "Development of the MnririiBoiler in the Last Quarter Century," The Engineer, 1!)1 I 2 G. W. Kessler, "Procedures and Influeriairtg Factors in the Design of Marine Boilers," Z ' T ( L ~ I ~ SNAME, vol. 56, 1948. 3 W. I. Signell, "Marine Boiler Design Tod:~y," Trans. SNAME, vol. 76, 1968. 4 Carl D. Shields, "Boilers, Types, Characterisbieo and Functions," F. W. Dodge Corp., 1961. 5 W. A. Fritz, Jr., and L. Cohen, "Development r u ~ t l Evaluation of a Supercharged Steam Generafi,irl~ System," ASME Paper 62-WA-279, November 196%. 6 Steam, Its Generation and Use, The Babcock r ~ i i t l Wilcox Co. 7 F. P. Bergonzi, G. Cooper, and J . F. Moorny, "Heat Recovery Equipment for Modern Marine Cyeltm," SNAME Southern California Section, November lNitl.
BOILERS AND COMBUSTION
H "Lexicon-Steam Generating Equipment, " AmerIjoiler Manufacturers Association, Newark, N. J. i t "Marine Steam Power Plant State of the Art riec~~it~ar," General Electric and Babcock & Wilcox, 1969. 10 "Boiler Furnace Performance Criteria," SNAME 'I'd It Bulletin No. 3-14, December 1963. 1 I Code of Federal Regulations, Title 46-Shippiw, p~~ldinhed by the Office of the Federal Register. IS W. 0, Nichols, M. L. Rubin, and R. V. Danielson. "Homo Aspects of Large Tanker Design," Trans. IYNA ME, vol. 68, 1960. 13 "Recommended Practices for Preparing Marine WI~rrrn Power Plant Heat Balances," SNAME T&R ll\lllotin No. 3-11. 14 R. P. Giblon, K. M. Shauer, and I. H. kolih, "L jnnign Considerations for Boiler Forced-Draft Systems, " lllatb
Marine Technology, vol. 6, no. 4, October 1969, p. 406. 15 J. J. Banker and M. G. O'Harra, "Some Considerations for Automation in Marine Boiler Desim, Combustion Equipment, Boiler Control and Burner Control, " SNAME, Great Lakes Section, January 1966. 16 "Section I ASME Boiler and Pressure Vessel Code-Power Boilers," the American Society of Mechanical Engineers. 17 William H. McAdarns, Heat Transmission, McGraw-Hill Book Co., Inc., 1954. 18 G. R. Fryling, Combustion Engineering, Combustion Engineering Inc., New York. 19 Eskel Nordell, Water Treatment for Industrial and Other Uses, Reinhold, 1961. 20 "Betz Handbook of Industrial Water Conditioning, " Befi Laboratories, Philadelphia, 1962.
NUCLEAR MARINE PROPULSION
CHAPTER I V
Sectio~~ 1 Basic Fur~damentals The objective of this chapter is to present a survey of the subject of nuclear marine propulsion that is directed towards persons having engineering backgrounds but no experience regarding nuclear reactors. For more detailed treatments on the subject of reactor engineering, including such highly specialized considerations as fuel dwign, reactor design, coolant chemistry, nuclear instrumentation, and reactor safety, reference is made to the comprehensive textbooks available on the subject [I-71.' I n order to avoid security classification problems and problems with 15 CFR Part 385, U. S. Export Regulations, p.rimary emphasis is placed on projected nuclear propuls~onapplications for commercial merchant marine ships as contrasted to naval propulsion applications. A discussion of the differences in these requirements is included in a paper prebented to the Society by ADM H. G . Rickover, et al. [81. 1.2 Introduction. The idea of utilizing the atom as a possible source of energy was first introduced around 1900 when i t was discovered that certain atoms could spontaneously discharge charged ions capable of effecting emulsion. This discovery of radioactivity was discussed in 1902 by Pierre and Marie Curie. Later, Einstein provided an explanation of the energy of radioactivity and of atomic energy in terms of the equivalence of mass and energy. I n 1939, the real possibility of converting mass into energy was demonst~atedfirst by the discovery of nuclear fission and later, in March 1939, by consideration of the possibility of a chain reaction. I n a chain-reaction process, sufficient neutrons are produced to provide for all system losses plus sufficient neutrons to maintain the reaction rate of second-generation fissions, each of which produces sufficient neutrons to continue to maintain the reaction rate. By 1941, sufficient knowledge had been accumulated to permit preliminary experiments on subcritical assemblies, and on December 2, 1942, the historical Chicago Stagg Field chain-reacting pile went critical. The most significant characteristic of nuclear power for maritime application is the compact nature of the energy source which has obvious advantages for many types of 1.1
Scope.
'Numbers in brackets designate References at end of chapter.
he power ~roducedby fission of mobile power plants. one gram of uranium per day is equivalent to about olw megawatt. I n other terms, the fission of one pound uranium is equivalent to the combustion of about 900 tons of 18,500 Btu/lb fuel oil. here fore, nuclcss power permits the utilization of very large power pla1ll1~ on board ship without the necessity for very lar~ll bunker storage or frequent refueling. ~ c o n o m i cstudicr indicate that the cost ~enaltiesassociated with nuclct~s power are sufficiently high that further innovations wil l be required before nuclear power for ship propulsioll will be able to economically compete with fossil-fueled power systems; therefore nuclear power is attractive only where the advantages of high power and enduranc;tl override purely economic considerations, as is often th" case for military purposes. With technological changrn in methods of shipping which would emphasize significantly improved efficiency through total system integrrhtion, it is expected that higher utilization and largola propulsion power requirements and revenue generatio~l would tend to make nuclear propulsion more attractivtr. I n general, the primary engineering ~ m b l e mfor nuclolu reactors is to provide under all circumstances for t l l ~ removal of heat from the nuclear fuel. The energy stored in nuclear fuel is in an extremely compact forni, and it has the theoretical potential to release its totnl contained energy in a short time. The high pow(:l. density potential results in the necessity to provitltq efficient, highly reliable, and sometimes unusual hen.1.transfer systems not only for the steady-state p o w ~ ~ operation and normal heat removal after shutdown, blr~ also for all emergency and accident conditions. A major difference between nuclear propulsion a~ltl foaril-fueled propulsion systems is the safety aspect 01 the nuclear reactor system; safety is a major considenltion with nuclear reactors due to the emission of radi~rtion, consisting primarily of neutrons and beta and gamma radiations, from the fission products. Further more, the fission product radiation must be considerotl for a long time after the reactor is shut down or the spcnl~ fuel elements are discharged. Operating personnel mu~l, be protected from the radiation by suitable shielding; the shielding may consist of lead, water, steel, concrettr, and other radiation-absorbing materials which in totll~l are equivalent to six or more feet of concrete. ~ h i n
131
~liitrl(1adds considerably to the size and weight of the
minimum energy required to dissociate it into its component
1~1111~rice and the provision of a sufficiently high level of mlinbility to ensure a long service life must be considered rial only in the formulation of the basic concept of the I J but~ ~ in ~t h e ~procurement ~ ~ of equipment and ci~~rtryjonents.Strict adherence to codes and standards a~rt1 aompliance with rigorous quality assurance programs rluri~kgconstruction are the means used to ensure plants
Biological Dose. The radiation dose absorbed in biological it is measuRd in rems, Biologiul The time required for a biological system such as man or an anim$ to eliminate, by natural processes, half the amount of a substance (such as radioactive materid) that has entered it. Biological Shield. A mass of absorbing material placed around a reactor or radioactive source to reduce the radiation to a level that is safe for humans.
~aa~otors.Concrete and steel absorb gamma rays and 111?11trons in reactor shields. A sheet of paper will absorb rls lhttenuate alpha particles and a thin sheet of metal will u b o all ~ except the most energetic beta particles. Absorption. The process by which the number of particles or photons entering a body of matter is reduced by interatrllion of the particles or radiation with the matter; ui~nilarl~, the reduction of the energy of a particle while bsr~versinga body of matter. This term is sometimes ~~~roneously used for capture. h~tlvation. The process of making a material radioactive by Immbardment with neutrons, protons, or other nuclear
&ha Particle. A positively charged particle emitted by crarlain radioactive materials. I t is made up of two rla~ltrons an&woprotons bound together,hence it is identical wihh the nucleus of a helium atom. I t is the least penetrat11lg of the three common types of radiation (alpha, beta, gemma) emitted by radioactive material. I t is not tlnrr~erousto plants, animals, or man unless the alphaemitting substance has entered the body. baekgromd Radiation. The radiation in man's natural
is the breeding ratio minus one. Burnable Poison. A neutron absorber (or poison), such as boron, which when purposely incorporated in the fuel or fuel cladding of a nuclear reactor "burns up" (is changed into nonabsorbing materid) gradudly under neutron irradiation. This process compensates for the loss of reactivity that occurs as fuel is consumed and fissionproduct poisons accumulate, and keeps the overdl characteristics of the reactor nearly constant during its use. C a p e . A process ih which an atomic or nuclear system acquires an additional particle; for example, the capture of electrons by positive ions, or capture of electrons or neutrons
absorbs a neutron and fissions, releasing additional neutrons. These ih turn can be absorbed by other fissionable nuclei, releasing still more neutrons. A fission chain reaction is self-sustaining when the number of neutrons released in a given time equals or exceeds the number of neutrons lost by absorption in nonfissioning material or by escape from the system. Cheinical Shim. Chemicals, such as boric acid, which are placed in a reactor coolant to control the reactor by absorbing neutrons.
132
prevents the neutrons from causing neutrons, a control further fission. of a nuclar reactor containing the core. The centn] but not the fuel elements and usually the
Cdtid M ~ ~h~ ~ ~ . mass of fissionable material that will supporta self-sustaining chain reaction under stated conditions. Cross Section. A measure of the probability that a nuclear reaction will occur. usually measured in barns, it is the area presented by a target nucleus apparent(or (or to an or other nuclear radiation, such as a photon of gamma radiation. Cluie. ~h~ basic unit to describe the intensity of radioactivity in a sample of material. The curie is equal to 37 billion disintegrations per second, which is approximately the rate of decay of 1 gram of radium. A curie is also a quantity of any nuclide having 1 curie of radioactivity. Named for Marie and Pierre Curie, who discovered radium
D e w Heat The heat produced by the decay of radioactive fission ~~ h ~ ~Neutrons ~ ~ emitted d t by radioactive ~ ~ products in a reactor over a period of seconds or minutes after a fission takes place. Fewer than 1 percent of the neutrons are delayed, the majority being prompt neutrons. Delayed neutrons are important considerations in reactor design and control. Depleted U r ~ u m . Uranium having a smaller percentage of uranium-235 than the 0.7 percent found in natural uranium. ~t is obtained from the spent (used) fuel elements or as by-product tails, or residues, of uranium isotope separation. ~ ~ ~ ~ isotope ~ off hydrogen i ~ whose . nucleus contains is therefore about twice as one neutron and one proton heavy as the nucleus of normal hydrogen, which is only a Deuterium is often referred to as heavy single hydrogen; it occurs in nature as 1 atom to 6500 atoms of ndrmal hydrogen. I t is nonradioactive. D h t - c Y d e ~~~~t~~ Syptem- A nuclear power plant systemin which the coolant or heat-transfer fluid circulates first through the reactor and then directly to a turbine. Doppler Effect. The shift with temperature of the interaction rate between neutrons and reactor materials, such fertile materials. as fuel rods, structural materials, The shift can appreciably affect the neutron density and hence the reactivity of reactors. D~~~ Rate. The radiation dose delivered per unit time and measured, for instance, in rems per hour. xleCtmn volt, ~h~ amountof Lioetic energy gained by an through an electrical potenelectron when it is tial difference of 1 volt. ~t is equivalent to 1.603 x 10-11 erg. ~t is a unit of energy, or work, not of voltage. Enriched Material. Material in which the percentage of a givell isotope present has been artificially increased so that it is higher than the percentage of that isotope naturally foulld in the material. Enriched uranium contains more of the fissionable isotope uranium-235 than the naturally occurring percentage (0.7 percent). than that needed to E~~~~~~ ~ ~ ~ t More i ~ treactivity y . is built into a achieve criticality. Excess reactor (by using extra fuel) in order to compensate for fuel burnup and the accumulation of fission-pro4uct poisons during operation.
~
NUCLEAR MARINE PROPULSION
MARINE ENGINEERING
Exclusion Area. An area immediately Surrounding a reactor where human habitation is prohibited to assure safety in the event of an accident. Excursion. A sudden, very rapid rise in the power level of a reactor caused by supercriticality. Excursions are usually quickly suppressed by the negative temperature coefficient of the reactor and/or by automatic Control Fast Neutron. A neutron with energy greater than approximat el^ 100,000 electron Fast Reactor. A reactor which the fission chain reaction is sustained primarily by fast neutrons rather than by thermal or intermediate neutrons. Fast reactors little or no moderator to slow down the neutrons from the speeds at which they are ejected from fi~siOning Fissile Material. While sometimes used as a synonym for fissionable material, this term has also acquired a restricted meaning; namely, any material fiss~onableby neutrons of all energies, including (and especially) thermal (slow) neutrons as well as fast neutrons; for examplci uranium-235 and plutonium-239. Fission. The splitting of a heavy nucleus into two approxirnately equal parts (which are nuclei of lighter elements), large amount of accompanied ~ . by the release of a energy and generally one or more neutrons. Fission occur spontaneously, but usually is caused by particles' absorption of gamma rays, neutrons, Or The nuclei (fission fragments) formed by Fission Roducts. the fission of heavy elements, plus the nuclides formed by the fission fragments' radioactive decay. Flux (Neutron). A measure of the intensity radiation. I t is the number of neutrons passing through I square centimeter of a given target in 1 second. Ex~res'ncl as nu, where n = the number of neutrons per cubic centimeter and V = their velocity in centimeters per second. fuc'l Fuel Cycle. The series of steps involved in refini1114~ It includes for nuclear power reactors. the original fabrication of fuel elements1 their use in reactor, chemical processing to recover the fissionah' material remaining in the spent fuel, re-enrichment of th'' fuel material, and refabrication into new fuel elements. Fuel Ellment. A rod, tube, plate, or other mechanica1 sh'~)" or form into which nuclear fuel is fabricated for use in lb reactor. Fusion. The formation of a heavier nucleus from two lightul' ones (such as hydrogen isoto~es),with the attendant relenfln of energy (as in a hydrogen bomb). Gamma Rays. High-energy, short-wavelength electromagnetic radiation. Gamma radiation acconlpanics alpha and beta emissions and always accompa11icrm fission. Gamma rays are very penetrating and are bc'flIa stopped or shielded against by dense materials, such t~ lor'' or depleted uranium. Gamma rays are essentially sirnil&' to X-rays, but are usually more energetic and are ~uclolll in Half-Life. The time in which half the atoms of a particullw radioactive substance disintegrate to another nuc1ear a seconc' "' Measured half-lives vary from millionths billions of Years. Heavy water. Water containing significantly more thl~ll heavy hydro~"l' the natural proportion (one in 6500) (deuterium) atoms to ordinary hydrogen atoms. Hcnvy water is used as a moderator in some reactors becau'o I' slows down neutrons effectively and also has a low crO" section for absorption of neutrons.
I"tamediate
(e~ithermal) Neutron. A neutron having Neutron Economy. The degree to which neutronsin a "IergY greater than that of a thermal neutron but less than reactor are used for desired ends instead of being lost by a fast neutron. The range is generally considered leakage or nonproductive absorption. The desired ends may b~ between about 0.5 and 100,000 electron volts. include propagation of the chain reaction, converting Ion* An Or molecule that has lost or gained one or fertile to fissionable material, or producing isotopes. r ~ ~electrons. ~ r o BY this ionization it becomes electrically Nudem Reactor, A device in which a fission chab reaction Examples: an alpha Particle, which is a helium can be initiated, maintained, and controlled. lts essential minus two electrons; a Proton, which is a hydrogen component is a core with fissionable fuel. ~tusually has a lttom minus its electron. moderator, a reflector, shielding, coolant, and control t"nlzatlon Chamber. An instrument that detects and mechanisms. Sometimes called an atomic ~ f ~ it ~ llloaBures ionizing radiation by measuring the electrical is the basic machine of nuclear energy. OIlrrent that flows when radiation ionizes gas in a chamber, Plutonium. A heavy, radioactive, man-made, metallic making the gas a conductor of electricity. element with atomic number 94. Its most important Irotope* One or two or more atoms with the same atomic isotope is fissionable plutonium-239, produced by neutron llllmber (the same chemical element) but with different irradiation of ~rani~rn-238.~tis used for reactorfuel and hlomic weights. Isotopes usually have very nearly the in weapons. y of high absorption cross section that nllme chemical Properties, but somewhat diierent physical Poison. ~ n material absorbs neutrons unproductively and hence removes them t@akage. In nuclear engineering, the escape of neutrons from the fission chain reaction in a reactor, decreasing its from a reactor core. Leakage lowers a reactor's reactivity. reactivity. Credible Accident. The most serious reactor Power Density. The mte of heat generated per unit volume llucident that can reasonably be imagined from any adverse of a, reactor core. c'ombination of equipment malfunction, operating errors, Pressure Vessel. A strong-walled containerhousing the ~ l other d foreseeable causes, The term is used to analyze core of most types of power reactors; it usually also contains '"la characteristics of a reactor. Reactors are the moderator, reflector, thermal shield, and control rods. rlesigned to be safe even if a maximum credible accident Ressurized-Water Reactor. A power reactorin which heat nhould occur. is transferred from the core to a heat exchanger by water Moderator. Material used in a nuclear reactor to moderate, kept under high pressure to achieve a high temperature I-Q-i slow down, neutrons from the high energies a t which without boiling in the primary system. steam is generated t'iray are Neutrons lose energy by scattering in a secondary circuit. Many reactors producing electric fi'llisions with nuclei of the moderator. A good moderator power are pressurized-water reactors. 111~sa high scattering cross section and low atomic weight. Production Reactor. A reactor designed primarily for 'I1 each there is a chance of absorption. +TO large-scale production of plutonium-239 by neutron irradirt~ d u c ethis loss of neutrons during the slowing-down tion of uranium-238. ~l~~ a reactorused primarily for the I)roceSs, the moderator atoms also should have a low production of radioactive isotopes. ~l@utron-absorPtion cross section. A high-scattering cross Prompt Criticality. ~h~ state of a reactor when the fission roction implies frequent collisions; these give the neutron a chain reaction is sustained solely by promptneutrons; that [latter chance of being slowed down before it is captured is, without the help of delayed neutrons. lt1ld also reduce the average net distance traveled in slowing Rad. The basic unit of absorbed dose of ionizing radiation. clown so that leakage is reduced. Small mass results in a A dose of one rad means the absorption of 100 ergs of average energy loss Per collision (requiring few radiation energy per gram of absorbing material. c'ollisionfJ)and 80 reduces both opportunities for capture Radioisotope. A radioactive isotope. unstable isotope ll'ld distance t~aveled. fhme practical materials are' of an element that decays or disintegrates spontaneous~y, (used in the form of graphite), beryllium and its emitting radiation. More than 1300 natural and artificial ciompoundg, and water. radioisotopes have been identified. Yolecde. A p u p of a t o m held tosether by chemical Reactivity. A measure of the departure of a nuclear reactor foroes. A molecule is the smallest unit of matter which from critic$ity. ~t is about equal to the effective multiofin exist by itself and retain all its chemical properties. ~licationfactor minus one and is thus precisely zero a t NaturalCirculationReactor. A reactor in which the coolant criticality. If there is excess reactivity (positive reac('lsua1ly water) is made to circulate without pumping, that tivity), the reactor is supercritical and its power will rise. IR, by natural convection. Negative reactivity (s~bcriticalit~) will result in a decreasNatural Uranium. Uranium as found in nature contains 0.7 ing power level. IIercent U-235, 99.3 percent of U-238, and a trace of Reflector. A layer or structure of material the U-234. It is also called normal uranium. core of a reactor to reduce the escape of neutrons. I t is '@utron. An uncharged elementary particle that has a mass located between the core and the shield. Neutrons enterfllightl~.greaterthan that of the proton and is found in the ing the reflector are scattered randomly, some of them l'ucleus Of every heavier than hydrogen. A free many times; and a large fraction of them may lleutron is unstable and decays with half-life of about 13 return to the core; it is possible to design a reflector by lainutes into an electron, Proton, and neutrino. Neutrons which more than 90 percent of neutrons that would be lost wstain the fission chain reaction in a nuclear reactor. may be returned. The returned neutrons can then cause Nautmn Caphue. The process in which an atomic nucleus more fissions and improve the neutmneconomy of the Or captures a, neutron. The probability that a reactor. Common reflector materials are graphite, berylgiven material will capture neutrons is measured by its lium, and natural uranium. l l ~ u ~ r capture on cross section, which depends on the energy Rem. The unit of dose of any ionizing radiation which of the neutrons and on the nature of the material. produces the same biologic$ effect as one roentgen of
~
~
134
MARINE ENGINEERING
power excursions. with h~ ~ ~ Neutrons ~ ~ ~in thermal tl equilibrium ~ ~ the substance in which they exist; most commonly, neutrons with a kinetic energy less than 0.5 electron volts. Thermal Reactor. A reactor in which the fission chain reaction is sustained primarily by thermal neutrons. Most reactors are thermal reactors. Thorium. A and, naturally radioactive with atomic number 80 as found in nature,element an atomic weiCt of
~
approximately 232. ~h~ fertile thorium-232 isotope is and can be transmuted to fissionable uranium-233 by neutron irradiation. uranium.A metal, symbol U, ninety-second element of the atomic series. Natural urallium is a mixture principally of the isotopes U-235 and U-238, the former being about 1/140 of the total. The nucleus of TJ-235 is capable of absorbing a neutron of thermal energy and thereupon undergoing fission into two fragments, which fly apart with great energy. The fragments are highly radioactive. neutrons are released almost immediately in each fission (the prompt neutronsf. A small fraction (delayed neutrons) is released later in the radioactive decay of some of the fission products. The fact that fission is induced by
I
NUCLEAR MARINE PROPULSION
bo a state Such that their position ~ of excitation ~ . exactly determined. Atomic nuclei are built UP of two kinds of primary particles; namely, protons and neutronuThe proton carries a single-unit positive charge, equal ill to the electron charge. The neutron is electrically neutral particle carrying no charge. For Ib given element, the umber of protons present in th(' which is the same as the number ahmic positive charges it carries, is called the atomic numb('r of the element. It is identical with the ordinal numberof the element in the familiar periodic table of the Thus, the atomic number of hydrogen is one, of helium21 of lithium 3, and so on up to 92 for uranium, the elemelr'l of highest atomic weight existing in nature to arY appreciable extent. A number of elements heavier thull 92 have been made artificially. The total number of pmtow and neutrons in an atomi(' nucleus is called the mms number of the element. the of both and are ''(' and the mas.l electrO1'r unity on the atomic maSs
and is thereby slowed down.
135
NUCLEAR MARINE PROPULSION
MARINE ENGINEERING
3 37
(Icfinite probability of decaying in a given time; this charge collected on the is equal to that carried docay probability has a constant value characteristic of by the primary ion-pairs. ~ l t h there ~ ~ are ~ marked h Lhe particular I t remains the same differences even among the instruments in each category, imspective of the chemical or physical state of the i o n i z a t i ~ n - ~ hinstruments ~ ~ b ~ ~ fall into two types; ('lement at accessible temperatures and namely, integrating and nonintegrating. with the integrating instrument, the total.quantity of charge due In a given the rate of decay at any instant to a number of ionizing particles is collected over a ~ m ~ o r t i o nto a l the number of parent period of time. I n the nonintegrating (or counting) rrldioactive atoms of the isotope under consideration devices, on the other hand, each capable of present at that instant. causing ionization is recorded separately. The decay constant of the radioactive species is a Integrating devices also can be divided into two Ineasure Of its decay probability. Radioactive decay is classes, which may be referred to as electrostatic and nn pmcessi the actual decay rate being electrodynamic. Those the type are by the decay constant and by the number of often called electroscopes, since they operate on the the particular nuclei present. same principle as the familiar gold-leaf electroscope, The most used method for representing the I n one form of this instrument, two thin sheets of rate Of radioactive decay is by means of the half-life (the me given an electric charge. long as no ions are time required for the number of radioactive nuclei of a present, the leaves will retain their charge and will given kind to decay to half its initial value). Because remain in the same position. ~ f however, , ions are the nature of the decay, this time is formed, those of the appropriate charge will be attracted illdependent Of the amount of the radioisotope present. to the gold leaves, causing a reduction in the charge and Other Types of Radioactive Particles. Gamma a decrease in the actual repulsion of the leaves. rliys are theelectromagnetic radiation released when the For the detection of beta (or alpha) the radianucleus emits its excess energy. These rays are tions are allowed to enter the chamber through a limilar in character to X-rays; they are highly penetrat- "window" of thin aluminum or other light material. itkg and have short wavelengths. Although the term With the window "open" the instrument measures both "gamma ray1' was originally used to describe the electro- beta and gamma radiations; but if it is gamma lnagnetic radiation which frequently accompanies radio- rays only will be detected. llctive decay, the definition has now been extended to Simple forms of integrating ionization chambers of the irwlude such radiations of nuclear origin. The rays electrostatic type, not larger in size than a fountain ltre emitted when a nucleus undergoes transition from a pen, are used extensively in health physics work to higher-energy to a lower-energy state. determine the total amount of radiation (or dosage) to Gamma radiation is described in terms of its photon which an individual has been exposed over a period of UnergY; for exampie, "1-Mev gamma rays." Apart time. fmm the fact that X-rays frequently have lower energies, I n an electrodynamic type of integrating ionization the difference gamma rays and X-rays chamber, a constant potential is maintained between the lwthat the latter are produced outside the atomic nucleus. electmdes by means of a battery. ~f ionizing radiation The X-rays which, as their name implies, enters the chamber at a suficiently high rate, the ions have definite energies (and wavelengths) characteristic produced are swept continuouely to the respective elecof the particular element concerned result from transi- trodes and a steady current flows. The strength of this tion8 between electron energy levels of the atoms. ion current is a direct measure of the rate of entry of the Radiation Detection Equipment. The introionizing particles and hence of the radiation duction of nuclear Power devices introduced a stringent ~h~ ion current may be measured directly, or it may be tiomand for a ~ u r a t eand reliable instruments for the determined by means of a high-impedance voltmeter connected across a mistance through which the current
'
significantly increased.
A ~ nuclear ~ b reactor l ~ consists of fuel containing fissionable material, a moderto slow down neutrons (except in the case of f m t reactom), a coolant to remove the heat generated by fission, a neutron absorber or neutron leakage control Although device, and the necessary the fuel form is different for vmious reactor types and is of the reactor closely with the coolant, the choice of the fuel material subject to fission is limited to three possibilities; these are uranium-235, produced in a uranium-233 (an artificial reactor by irradiation of the fertile material thorium), and resulting from plutonium-23g, or (an artificial conversion of the essentially nonfisionable U-238 by neutron absorbtion). -7 bnvenion and Breeding. Uranium-235 is the in nature to any only fissionable material extent, but in plutonium production reactors (such as those operated by the U. S. Atomic Energy Commission) excess h i o n neutrons are captured by fertile uranium238 which is thereby converted into fissionable plutoonly 0.7 nium-239. since natural uranium percent of uranium-235, it is expected that, with largescale use of nuclear reactors for power production the available uranium-235 will be consumed at some point. since nearly 140 times more nonfissionable uranium-238 is available than is the fissionable uranium-235, it is of ways in which uranium-238 major significance to could be utilized for power production. Although not of immediate interest for marine propulsion, one promising would be to use plutonium-239 to maintain the fission ,,hain and at the same time regenerate more plutonium by utilizing the available ura,nium-23s. If a reactor were to regenerate the same amount of plutonium-239 as it loss by fission, i.e., if the conversion -6 lypeS of ~ i ~ ~Material, i ~
plies in nature are to be utilized as a source of Powor, in then eventually reactors must be uranium-233 serves to maintain a fission chain and at same time supplies neutrons for its regeneration from thorium-232. 1.8 Isoto~es and Rate of Decay. In a nuclc'r reactor the fission process results in liberation of and also in emission of nuclew radiation of differcllll kinds. In general, the remarkably large amount per fis4i011 energy released in fission (about 200 nucleus) manifests itself in the form of heat result ill^ from the kinetic energy of fission fragments. Th'' radioactive decay, neutmnic reactions, and radiativr emissions are not pmductive in a primarily pow('rproducing reactor and must be provided for in the of shielding, reactor operating characteristics, alltl otherwise in reactor design. The majority Of tlv' naturally occurring elements are stable except for a of the high atomic weight elements such radium. "l unstable element undergoes spontaneous radioactin' disintegration at a definite rate with the emission the nucleus of an electrically charged particle (either ''I1 alpha particle, i.e., a helium nucleus or a beta particlr1 i.e., an electron). Often, the products Of decay ' themselves radioactive, expelling either an alpha Or I' beta particle. After a number of stages of disintegrati0l1l an atomic species with a stable nucleusisformedIn addition to naturally occurring radioactive slll) stances, there have been pmduced artificial radioisotop(* of all the known elements. These have been obtai'l("l either by bombardment of stable elements with particles in cyclotrons, etc., by the capture of neutro"n~ or as a result Of n ~ ~ l e fission. ar A few Of them exl'"l alpha particles, but a large number, including most Of t'llr fission products, are beta emitters. For a given radioactive species, every nucleus h&rI' wl('
Ionization chambers with current-measuring devices
container (e.g., a cylindrical vessel, which serves as the a. Ionization Chambers. A number of different other electrode). The electrode to which the measuring J1lstruments for the measurement of nuclear radiations instrument is attached is called the collecting electrode. 'perate On the ionization chamber principle, where the This is frequently maintained at ground potential while
NUCLEAR MARINE PROPULSION
MARINE ENGINEERING
Table1 Gas
IUolocnlar weight. . . . . . . . . . . . . . . . . . . 'I'tlormal conductivity, ~ t u / h r - f t e ~ / f t 300 F... . . . . . . . . . . . . . . . . . . . . . . . . 700 F . . . . . . . . . . . . . . . . . . . . . . . . . . 1330 F . . . . . . . . . . . . . . . . . . . . . . . . . . Vinoosit centipoises at: ......................... 200 700 F . . . . . . . . . . . . . . . . . . . . . . . . 1330 F.. . . . . . . . . . . . . . . . . . . . . . . . . . Hl)o(?lficheat, Btu/lb-F, at: 200 F . . . . . . . . . . . . . . . . . . . . . . . . . . . 700 F . . . . . . . . . . . . . . . . . . . . . . . . . . (330 F . . . . . . . . . . . . . . . . . . . . . . . . . . 1 ' O I I H ~ ~ Ya>STP,lb/fta . . . . . . . . . . . . . . Volumetric speclfic heat at STp, I#1u/fta-F. . . . . . . . . . . . . . . . . . . . . . . lltrl~~tive heat transfer coefficient ('ompared to He for same gas (,ur?peratureand same power output 1tt31atlvepumping power compared to 110 for same gas temperature and mne power output. . . . . . . . . . . . . . . . lbflhtivepumping power compared I,o He.. .......................... lfdlbtlve cost of gas per IOOO fta at STP Italative total activity. . . . . . . . . . . . . . . llflll~tivegamma activity. . . . . . . . . . . . .
g..
are sealed to prsvent a loss of fuel or The fuel fission products to the reactorcoolant under all normal *he fuel elements are rnsembled in fuel bundles that mnsist of from 36 to 1~ fuel m h some assembled in a square array. zirmnium steel for to economic advantages as the same fuel bumup bemuse of its lower neutron capture characteristics. The selection of clad materiall pellet diameter, and other details of the fuel assembly depends upon the design optimization for the particular application. There is a significant amountof experience with metallic fuel typesfrom the naval remtor program. by a high enrichment of These fuels are uraium-235 and are usually fabricated in the form of a msembled into a single fuel multitude of fiat a so-called ~~~h fuel plate is composed assembly. picture frame construction where the uranium metal is a sandwich with ,,ladding material on each edge and on both front and back surfaees. The uranium metal is usually metallurgcally bonded to the fuel cladding to improve heat transfer. ciharacteristiCs generally attributed to metal fuels are: (1) high heavymatomdemity; (2) a significant and reliable thermal-expansion coe~cient;(3) amenability to potentially inexpensive fabrication rnetho&j; and (4) high thermal conductivity. ~~~~~~ldisadvantages of metal (!) low melting temperatures; (2) high rates of fuels radiation-induced swelling; and (3) poor high-temperature compatibility with austenitio stainless steels. A high thermal conductivity and low melting temperature tend to ofisat each other in terms of the specific power attainable, but metal fuels have the potential for somewhat higher specific powers than oxides. may be considered. ~h~~~are other typesof fuels
conventional plant where the maximum temperatuE in limited by the chemical reaction of fuel oxidation and rate of energy release is a direct function Of the rate of fuel injection, a nuclear reactor has no such limitation A nuclear reactor hm a large quantity of in the fuel contained within the Emtor) and the maximum temperature of the reaction is limited only by th'! ability to remove heat or, more properly) by the by spondence between the heat removed fmm the as a functio'l the coolant and the Power level of the ,f the excess reactivity or neutl'0niCs Of the systemThis should be recognized as being true only On a interest for retical basis, since for Power reactors marine pr0pUlSion the neutr0nics Of the system are 'I' power that operation at Power levels above removal of the moderator, which has a negative Or "shutdown" effect on the mactor, and all major 'ystemH are designed to fail safe or shut the reactor dew''' Nevertheless, the point is still valid that generally 'I1'' removl'l most important aspect of l'HiCtor design is Of heat and the most important single the coolant selection. A number Of possible gas molants for reactor systen'* have been considered. However, most Of the pOn*iOr bilities can be eliminated, either by 'Ir lurgical evaluations (air, hydrogen, carbon by heat-transfer considerations (neon! argon). n''' ties of gases which are suitable for reactor shown in Table [lo]. I n addition to relatively Poor heat transfer,am)1' and neon also has problems of neutrofl very expensive. N i t w e n has a high n'utrOn-absorptiO1' cross section and might cause nitriding at high tures. Thus, the list of gaseous Coolants of interest fol marine pmpulsion can be reduced to carbon dioxi(it"
''
141
Properties of Gases Suitable for Reactor H2 2
He 4
0.125 0.199
...
0.097 0.135 0.172
0.010 0.015 0.020
0.023 0.033 0.044
3.47 3.51 3.60 0.0052
1.24 1.24 1.24 0.0104
0.249 0.259 0.279 0.0727
0 0178
0.0129
1.19 0.17
Na
cooling
Air 29
CO 28
0.018 0.028
COr 44
A 40
0.017 0.027
0.013 0.042 0.028
0.012 0.025 0.018
0.020 0.031 0.044
0.017 0.041 0.028
0.027 0.054 0.041
0.241 0.254 0.272 0.0748
0.250 0.262 0.283 0.0727
0.217 0.262 0.295 0.114
0.124 0.124 0.124 0.104
0.0180
0.0179
0.0180
0.0238
0.0129
1 .OO
0.73
0.73
0.72
0.79
0.68
1 .oo
2.2
2.2
2.2
4.0 10 9294 0.0456
4 0 0 7225 1284
4.0 60 0.51 0.5
0.17 1 .O 6 22.7 4.53 X 10-4 18.5 0 0
28 0.018 0.028 0.037
0.020 0.031 0.041
0.039 0.021 0.032 0.042
...
\
0.88
10
1.8 5 1 .O 1 .O
24
40 1392 137,065
(excess N?aCtiVityto overcome the poison effect Of shortlived radioisotopes immediately after shutdown) Since the excess fuel a t start-up provides reactivity in an excess of that required to maintain essential aspect of reactor control is to provide margin There are, however, a number of disadvantages of for shutdown at all conditions. In addition, since the water as a reactor coolant. As more advanced tech- power output of a given reactor is directly pmportional nology is developed, it is probable that water will be to the neutron density or the number of neutrons Per replaced by a reactor coolant that will permit more unit volume fuel, the control system must sense and reactors. The general limitations associated limit any excessive rise of neutron flux during power level with the use of water as a reactor Coolant are:
A water coolant provides the capability of direct steam generation in a boiling-water reactor. Water technologY is well known and system cornponents are available, reliable, and relatively inexpensive.
hecame, although the fast neutrons are slowed down to herma1 energy, there is excessive neutron absorption in tho water as compared to fissile capture of neutrons in NEUTRON
-
ABSORPTION Several important control characteristics of lightI N MODERATOR water reactors Can be observed from Fig. 3. The most l'nportant is that light-water reactors are nonauto(jlltal~ticin that, if the reactor power is increased (even transiently) above the ability of the cooling system to IVmove heat, the moderator-to-fuel ratio is reduced, a I ~'roviding a negative reactivity or shutdown effect. Itemoval of moderator from the fuel region may be llccom~lishedby either steam void formation or by njoction of water. I n the case of boiling-water reactors which are normally designed to operate slightly underWATER-TO-FUEL RATIO and be provided Fig. 3 Variation in reactiGv as a function of wo+er-to~fuelratio for an '10take care of reactivity lost due to steam voids. From idealized, homogeneous, thermally critical lightmwater Ipig. 3 it can be seen that for undermoderated systems a #hamvoid would displace some of the moderator, ""'ulting in a 'light reduction of reactivity. I n addition, good moderators, they the energies of very fast in temperature (mide from spectral effects) and neutmns ss a result of inelasticacattering collisions~ i'mssure a moderator (and therefore Elements such as lead, barium, or imn readily decrease a)o1ant) can be expectedtoresult in changes in reactivity. the neutron energy down to about 0.05 M~~ where the r'ight-water are designed to have a negative hydmgen (elmtic) scattering cross is relatively moderator temperature coefficient. Themfore, a Cold large. Hence a combination of a moderately heavy or r"ctor that is but has not reached operating heavy element with hydrogen will slow down lamperature will be subcritical a t operating temperature. even neutmns of very high energies. this provides good operating characteristics, Essentially, every neutron that undergoes an inelastic da~ending 'POn the magnitude of reactivity swing collision is because of tfie high between hot and it does require sufficient excess probability subsequent slowing down and capture. rwctivity to shut down in the cold condition. Further, even in an elastic collision, in which case the 2'5 Shielding' For such as marine decrease in energy may not be large, the acmmpanying m"ctors, "IMiderable design attention must be given to change in the direction of motion of the leads to the attenuation of emitted nuclear radiations by an i n c m e d length of path through the shield such that lome Of Not only is such shielding the probability of slowing down and capture ia thereby n@oessaq for the protection of Personnel, but a high increased. Consequently, as a first approximation the r"diation backmund will interfere with the operation effectivenWsof a particular material for the attenuation of used in various aspects of reactor Opera- of fast neutrons is determined by the total fast-neutron cross section, which includes both inelastic and elastic the radiation a reactor System includes scattering as well as direct capture. 'Ipha and beta particles, gamma rays, and neutrons Of For maximum efficiency,a shield should attenuatefast 'brious energies, Only gamma rays and neutrons need be neutrons and gamma rays at such a rate that their fluxes since these are far the most penetrating. will be reduced to the maximum permissible values at Any which attenuates these radiations to a the exterior of the shield. hi^ requirement be "lfiCient reduce all the others met if a of high mavl number and hydrogen (or b negligible value. hydrogen pompound) were uniformly distributed in the the reactor three aspects; namely, proper proportions throughout the shield; hbwever, 'lowing down the fast neutrons, capturing the slowed- this is generally not a possibility for shipboard shielding down and forms of gamma both because of ship arrangements and also because of ruliution. since of low mass numbers are the structural requirements for the heavy used. bat moderators, hydrogen in the form of water can In marine propulsion maohm, the weight of shield @'Itably be used ss the shield constituent for slowing is of major importance; if the shielding is too heavy, the fast neutrons' However, at high neutron energies reactor may not be suitable for its intended purpose. 'Iis acattenng cross Of h ~ d m g e nis Very small; this instance, the cost of the shield may be considerable thicknm of hydrogeneous secondary in significance. In addition, shielding that lluterial be required to down the fission results in a relatively concentrated loading distribution '@'limns Of highest energy. The situation can be must be carefully considered since such load distributions impmved by an element of fairly may lead to problems with the ship,s structure. 'lKll mass such substances are not Wherever possible, advantage is taken in shield design
NUCLEAR MARINE PROPULSION
MARINE ENGINEERING
of the attenuating effect of distance, according to the categories according to their functions: (1) heavy elements to absorb the gamma radiation and slow down inverse-square law, on the radiation intensity or flux. very fast neutrons to about 0.05 Mev by inelastic colliIf the operating personnel can be kept a t an appreciable sions; (2) hydrogenous substances to moderate neutrons distance while the reactor is in operation, a significant saving in thickness of the shielding may be feasible. For having energies in the range below about 0.05 Mev; and example, a shield may be made thinner at the top and (3) materials, notably those containing boron, which capture neutrons without producing high-energy gamma bottom if access is restricted to the sides. To protect the heavy structural components surround- rays. Heavy elements which have been employed in metallic ing the core from possible damage from the heat form for ship shielding are iron and lead. Iron turnings liberated upon absorption of radiation, a so-called thermal shield is frequently introduced close to the or punchings, as well as iron oxides, have been incorreactor. It consists of a substantial thickness of a porated in concrete for shielding purposes. Because of its high density and ease of fabrication, dense metal of fairly high melting point (e.g., iron) lead is a good shield component. For gamma rays with placed between the reactor core and the main shield, or biological shield (see Fig. 1). The thermal shield energies in the region of 2 Mev, roughly the same mass lead as of iron is required to absorb a specified fraction consists of a material which effectively absorbs gamma of radiation and inelastically scatters fast neutrons. Since of the radiation. However, a t both higher and lower these two types of radiation carry most of the energy energies, the mass absorption efficiency of lead ill leaking from the reactor, a large amount of the heat appreciably greater than that of iron. The disadvantages of lead in reactor shields are its ~roducedin the shield will be released in the thermal relatively low melting point and its softness. It cannot shield. carry any appreciable portion of the reactor system If the circumstances are such that passengers or other load and, because of relatively low temperature limits, it ship's personnel can be kept at a good distance from a may require cooling. reactor when it is in operation, it is usually desirable to Masonite, with a density of about 1.3 g/cu cm, wun do so. This may be accomplished by designating used as the hydrogenous material in some of the early exclusion areas of several maximum permissible radiation reactors. The number of hydrogen atoms per cubio levels for passengers, ship's crew, and reactor operators centimeter is not much less than that for water. Is on watch. The reflector makes an important contribution to fast- addition it contains both carbon and oxygen, which can neutron shielding. The reflector, especially for a thermal act as moderators. As a general shield material, there is much to recomreactor, is invariably a good moderator (e.g., water, mend concrete since it is strong, inexpensive, tllltl heavy water, beryllium, beryllium oxide, or graphite) adaptable both block and monolithic types of tollso that it will slow down an appreciable portion of the struction. toOrdinary concrete of 2.3 g/cu cm density moderately fast neutrons escaping from the core. contains somewhat less than 10 percent by weight of Because of scattering, many of these slowed-down water when cured. Although the hydrogen concentrrbneutrons are returned to the core, thereby easing the tion in concrete is considerably less than the concenh~lshielding problem. tion in water, the larger proportion of oxygen (whioh ---An imoortant function of reactor shielding is to acts as an additional moderator) and the calcium ntlrl capture the neutrons after they have been slowed down. silicon in concrete compensate, to a great extent, for thcr This is done by inelastic scattering and subsequent difference. Nevertheless, ordinary concrete alone is nol capture by materials in the shield that have a large very efficient a s a reactor shield material since it normtd ly neutron capture probability. This is accomplished if a contains no element of high mass number. good moderating element such as hydrogen is present in Various special ("heavy ") concretes incorporati IIK addition to materials of medium or high mass number. heavy elements have been developed for reactor shieldi~~y. I n addition, an effective shield provides for the absorpIn barytes concrete, for example, the mineral b a r y h ~ ~ , tion of the various primary and secondary gamma rays. consisting mainly of barium sulfate, largely replaces t h ~ The penetration of gamma rays is a function of their sand and gravel aggregate in ordinary concrete. 'I'l~n energy but they are effectively absorbed by a material density of barytes concrete is about 3.5 g/cu cm. Thw ti of high density. The shield material, such as iron or shield of barytes concrete would have to be no thiokrr lead, which serves as the inelastic scatterer of neutrons than an iron-water shield of the same effectivcncn~ will also function as the absorber of gamma radiation. although the total weight of the barytes concrete shioltl Within the energy range of interest, gamma absorption would be greater. is determined essentially by the mass of the shielding 2.6 Safety [14]. Nuclear ships must comply wiC11 material. The thickness of shield required to produce a the rules and regulations of the cognizant agencierr, specified absorption of gamma rays is inversely proportional to the density of the shielding material. Thus a inc1uding:athe United States Atomic Energy Comrni~~iol~ smaller volume of lead than of iron would be required, [15-191; United States Coast Guard [20-231; Uniloll States Department of Commerce; National Bureau of but the masses would be approximately the same. Shield materials may be divided into three broad Standards [24]; International Convention for the Snfnbr - --A
life at Sea [25]; the classification societies [26]; and I411t: rules and regulations of agencies having cognizance over the ports of call [27]. Attention is called to indusi8t+ir~l safety codes, which may be applicable in part. ( lodes of this type include the American Society of Mtwhanical Engineers Boiler and Pressure Vessel Code JYNJ, and the applicable ANSI and IEEE codes )2!),301. A riuclear ship should also provide a degree of safety foruthe non-nuclear portions sufficiently high to ensure rrdo operation of the entire ship. I n this respect proviriolls such as watertight subdivision, stability, fire protection, bilge pumping, fire extinguishing, electrical I~intullations,steering gear, astern power, and navigal,iotlul aids should be evaluated in order to provide for tjl~omaximum practicable safety for the ship. a. Containment. Containment constitutes the outer rrl~c:losureor other systems or arrangements which are provided to prevent the uncontrolled release of hazardous atnounts of radioactivity to normallv accessible snaceu --or t,he ship's environment in the of an accident or ,tl,\lfunction of the nuclear system. I t is tlInt any one of several containment methods may be ,no& suitable for a particular application. Separate prossure-tight containment vessels or containments ~~bilifiing integral portions of the ship's structure are rrtttnples of containment systems that may be utilized. 111 the design of a containment system, the effects of pul-ification of radioactive loops, pressure relief or luppression systems,and systemswhich effectively pmvont core meltdown or its consequences, should be oI'
event
r -
145
should, therefore, be designed to contain, control, and possibly suppress the release of radioactive material which could result from any credible accident. Consideration should be given to (1) the pressure and temperature of the coolant, (2) the energy released as a result of any chemical reaction within the system, (3) the nuclear heat generation, including afterheat, and (4) the energy stored in the structure. The processes involved in the release of this energy are heavily dependent upon the type and specific design of the nuclear power plant. Each system should be evaluated on an individual basis to determine the pressure buildup in relation to the containment d-esign. Missiles resulting from a malfunction of the system components should not result in the release'of hazardous amounts of radioactive or toxic materials to occupied spaces or the ship's environment. The following components are typical of those which may be considered as potential sources of missiles: High-speed rotating equipment. The installation such withinthe should be kept to a minimum, but, if installed within the containment~ be to reduce the probability of rupture of the containment wall due to a failure of any "tating Rods. Positive means be provided to prevent rods from being ejected. ' within the pressurized system. These should be located or protected so as to minimize the p"bability damage to the containment walls in case of failure.
'he containment system should be designed to ensure The primary objective under these circumstances the basic integrity of the containment will be should be to maintain the integrity of the containment tained for any credible operating or twcident and, insofar as practicable, to prevent impairment of the The following factors are typical of those secondary shielding when materials particularly sushiah should be considered: ceptible to fire damage are used (e.g., lead, polyethylene, Or r Maximum credible pressure buildup within the The containment should be designed to remain intact b~tninmentdue to an accident to the nuclear system. if the ship sinks in shallow water, and consideration r Maximum credible internal missile. should be given to provisions for decay heat removal. 0 Location as regards collision or grounding damage. r Itupture of piping, ducts, or similar components Containment integrity should be maintained for a ~ i d eof the containment, and such components con- period of several years following such an incident in order to provide sufficient time for salvage operations. tod to and passing t h r o ~ g h ~ t hcontainment. e b. Shielding and Radiological Safety. Shipboard r External fires and explosions on board. shielding and radiological safety are intended to provide Fires within the containment. standardd for protection against nuclear radiation for 0 Binking of the ship. personnel on board ship and for persons in the vicinity Forces due to ship motion. of such ships in conformance with the cognizant regular Itemoval of reactor decay heat in the event of loss 81 aeolant circulation and provisions for preventing the tory agencies. Inasmuch as all regulatory agencies normally follow the recommendations of the Federal !@&atorcore from melting through the containment. Radiation Council [16, 171, the recommendations of the e Leakage and measurement of leakage rate. Federal Radiation Council should be considered to All nuclear systems producing useful power contain anticipate changes to the criteria specified by the regula@ ~ O T Oenergy ~ indicated by pressure and temperature. tory agencies. It is the intent to provide standards for l ~ d d r nuncontrolled release of this energy and any protection by means of shielding and control of personnel &idltiunal energy that might be generated in a nuclear access so that passengers and shore personnel will not be ~ ~ l d ( !provides nt a potential mechanism for the diaper- exposed to radiation exceeding recommendations for the ~lo11of radioactive material. The containment system general population, and so that operating, maintenance,
.
MARINE ENGINEERING Limits for Table Liquid waste Disporal Discharge to h e Sea as Specified for the NS Savannah
Table 3
NUCLEAR MARINE PROPULSION Radioactive GasWane Specified for the NS Savannah LIMITSON
limb as
a~aseouswaste discharges are to be made while the underway.
is
from shoreline at
depths greater than 200 fathomsd
be used to pmtect against missiles and to provide con[311' The prime function "I d- Health tainment in case of an accident to reactor co~pOneIltS. individu'd' waste DiSposal. Radioactive wastes health physics is to safeguard the to nucl"lL' c. resulting from the hSion whose work is likely to are defined rn the end which contain radioisotopes in significant radiations by taking all steps that are 'Onsidered ncc(u liquids, sary to minimize such exPosure. In addition there "11' quantities. Radioactive wastes include and grnes. Some examples of solid wastes are con- responsibility of making sure that nothing escaping fr''ttl p ~ i C land ~ ,spent the nuclear plant, even in the event Of an accid('l'l" taminated dirt, ,,hips, or other a ~ ~ ~ which i has ~ become ~ e Would ~ represent t ion exchange the ma*mum ranit'' The regulatory bodies contaminated or radioactive may also have to be treated tion exposure limits for personnel, maximum Permis~il''' in The the same manner a waste. purpose of a radiosetive wastedisposal system is concentrations of certain radioisotopes in air and wd''l' and dispose of waste material to mlleot, audit, of any area in a manner that limits the
and maximum permissible amounts Of such Such may accumulate in the human body.
"lN'
rcc('lll
"'cl'datiom are subject to regular review a. increasing ""owledge is g ~ n e Of d the effects of nuclear radiation on '"lf'''lrt ' human body' Dosages are set at such low levels Over many years is unlikely to cause injury. On the Other hand, the levels cannot be so low as to make operation of a plant impossible. 011 board One of the Primary r~pomibilitiesof I'O"lthphysicsisto monitor radiation. This involves the 'l('tcrmination and recording of radiation dosages and (lt)~c rates at nUmerOUS locations. Radiation dosage is rrloasured in terms of the energy absorbed from the radiaIli(l11~ and the dose rate is the time rate at which such f'Jlorg~ is absorbed. I n general, the total dose (or dosage) ~*nouived is the product of the dose rate and the exposure
curve observed on large central-station nuclear plants has not been apparent in unitsof lower power levels. I n fact, in 1969 the product lines for several manufacturer's of central-station units did not include power ratings as low as 300 megawatts electric lmw(e)l. ThereforeJ the Capital costs of around $2@)-$220/kw(e) for to capital costs of grestermw(e) than units must be for unitsthat, of 50based to onmw(e). In addition, it may kw(e) be expected parisom of fossil-fired marine and stationary units, mobile Power Plants will cost about 35 percent more than land-based units of the same rating. Studies 133-351 have indicated that fol large marine reactors of 70,000 to 50,000 shp, fuel costs will be as low as 2.2-2.0 mill/shp-
NUCLEAR MARINE PROPULSION
1A8
Table 4 Summary of Fuel Cost Data P L A N A T CORE Basis-year Natural U $/lb UIO~ Conversion $/kgU Separative work $/unit Tmls com osition % (U-235g f t after . separation processing) Pu credit $/gm fissile Fabrication $ / I t @ Spent fuel shlpplng Reprocessing, $/kgU Reconversion, $/kgU Capacity factor
MARINE 1 2 1977 1974 8.55 8.10 2.29 2.29 26.00 20.00 0.20 0.20
Nf3
Swannah
3 1981 8.80 2.29 20.00 0.20
1 1968 8.00 2.50 30.00 2.53
8.55 7.89 8.00 114.00 100.50 87.50 6.00 6.00 6.00 29.10 31.80 31.80 3.00 3.00 3.00 70% 70% 70%
10.00 88.75 6.00 52.35 5.60
.. .
Table 5 Nuclear Fuel Costs, mills shp-hr PROPOSED PLANT A
NS
Savannah
CORE
1 1974
2 1977
3 1981
1
Resis 1968 - -Direct costs 1.107 0.957 0.934 1.777 net uranium -0.246 -0.200 -0.194 -0.450 plutonium credit 0.611 0.426 0.335 1.667 fabrication shipping, repro., recon. 0.204 0.173 0.156 1.201 subtotal direct costs 1.676 1.356 1.231 4.195 Working capital Outof-m-<
on uramum on fab, Pu,shipping and reurocessmg Imre ' on uranium on fab, Pu, shipping and reprocessing Total cost, mills/shp-hr
0.132 0.107 0.096
0.528
0.025
0.017 0.014
0.030
0.317 0.323 0.320
0.912
0.072 0.064 0.058 2.222 1 .867 1.719
0.061 5.726
to UOz, and the cost of fabricating the U02 into fuel rod assemblies. Uranium occurs in nature in ores which grade from 2 to 5 lb/ton of ore. Yellow-cake can have several chemical forms, including U308or Na2U207,a11 of which of U-235 is fissioned per mwd, and about 0.6 gm of plutoare yellow, hence the name. It is conventional to express nium is produced per mwd (i.e., the conversion ratio i H costs in the units of $/lb UaOs. Historically, U30a 0.50). Later in life, due to fissioning of some of tho prices were over $10/lb in the late 1 9 5 0 ' ~$8/lb ~ in 1968, plutonium, only about 0.6 gm of U-235 is consumed pclr and $6/lb in 1969, all under AEC contract. Free- mwd, and no net gm/mwd of plutonium is producctl; market prices have historically been comparable. The that is, plutonium is fissioned at the same rate it irr U-235 content of natural uranium is 0.711 weight produced. Lifetime averages are about 0.9 gm/mwd 01 consumption and 0.3 gm/mwd of net fisuilo percent, and it takes 14.3 lb of U&a to make 1 kg of 3.0 U-235 plutonium production. Therefore, the net consumptiolr weight percent (w/o) enriched uranium typically required of fissile material (U-235 consumption less net fisnilu as fuel for a marine nuclear plant. The process of enriching the U-235 isotope from 0.711 plutonium production) is only weakly dependent olr w/o to the 2 to 5 w/o required by light-water reactors burnup, its primary dependence being on conversiot~ ratio. As a general rule, what burnup dependor~tn requires that the uranium be in the form of UFs. The process of enriching makes use of the mass there is tends to decrease the net fissile material depletiotr difference between U-235 and U-238 isotopes. The cost as burnup increases. When fuel is discharged from the reactor, it is cool(d greater speed of the U-235 F6 molecules enhances diffusion through membranes more easily. The gaseous for about 6 months and then is shipped to the reproccrlpsing plant for recovery of residual fuel values. A enriched UFs which comes out of the diffusion plant is delivered in standard cylindrical gas bottles. The UFO shipping cask can be rented and will make several tl*il)r is then converted to U02,the fuel form in wKich it is used. back and forth between the reactor and the procexsirrg Direct fabrication costs include the cost of making plant. The shipping cost, therefore, is made up of C I L U ~ ~ UOz pellets, the cost of cladding and end fittings, the rental costs and transportation charges in about ~ ( ~ I I I L ~ cost of assembling pellets into fuel rods and assembling proportions. Spent nuclear fuel has substantial residual value i n ilr fuel rods into fuel assemblies, and the costs of inspection, uranium and plutonium, and ~ossiblyin other f i ~ ~ i o l ~ losses, and scrap recovery. Quoted fabrication costs products or transuranic elements. The spent futd ir can be specified to include all post-enriching processes mechanically chopped and dissolved in acid, and the S ~ B O I ~ ~ and services up to delivery a t the plant. fuel solution then proceeds through several chomicrkl The net cost of fissile material consumed in the reactor process steps to purify the uranium and plutor~i~lrll is based on the reduction in U-235 enrichment associated Provision can also be made to recover desired spent-l'\~~l - - -. with hsioning of the U-235, less the credit obtained due The uranium and plutonium produot 11f by-products. to the production of fissile plutonium, the latter being the plant is in the form of a nitrate. The uraniur~r,i ~ t produced from neutron absorptions in U-238. In a practical reaction system, fission of about 1.3 gm of order to be marketable, must be converted to UFO. Mining and milling costs and ore grade provide 1411c U-235 will produce 1 megawatt-day (mwd) of energy. I n an actual reactor, however, about 8 percent of the minimum cost of production, allowing nothing for I1rs energy is produced by fissions induced in U-238 by high- exploration, depletion of reserve, plant write-off or prolil energy neutrons, and some of the plutonium production A significant assessment of what represents reasorl~~lllu I~ is fissioned. Therefore, a t the start of life about 1.2 gm prices (as distinguished from operating costs) C E ~ LIIY
round in the USAEC purchase agreement with United Ntatesproducers. Through 1968, theAECcontinuedto 1 1 t h ~the W/lb for later years however, the price puid was equal to 85 percent of allowable costs plus $1.6O/lb. Among the allowable coats was a fixed M.64/lb for royalty and exploration costa. Based on iistorical data, an allowable overhead of about $0.3O/lb cjould be included among allowable costs. Tables 4 and 5 are the results of a fuel cost analysis (jonducted by the U. 8. Maritime Administration (351. 'I'he purpose of these comparisons is to project fuel cost latimatea for advanced marine nuclear propubion plants
149
in the 75,OOU-shp range for three reactor cores and compare them with NS Savanruzh data. For the advanced marine propulsion systems, core 1 is designed for 20,500 megawatt days per metric ton of uranium (mwd/mtu) for 2.72 full-power years; core 2 is designed for 25,900 mwd/mtu for 3.45 full-power years; and core 3 is designed for 28,700 mwd/mtu for 3.82 full-power years.
An analysis of Table 5 indicates thatathe marine propulsion system would have fuelfirst cost advanced incentive of 3.5 mills/shp-hr over the NS Savannah with potential for an additional 0.5 mill/shp-hr fuel cost incentive with improved fuel performance.
Section 3 luclear Propulsion Applications 3.1 NS Sawnnah-Pressurized-Water Reactor [ 3 6 9 I n naval circles, the intereat in using nuclear finorgy a t sea led to the launching of the nuclear submarine Nautilus in 1954 and subsequently to today's r~~rolearNavy. The potential of nuclear energy for rrommercial shipping influenced President Eisenhower to moommend construction of a nuclear-powered merchant rhlp: I n 1956 the Department of Commerce and the Atomic Energy Commission (AEC) were authorized to develop and construct such a vessel. On National Maritime Day, May 22, 1958, the ship's ksol was laid, and a little more than a year later the NS duvcmnah was launched as a successor to the earlier ~uvannah. The NS Savannah has a length of 595 ft, beam of 78 ft, &id draws 29.5 f t of water. She carries a crew of 110 L I I 9300 ~ tons of dry cargo. Fully loaded, she displaces 90,000 tons. Like many modem cargo ships, she has bkc, capability of carrying passengers and has cabins for W. Her turbines develop 22,000 shp and her cruising %peedis about 21 knots. The Savannah's nuclear power plant is simple in principle. Uranium, artificially enriched to ensure nadily fissionable atoms, is contained in fuel elements within the core. When the rods are withdrawn (see Ng. 4) a chain reaction starts in the fuel. Fissioning Unnium quickly heats the surrounding water to a high bmperatlire; however, a pressurizer keeps the water ~ ~ l d,enough er pressure to prevent boiling (hence the tlrrtne pressurized-water reactor). The hot water is @lmulatedthrough the boilers, as illustrated by Figs. 5, 8, and 7, where it gives up part of its heat to generate obaem. Steam from the boilers drives the main turbines a d the turbogenerators. After passing through the turbines, the steam is condensed and fed back to the kllara. At full power the Savannah's reactor core gives (IIheat energy equivalent to 80 megawatts. Fuel for the Savannah is uranium enriched to an bverltge value of 4.4 percent. This means it has more Unaium-235 than the 0.7 percent in natural urtmniull..
This slight enrichment simplified the design of the fuel elements and the reador by permitting the use of structural materials that are resistant to corrosion and radiation. The uranium, in the form of uranium dioxide, is compreased into pellets. These are slipped into tubes of stainless steel called. fuel pins. Uranium dioxide was chosen because it does not react chemically with water, has a high melting point, and can hold its shape a t the high-temperature and high-radiation levels within a reactor. The fuel pins are assembled into 32 fuel elements, each containing 164 pins, for a total of 5248. These fuel elements are designed to the standards
UPPER GMD PUTE
LOWER GRID P U T
LamR FLOW RAFFLE ASSEMBLY
F:. i Cutowoy of NS Awnnah'r complete reactor. Note cross-shopad control rods (hat fit between fuel elernenh
I
i
NUCLEAR MARINE PROPULSION
MARlNE ENGINEERING
TOR VESSEL
uRBOGENERATOR
I--HEAT EXCHANGER 4-CONDMSING TANK +LETDOWN COOLERS S-CONTAINMENT DRAIN S-STEAM DRUM TANK 6--PRESSURIZER
7-CHKK ~ A L ~ E a-pu~p -ATE VALVE
Fig. 7 NS Savannah steam generating equipment
n9. 6 NS Savannah system arrangement
~ig.5
Shematic diagram of NS Savannah reactor circuit
set for land-besed nuclear power plants and in addition the shock and vibration from are made to motion of the ship. ~h~ ~ a v a n first ~ ~ core s contahed 17,000 lb of uranium-~35. ~ u r i n gits ,anium, of which 668 lb useful life, about 130 lb of uranium-23S could be fissioned cruise within the core. with one loading she 300,000 nautical miles at a speed of 21 knots, increasing to 23 knots when necesssry. This is equivalent to 12 the e&h at the quator. On such a journey, trips in whhh any conventional ship of the same size would four to five times it. Own weight in fuel, the saVannd would use a quantity of nuclear fuel less than one of her passengem! ~~~~~~~d among the fuel elements within the core rods (see ~ i 4 and ~ 5).~ These . contain 21 neut.nsbsorbing boron. Depending on the position within the core, a nuclear chain reaction can of the or shut down. Neutron detecbe Cirmib that govern the drive tom can mechanism for the mntml These maintain the ,,hain resction automatically at a d&d level, be it full power for top speed, or just enough power to run the ship's generating system. reactor vessel contains the Savannd's A massive nuClem core. ~t is 27 ft high and hm an inner diameter are 6.5 in. thick, with the of more than g ft. ~ t s to prevent clad ~ t h
jH upper head Of the reactor the core. removable to permit loading and ~ o water t from the reactor (hot-water) loop is circulated through a boiler wheresome of its heat is giver' up to make steam in a separate (steam) loop. This design isolates the turbine and engine room from any radioactive materials in the reactor, for there is no 'per' loop. Two path from the hot-water loop to the heat exchangers with independent pumps Were to ensure reliable cooling of the reactor core. per hour boilers generate UP to 265,850 lb Of PRSSUW varying from TI5 to 445 psi. The Savannah's reactor and other system that may contain radioactive materials are enclosed in containment vessel. I n it are the reactor core in it' the stem' pressure vessel, the water Pumps, the d r u m , and the p m u r k e r . F*re shows how 'Omp ~ t l tyh a e parts tWe fitted into the containment , ft long and 35 f t in It a 1-e s t r ~ t u r e50.5 and its walls are carbon steel UP to 28 in. thick that designed to contain an intemd pressure Of 18' psi' a completO This is more p m s w than would result system' rupture of the reactor's When in operation, a nuclear Core gives off neutron# unlavl properLy and gamma rays that could cause confined. Gamma radiation also is given Off by thn Th" radioactive materials that result function of shielding to confioe these radiations
CO~~OS~O The ~.
permit routine operation of the ship and to protect the passengers and crew- The Savannah has two distinct rats Of shields. The first is built around the reactor vassel. It reduces the escape of neutrons and gamma *diation s a c i e n t l y to permit the crew to enter the Wntainment v ~ s e for l short times after the reactor is #hut down. The secondary shielding is outside the sontainment vmsel. It would serve to reduce personnel exposure to radiation should a reactor accident release radioactive materials within the containment vessel. The primary shidd consists of a layer of water 33 in. Chiok, by a layer of lead. The secondary lhielding is a combination of lead and polyethylene lPreund the upper pa* of the containment vessel and oanorete around the lower portion. Any power plant creata radioactive wa~tes. "'la and 'lothing, used m i n s from water purifiers, wiping and other items may become contaminated with rradiwtivit~. The standard practice is to collect most such wastea for disposal, usually by burial on land. liquid. and gases are also created during BlMtor 'peration. Some wastes are so slightly radio&Otive that can be mixed with air or water and be di'ohW& to the air or sea- The Savannah is equipped OO1lect and store wastes, or to release them in diluted
Because there was no practical experience with marine Of Power, the Savannah was equipped with WJXiliary electric "take home" power to bring her back port there be difficulty with her ~eactor. An @isotric motor was coupled to the reduction gear by means a so that the electric motor could drive the m ~ l l e ar t a modest speed in the event of a failure in mode of operation. Two diesel generators WeM provided furnish electrical needs and operate a meling pump for the reactor.
BY virtue of its history of application in U. S. naval vessels and the NS Savannah, the pressurized-water reactor is a prime candidate for merchant &ips. ~h~ PWR offers the attractive characteristics of having a light-water moderator and coolant, high power density, and ability to follow the load. ~t is also by high capital cost, high stored energy in the coolant, and production of low-pressure saturated steam. Considerable study work has been performed to improve the design of marine prmurized-water reactors. The primary innovation hae been the inclusion of a once-through type heat emhanger in the reactor pressure veasel. The operation of such a system was tested by the German ship 0th Hahn in late 1967. hi^ change should significantly reduce the capital costand size of a marine PWR. 3.2 The Babcock 8 Wileox CNSG pressurized-water Reactor [40, 41, 421. The Consolidated Nuclear Steam
Generator (CNSG) is a compact gressurized-water reactor that is designed for merchant marine appliestions. The CNSG incorporates a once-through steam generator which produces superheated steam at a constant pressure over the entire operating load range. The complete CNSG system consists of the reactorwith its integrhl steam generator, pressurizer, reactor coolant pumps, control and safety systems, amiliaw systems, and instrumentation. The reactor can be shopassembled to impmve the quality control and minimize the erection time. The compact vapor suppression system provides both neutron attenuation and enew containment at greatly reduced post-accident The CNSG uses low-enriched fuel fuel costs approximately 40 percent below conventional fuel costs. The reactor for the Geman nuclear ship 0th ~~h~ is of the CNSG type. The CNSG discussed herein was designed to power a
MARINE ENGINEERING
NUCLEAR MARINE PROPULSION
153
dance with the ASME Boiler and Pressure Vessel Code, Elimination of field assembly and welding of the Hootion 111, Nuclear Vessels. The vessel is fabricated primary system with the attendant problems of quality Prom carbon steel and is clad on the inside with stainless control for the reactor coolant boundary. Elimination of any significant source of cold reactor The pressure vessel consists of a 162-in. I D cylindrical coolantkater that might cause a reactivity excursion. nl~ellapproximately 69 in. thick, having an ellipsoidal Limiting of the physical "target" size of the reactor bottom head and an ellipsoidal top section. coolant system as a consideration in ship collision The suppression chamber of the containment is formed accidenb. by the dry-well vessel and an outer concentric cylindrical 1, addition to the foregoing feat-, the preesure vesael- Venting from the dry well pmw-suppremion mangement for the CNSG proto the s u ~ ~ m i chamber on consist of pipea attached to for m e r e containment at greatly reduced (~~enings in the d r ~ e lVW8el l wall and extending into incident p-ures thus decreasae the magnitude of the suppression water. l°K-pressure rupture disks the driving force forand h i o n product leakage and of resultt l o m d l ~seal these vent penetrations for humidity ing fission dispersion. ~h~ decreased size of the nontrol in the dry and prevent bakeow of the containment structure also enhances the ability of the fluppreasionwater at extreme ship roll attitudes. surrounding ship's structure to provide collision protecThe shielding design of the CNSG includes the use of tion. I(jd, water, concrete, and steel. Operational shield 3.3 . The c~~~~~~~~~ E~~~~~~~~~~ U N I M ~ presD r"quinmsnts at elevation a= met by the Con- surized-Water Reartor 1431, with the ~ ~ ~ liderable quantity Of in the equipment, Engineering UNIMOD preasurized-water reactor, plant "le s"pp-ion water in the mntainment, and the compaction is achieved by employing a self-pre88ul-ized noncmte The suppression water e f f ~ t i v e l y reactor with the heat exchanger located within the *ttenuatea the flux, and the shell is reactor vessel. The elimination of external primary-loop "lsd to gamma rayS and opera ti^ components reduces the radioactive volume requiring (fimw) gamma rays to levels- The lead shielding and, hence, the shield size and weight. The clllield furnished is for post-accident shielding. reactor vessel with surrounding shielding is completely The 270-mw(t) CNSG core is composed of 32 fuel encapsulated by the containment vessel. Water within uaaembly and 32 Cluster The core has an the containment vessel provides a rwemoir for vapor nlluivalent diameter of about 70 in. and an active length suppreeaion as well as providing part of the shielding. of 84 in. The total contained uranium is 12,583 kg at The reactor fuel is uranium dioxide in stainaverage enrichment of about 4.1 percent. Radial lem steel tubing. ~h~ active fuel region is 42 in. in Power flattening is accomplished by zone loading with diameter and 60 in. hi&. ~h~ twepasscore contains 61 fuel assemblies of which 36 are in the first pass.
CONTROL ROD DRIVE
the turbine at a 'Onstant pressure OVUr power level videe steam permitted "(' 105,m-shp containership. The the entire load range, the steam system is for thh Shp requirement is 270 mw(t)pressure that is lower than of the CNSG. The unique feature of have a desi%n the steam design p"ssmes the more the CNSG isthat the major components (which are ~ h a r a c t e rby i ~increasing ~~ pressures at vessel as shown on system are located within the reactor on powers). he steam generaor is made up drives Fig. 8. Primsry and coils of tubes connected to feedwater and reactor fuel is lowsection mlLy the top of the vessel. ~ i The ~ Reactivity ~ ~ contml l ~ is~ sheets . in four separate cimits' Each emiched UOr with movablec~uster control rods and f&d operate independently Operation with One' two' "I accomplished by three sections is feasible if one unit must be isolated lumped burnable poisons. of the oncethrough, forced- any reason. The tubes are made Of Incone'' whi(ll' better heat transfer' and lig""a The generator is in the aonular space permits thinner walls, circulation type and is than 'Ore and cOntr"l betweeo the core and pressmv ~ e l .The once- weight The reactor Vessel contains the superspace' and ''I thmugh desip enables the to turbine rods, steam generator, prmsurizer heated which permits an improved generahr' I'''' provides a greatermargin for load changes internal suPPo*s for the 'Ore, eECiency and is desiwd in *rithout moisture carry-over. since the generator pro- control rod drive line. The
of
b
MARINE ENGINEERING
NUCLEAR MARINE PROPULSION
most desirable, consistent with available
-
CIRCULATING
S t e m flow to the propulsion turbine is regulated at the turbine throttle. The steam pressure is maintained within a relatively narrow central band by regulating the feedwater flow through a vanabledelivery feed pump' h. crm ueim c#nbus+im E&me,-ing UNIMOD r e d o r With a decrease in steam demand, for example, the steam pressure will tend to increase, which provides the signal to the feed pump for decreased flow. As a result, 10s~ along the total length of the exchanger, heat is removed from the stem generator and tho an into three downcomers located between each pair of primary water leaving the steam generator and heaterchanger out through the three pumps, the 'Ore in temperature' The is a decresae in 'Ore power to match demand' back into the r e ~ t o vessel r and down to the core idet. The resotor vessel is inside the h i c With reduced load on the steam generator, the extends mund the vess%l, pumps, md the cold shut- inlet temperature rism, the reactor power d e c r e a s ~and ) The average 'On' the core outlet temperature falls. ~h~ space between the canned down mechanisms. compensate for the reactivity vessel assembly and the containment wall is filled with temperature rism slightly gain due to the Doppler broadening effect and the berated waterto a level above the reactor vessel h d . remains critical a t the reduced Power level. With by iron and ~h~ principal radiation shielding is water with lead added in local m a s for additional m a self-pressurized rmctor, the reactor prmsure is detarattenuation. The reactor vessel, heat exchanger, and mined by the core outlet temperature. Hence, the vessel intemals a measurable contribution to the reactor pressure d r o p a t decre&aed power. Tho shielding. Steel slabs rn employed in the annular primary loop, consisting only of the leactor, pumps, allll region between the pressure vessel and Containment once-through steam generator, forms a self-contairld vessel for the additional attenuation. The containment power regulating system with the negative feedbaok by the inherent characteristics of the syst~nl vessel is 16 ft in diameter and 34 ft high. I n UNIMOD a major step h~ been made toward components. NO separate pressurizer is required and 1 1 ~ sidplifying principal components to permit maximum functions need be imposed on this system. utilization of factory preassembly. The total assembled tion on for the much slower changes in nuclaw weight of the plant is only 325 long tons and offers the characteristics due to fuel burnup is similarly ~rovidotl This would mean choice of could one-piece installation. UNIMOD be prassnted aa a complete uoit prior by self-regulating changes in moderator density. Fix*' burnable poison in the core minimizes the total the to shipboard installation. adapt For naturally most to assembly change due to fuel burnup. The excess reactivity plant has been designedminimum interoonneotion. F~~ meawed relative to the end of the Core lifetime. by modules requiring vessel assembly, shielding, and cold-to-hot reactivity control is provided by the poiw'l'example, the four rnodulm, surmounted movable fuel clusters and represents tt'r containment vessel can be installed each weighing less than 10 long tons. The optimum only mechanical rod motion required for leactor number of modules for a p&iculm application will The rods are fully withdrawn in coming up to the '''I' depend upon the deof concumnt fabrication and condition and remain So during operation.
'
155
MARINE ENGINEERING because Of the phase larger pressure vessel would in th(' change and the requirement to condense the internal once-through stetirn generator. ~ l t h considerable ~ ~ ~ h experience with central-statio~l direct-cycle plants has been obtained, there remains reserve with regard to their application aboard ship. The fact that the engine room would be a area and thus require more planning of mutine inspectioll is contrary to the usual ship policy. There also exist'' doubt whether isolation valves can be provided whici'~ tb'! in the event of a nuclear accident, could reactor compartment from the mmhinery mom. Rul" NO. 501 of the "Provisional Rules for the C1assificatiO'l of Nuclear Ships" of Lloyd's Register of ShippiT'~ whiol' essentially states that plant components the primary coolaht flows must be placed in the ment vessel. Under this ruling, one would conclude that only moto~generahrdrives could be considered for direct-cycle marine BWR. It is anticipated that thca' restrictions and reservations with regard to diroctl nuclear marine cycles Can be removed with engineerilly and development work, but the cost Of such cannot be estimated. A BW1' f. Direct-Cycle BWR Radiation reactor can be designed to operate within the allowable dose rates set by the AEC and can be operated board ship. However, the indirect-cycle PWRsystoTr'~ with suppression of disassociation gases by the use hydrogen, stores small volumes of noncOndensibl'!*l " including radioactive gases, prior to I'c%u~~, gaseous waste discharge to the atmosphere il' " negligible- Nevertheless, tl'i' PWR system is low level is far beyond the requirements for safety. ' I ' t~l!a the case of the BWR system, fission gas leakage fuel is unavoidably diluted with disassociated hydroW is I' oxygen. With normal air inleakage, the me effectthat forced circulation produces on the and w'''' measurable r e k a ~ Of e gaseous Waste which, reectormpome is dependent on the pump characterand whether primary-driven jet pumps or within allowable limits, is significantly higher than It istic It may be expected t"ntl feedwater-driven jet pumps are used. Feedwater-driven comparable P w R system. jet pumps have the large advantage of eliminating difficulty will be encountered with licensing and i'*ifil recimulation lines. The volume savings and reliability port entry without the benefit of a pmtotype react"r which result from using feedwater-driven jet pumps are demonstration. Concept'r 3.5 Appraisal of other Marine more significant in a marine reactor than the capital C O S ~ reduction which may result from the removal of the All reactor Concepts other than the PwR, BW1'l 'Ir gas-cooled thermal reactor have the disadvantage of ~ ' ( l b recirculation lines. lL irr additional advantage of forced circulation is the having land-based predecessom (46, 479 481. flow, his means highly unlikely that such systems would be abl(' '*I of contmlling that the plant can be instrumented to make a load change obtain acceptance without first being operated as cer"ml movement. Reduced md move- station plants, unless they possessed extremely attractive without requiring merit makes the system safer and more reliable while characteristics. However, since consideration has b'n'h given to these plants by othen, some remarks concerrli~k~ reducing wear on the control rod drives. other possibilities are ifi order. There are two s~st('sla either naturalor forcede. ~ f i cycle. ~ ~ t circulation systems, the direct cycle must be used to be which have been considered in some detail, marinc gIs economically competitive with the PWR. The savings cooled reactors and organic moderated marine The marine gas-cooled reactor (MGCR) Program w"' of recirculation would mult from the undertaken to develop a high-temperature, gas-cOOll"l~ lines and pump in the naturd-ciroulation indirect systemwould not overcome the disadvantages i n d u c e d . closed-cycle gas turbine Power Plant [49]- The pndn"' I" l-hese disdvantages are in the form of a variable- was initiated in 1958, and in late 1960 was The reactor wua of reactor development. the status The demity moderator and a. larger prmure vessel.
of the effects ~ l t h there ~ ~ is ~a good h that ship motionhave on reactor operational characteristia, it is anticipated that substantial work in this area will be required in a marine BWR program. d. p o n e d - ~ f i C ~ ~ system. On me capital costs for the additional hardW- necessw for forced circulation on amdl land-based plants [less than 250 mw(t)] being too large in generally have been c o m p ~ s o nto the total capital investment to wammt using them. l-his generalization is probably not true in the case of a marine BWR, for a number of r e ~ O ~ One . originates from the of the more important problem of oontainment size. The P m s ~ ~ for a is fixed by the requiment ,turd-circulation that freesurface separation be used, and therefore an increased power density doee not decrease the vessel height. diameter and has only slight effects on in the p.essure vessel l-he amountof water for the natural-circulation plant could be much as three timee that contained in a. high-power-density system. l-his amountof additional water can have comiderable on the containment cost. Marine water reactors usually have high containment pmsures (about 144 psig) because of the limited space available for containment volume. separaA high-power-density BWR, using tion and foxed cimulation, can improve the situation considerably, since the amount of pressurized water can be reduced. ~t also has more flexibility in the pressure drop in its recimulation loop. For the latter reason, several methods are available to stabilize the reactor as well as to reduce the of ship motion. A comiderable reduction in the effect of ship motion on power has been obtained through use of forced circula-
NUCLEAR MARINE PROPULSION 'Ieo moderated and was a 10.5-~exentenriched UO. l'olium-cOoledplant. The reactor exit temperature was be 1500 F~witha thermal efficiencyof 36.2 percent. ('onsiderable development work in both the lmoactorand turbomachinery was necessary. The EBoRE project wa. an gas-cooled Program- Although the reactor "ld plant were attractive from a weight and volume 'handpoint, the BeO moderator and 10.5-percent "llriched fuel entail a fuel cycle cost and capital cost which are not Competitive with other nuclear systems. disadvantage is that the plant was designed for :r2,000 S ~ and P would require considerable development be scaled UP to the 70,000 to 100,000-shp range. organic-moderated, p~ssure-tubereact~malso have hencon~idemd~ but owing to the limited information nvsilable very little can be concluded regarding them 511. The mostattractive feature of this system is the possibility Of having low stored energy in a high power level system. This advantage must be weighed l'uainst the disadvantages of providing a moderator "loanup and makeup System, and the C O S ~of making-up lor the moderator decomposed.
14 '*SafetyConsiderations for ~~~l~~~powerPlants on Merchant ships, 9, SNAME T&R ~ ~ N l ~l3-18. . ~ ~ i 15 U. 8. Atomic E~~~~~ ~ Title ~ ~ i ~ of Federal Regulations; pa* 20, U. S. G~~~~~~~ Printing Office,Washington, D. C. 16 "Background Material for the Development of Radiation protection Standar&, 19 ~ ~N ~ 1,.~~ ~e rd Radiation Council, U. S. ~~~~~~~~~t printing Ofie, Washington, D. c., 1960. 17 '*Bsokground ~ ~ tfor ~the ~D~~~~~~~~~~ i ~ l of Radiation protection standar&, * R~~~~ N ~ 5,. Federal Radiation Council, U. 8. ~~~~~~~~t printing office? Washington, D. c., 1964. 18 "Report of Committee 11on Permissible Dose for Internal Radiation, 1959," publication 2, ~ ~ commission on ~ a d i ~ l protection, ~ ~ i ~ ~ Pergamon l Press, New york, 1960. 19 "General ~~i~~ criteria for ~~~l~~~Power Plant Construction Permits," ~ e d ~~~i~~~~ ~ ~ ~32FR10213, l july 1967. 20 upart 5 & ~ pmUre ~ ~ vessels,j, l ~ ~Title~ Code of Federal ~ ~ ~ ~ l ~ t i ~ ~ . 21 "Part 37-Tank Vessels," Title 46, Code of Federal Regulations. 22 ''Part 79-Pm~enger ~&sels,"Title 46, Code of Glasstone and A. Sesonske, Nuclear Re&r Federal Regulations. Dngineering, Van Nostrand, 1963, 23 "Part 99-Cargo Vessels," Title 46, Code of J. M. Harrer, Nuclear Reactor Control Bngineering, Federal Regulations. Van Nostrand, 1963. 24 "Maximum Permissible Body Burdens and MaxiA. Schultz, of Nudear Reactors and mum Permissible Concentratione of Radio-nuclides in l'ower Plants, McGraw Hill, 1961. Air and in Water for Occupational Exposure," NBS GLwtone and Edlund? Elements of Nuclear Handbook, 69, U. 8. Department of Cornmeroe, U. 8. lieactor Theorti, Van Nostrand, 1952. Government Printing Office, 1959. A. W. Kramer, Boiling Water Readma, Addison 25 'eInternational conference on safety of if^ at Wesley Publishing Co., 1958. Sea, 1960," Chapter 8 and Annex C. Bureau of Shipping.
printing Office, 1958. l1 of High Temperature Gas Resotom," WASH 1085, U. S. Atomic Energy commission, U.S. Government Printing Office, 1969. la 'Maritime Gas Cooled Reactor Project ED-118'ummary, ' EC-l S~AUS Report, Westinghouse B10ctric Co., November 1961. "LMFBR Pmgram WASH 1101-WAsH 1110, U. s. Atomic EnermCommission, Clearing House lor Federal scientific and Technical Information, flpringfleld, Va., 1968.
32 E. K. Sullivan and R. P. Goodwin, "The Nuclem Merchant Ship Prograq" to mi^ Industrial F ~ ~ Washington, D. C., November 1959. 33 N. B. McLeod, "The Economics of Nuclear Fuel in Maritime Application," summary ~ ~N U S~- ~ ~ ~~A , r Contract No. PI-MA6384 PB 16g-935. 34 "Nuclear Capital Costs and Co8ts Trends In Maritime Applications," 2784 Contrsct No. PI-MA66492, PB 173 348, Nuclear Utility Services, October 1966. 35 "Economics of Nuclear ~~~l in ~ a r i t~ i ~~ ~ ~ l i tiom, " 2658, Nuclear Utility Services.
I!\\!
158
MARINE ENGINEERING
36 "Power Plant Description-NMSR Project," BAW-1122, AEC Contract No. AT (30-3)-274, September, 1958. 37 G. E. Kulynych, "Description of the NS Savannah NMSR," BAW-1164, Final Safeguards Report, AEC, Vol. I, I11 through VIII, Contract No. AT(30-3)-274, June 1960. 38 "Technical Safety Evaluation of the NS Savannah," 4th Supp., Brussels, European Atomic Energy Community Euatom, 1966. 39 "The Power Plant for the First Nuclear Merchant Ship (NS Savannah)," Nuclear Merchant Ship Symposium, Contract No. AT(30-3)-274, August 1958. 40 "The Consolidated Nuclear Steam Generator 11-A Conceptual Merchant Ship Nuclear Reactor Design." BAW-1280, Contract No. AT(30-1)-3206, sep&ber 1963. 41 "The Consolidated Nuclear Steam Generator I11 A Conceptual Merchant Ship Nuclear Reactor Design -Revision. " BAW-1289, (Rev. I), Contract NO. ~~(30-11-3206,December 1963. 42 Proceedings of Conference on Nuclear Marine Propulsion, C0NF.-640810, August 1964, p. 55. 43 Proceedings of Conference on Nuclear Marine Propul&on, C0NF.-640810, August, 1964, p. 107. 44 V. A. Mize, B. G. Voorhees, and F. Weinzimmer, "Marine Boiling Water Reactor Nuclear Propulsion
System for 60,000 DWT Tanker," paper presented to the Philadelphia Section of SNAME, February 1960. 45 '22,000 SHP Marine Boiling Water Reactor Power Plant for Commercial Tanker, " Preliminary Plant Descri~tion, Contract No. AT(04-3)-196, General ~ l e c t k CO:, c August 1958. 46 "Nuclear Powered Tanker-Design and Economic Analysis-Pressurized Water Reactor,' George G. Sharp, Inc., Contract No. AT(30-1)-2379, NYO-2860, Combustion Engineering Co., January 1960. 47 Frederic de Hoffman, "Gas-Cooled Reactor Concepts, * General Atomics Division, Nuclear-Powered Merchant Ships Symposium, Washington, D. C., July 1957. 48 "Evaluation of Coolants and Moderators for the Maritime Gas-Cooled Reactor," GA-570, Contract No. AT(04-3)-187, General Atomics, December 1958. 49 K. A. Trickett, "A Review of the Maritime GasCooled Reactor Program," Maritime Gas-Cooled Reactor Program, General Atomic Division, GA-2603, Contract No. AT(04-3)-187, December 1961. 50 R. J. Gimera and R. E. Stanbridge, "Reference Design for an OMR-Powered 38,000 DWT Tanker," NAA-SR-1851, Atomics International Division, Contract No. AT(11-1)-GEN 8, March 1957. 51 "Maritime Organic Moderated and Cooled Reactor," NAA-SR-3859, Contract No. AT(l1-1)-GEN-8, Atomics International, May 1959.
CHAPTER V
-
I.
I
Steam Turbines
Section 1 Nonreheat Main Propulsion Turbines 1.1 Introduction. Although the early development of steam power utilized the reciprocating steam engine as s prime mover, the inherent advantages of the steam turbine soon became apparent and have made it the ohoice for all large modern steam propulsion plants. Turbines are not size limited and can be provided for any power rating up to the maximum likely to be encountered in marine service. High steam pressures and temperatures can be accommodated safely and are limited only by boiler problems. Rotary motion is simpler than reciprocating motion and the unbalanced forces (that produce vibration) which are present in many reciprooating machines can be eliminated in the turbine. I n sddition, the turbine can efficiently utilize a low exhaust pressure, and is characterized by light weight, minimum apace, and low maintenance. ~h~ -fine turbine operates in with the same basic fundamentals as its land-based counterparts in central station and industrial applications but differs in many important respects. This chapter emphasizes those features and characteristics peculiar to marine applications which are derived from the special requirements of marine propulsion or auxiliary drives. The science, and often art, of turbine design is a highly rpecialized field and a number of textbooks have been devoted almost entirely to this subject [I-81.l Space IImitations prohibit a complete, detailed treatment of turbine design and, therefore, the scope of this chapter is limited to information useful to marine engineers, owners, and operators in connection with the application and operation of marine steam turbines. Turbines are umd to propel many types of vessels having widely varying requirements with respect to power, economy, weight, and arrangement. In the dection and development of a turbine design for a 8geciiic application, the following factors must be oonsidered:
(3) The turbine throttle steam pressure and temperature. (4) The steam cycle arrangement, together with the number and location of extraction points and corresponding steam flows. (5) The turbine exhaust vacuum for design purposes. (6) The type of power transmission to the propeller. (7) The astern operating requirements. (8) Spacelimitations of the engineroom arrangement. (9) The importance of machinery weight and size. This chapter deals with the effects of these and other factors upon the design of turbines. 1.2 Steam Conditions. Ever since the early days of the steam turbine, efforts have been made fo improve the steam conditions in order to increase the economy of the power plant. Steam conditions of marine steam power plants have tended to advance about every ten years, with each advance followed by a period during which experience has been obtained and the knowledge gathered for another forward step. Increasing the steam Pressure will reduce the heat rate and steam rate until a pressure of about 2500 psig is reached. A rough rule states that "Doubling the pressure will reduce the heat rate by 4-6 percent." More accurately, a 100-psi increase in initial pressure will reduce the steam rate by the percentages given in Fig. 1. The gains decrease as the pressure increases because the turbine efficiency suffers a t higher pressures. As steam pressures increase, the specific volume of the steam decreases; therefore, the nozzles and blades become smaller a6d less efficient. A limiting pressure is reached for every capacity of turbine a t which the gain due to the improvement in Pressure is offset by the decrease in internal efficiency. For this reason, higher initial pressures may be used more effectively on large turbines. The initial pressures given in Table 1are recommended as practical upper limits for various sizes of propulsion (1) The maximum ahead power needed to provide units. It should be noted, however, that somewhat the desired ship's speed. lower pressures are determined to be optimum when all (2) The relative amounts of time spent a t maximum of the economic factors are considered. The pressures power and reduced cruising powers. given in the table have been chosen because thev ~ e r m i t reasonable utilization of the pipe flange andAvalve 1 Numbere in brackets deaignate References at end of chapter. dimensional standards of the American National
-
MARINE ENGINEERING
STEAM TURBINES
-
100 F INCREASE IN INITIAL TEMPERATURE
1100
-
VARY WITH CHROME.MOLY. CO#POSITION. LIMITS ARE INDICATED BY THE SHADED AREAS.
CHROMEMOLY.
MOISTURE EXCEEDS 12%
MOLY.
INITIAL 750 TEMPERATURE
I
R ~ u in & *am ~
rot. far i n u a d inHial prasure or -Proturn
-STANDARD STEAM CONDITIONS
CARBON
STEEL
Table 1
Recommended Initial Steam Pressures
INITIAL PREBBURE, RATEDBHP
PBIG
in Fig. 2. The r e d a t o n , 8tandsrds Institute as bodies require adherence to these standards for merchant marine propulsion units. Increashg the initial temperature will a h reduce the hest rate and steam rate. As an approximation, a 15 deg F increase in temperatwe at full power or a 25 deg F d l result in a 1-pnent d e increase at lower M~~ accurately, a 100 deg F crease in rate reduce the increase in initial temperature by the percentages given in Fig. 1. unlike a change in initial pressure, a change ternafiects the specifio volume only a relativdy amount. a rssult, the physical dimensions of the parts ,,hanged only slightly, and an increase in both initial temperature d l improve the economy large and small turbines about equally. Part of the from an increase in reduction in heat rate is i,,itial temperature (at constant initial from caud by an increase in turbine present in the lower pressure a reduction in the stages. ~h~ rem~nderis due to the increase in the available energy.
-
For a given initial pressure, t h m is a minimum initial temperature below which the moisture content in the low-pree8ure of the turbine is sufficient to 12aw r n d e a b l e erosion of turbine blades and lo= of &age efficiency. A moisture content of 12 pement in the exhaust is often accepted as limiting, and the c o r n spnding minimum initial temperature. m Y be noted from .Fig. 2. Certain combinstions Of P m u r e and temperature have become widely accepted; these s b &fied conditions are indicated in Fig. 2. Although 'perate at marine s t a m turbines can be designed Or higher) practical pmblems in tempemtms of of marine boilers tend to impose an the upper limit Of g") F. Concern has been expresad reg*ding the sudden imposition of high tempersture upon a astern turbine and exhaust casing; however, it is pomible to accommodate the thermal shock and rapid e*ansion without either distress or distortion1.5 Exhaust Vacuum. A moderate vacuum of Hg become pneraly accepted a deaiy' basis for merchant ProPuLaion turbines. This selectiol' mmprOmLa 'Onis considered a reasonable sidering the worldwide varbtionin seawater tempemtwo of h d h * and mndensin13 and the she, weight, and equipment. LOWseawater t e m ~ e m t m Perfit Uhvacuum conversely, high seawater t e m p e ~ l ~ ~ limit ~ attainable Vacuum. It is often mcdtat the 'Onremsin On ~trU~ti0n to assure that the specific trade route t h u g h o u t its useful life; therefom) it is generally conaidered drsLable to design for 1'
PRESSURE, PSIG
Rg
Standard 'team carditim in relatiail t0 preUUrshmperatura service ratings for ANSI s*dard valw, a d fltlings
good vacuum- In special cases where SOrviceis limited areas of low seawater temperature, the increased and weight of turbines designed for higher than vacuum should be subjected to an oconomic evaluation. The spec5c Of steam increases rapidly as the vacuum is impr0vd. For example, an ~ncreaseof from 28 in' meM:W' to a 30-in- barometer) practically doubles the sPec5c volume. TOhandle this at full load, it necessary to increase pr0porti0nallY the flow areas of the turbine atages at the exhaust end. A more detailed discussion of exhaust losses is given in Section 1.7. It is customry for high-speed, lightweight naval to I'nakesome sacrifice ih economy by accepting a hi@er exhaust Pressure (generally 2-5 P&) at hi& power in Order to reduce the weight and size of turbines condensers. At cruising Powers where economy is more important, the vacuum approaches merchant levels due to the reduced condenser loading. vacuum is for reasons other than Oconomy. When the element is developing power, the ahead blading is being driven backward in at essentiallJ' exhaust Premure. If the vacuum is poor, the wiodage losses of the ahead &ages d l cause
fiangas,
more rapid heating and may limit the allowable qeed or period of operstion. 1.4 Nonextmdon Steam ~ ~ when t ~the . rated full power, the initial steam conditions, and the exhaud vacuum have been selected, it is possible to establish the steam rate which may be from well-designed equipment. Figure 3 gives typical nondraction rates for merchant type, geared-turbine units designed for optimum pefiormnoe at full power with a balance between efficiency,size, weight, and cost. ~h~ reduction in steam rate with improved steam conditions is easily seen as well as the decrease in steamrate which is po&ble with higher powered units. A method for estimating s h m rates at other conditions is given in reference [g] and chapkr 2. Merchant vessels generally operate at or near full power most of their service life; therefore, pefiormence at partial loah tends to be leas important. A typical variation in steam mte at fractional that representative of turbine designs which incorporate no special features to enhance pefiormnce nozzle contml is shown by the other than fiwdage curve marked ~strsigh~thmugh,,in Fig. 4. When partbl-power performanCeis impofimt, as in the case of naval combatant vessels, Evemi means my be
\
163
STEAM TURBINES
MARINE ENGINEERING
RATED FULL POWER. SHP
h.
TurEne ,barn raw, nwxtrocting, for merchaWPe
wed
turEna
ENTROPY fig. 6 Turbine condition cum-reheat
cycle
steam plant cycle as discussed in Chapter 2. Connections for this purpose are provided at turbine stages where the pressure is appropriate to the intended use. Often, it will be found that the number of stages and the corresponding stage pressures selected by the turbine designer to give optimum turbine performance will result Rg. 5 Typical extraction stage pressure c u m in the desired extraction pressure falling between two stages. Ip this case, it is usually best to select either the higher or lower stage pressure rather than disturb the amployed to move the point of optimum steam rate to a turbine design. partial power and to reduce the low-power steam rates. In general, stage pressures vary almost linearly with Possibilities include an excess in total wheel speed at the apparent flow beyond the extraction point, and a full power, the interstage bypass, the series-parallel curve such as Fig. 5 may be obtained from the turbine turbine, and the two-row/onerow control stage. Dis- deaigner. In choosing the proper turbine stage for each
POWER FRACTION
,R
,
ramv-
power fra-
d various Wpes d tu&mn
STEAM TURBINES
hi = hw - E.L.
A straight line joining the "top point" and the "state l i l k ~end
initial pressure ( P 3 and initial temperature (TI). A state line "top point" is then plotted at initial enthalpy and 90 percent of the initial pressure. The "used energy" per pound of steam may be found steam rate X external efficiency EXHAUST PRESSURE
valve and piping arrangement to shift each system from one turbine stage to another as the power varies. The converse of extraction, called induction, wherein excess steam in the plant is introduced to the turbine, * generally discouraged. While this may improve the heat balance, it (1) tends to congest the last stages of the turbine, (2) may introduce large slugs of water from syshms and thereby cause bid damage, and (3) requires an automatic control valve to prevent induction when at standstill, operating astern, or in the event of turbine overspeed. 1.6 Condition Cuwe. The Mollier chart * a representation of pressures, temperatures, of s t , . ~t & consuperheat, and heat point" of steam venient to plot on this ,,hart the at any stagein the turbine; when th& & done for d l stages of the turbine, a line drawn through the= points is called the "condition curve" Or "state line. " cwesn for rated full power and for The powem are urnally obtain& fmm the turbine desiper. They are in defining the characteristics of throughout the turbine, at, extrtwtion points and at the turbine exhaust. In the eventthat a rnndition curve is not available, an approximation of the fdl-power condition c-e may be made for preliminary pwp"as indic&hd in Fi. 6. The initw point is esbbli&ed at the intersection of the
The external efficiency accounts for turbine bearing losses as well as mechanical losses in reduction gears, and for electric drive installations includes the motor and generator losses. If these external losses are not known, the following assumptions may be made: TYPE OF DRIVE EXTERNAL EFFICIENCY Gear drive, single reduction Gear drive, double reduction Electric drive External efficiency at ~ s r t i a powera l may be approxiby varying the external loss (one minus external efficiency) at full power as the 1.7 power of the propellor speed. ~ h ,thalpy , of steam exhausted to the main Condenser is h, = h, - used energy The exhaust loss is caused by the velocity energy in steam leaving the last row of blades which cannot b" is converted to recovered as useful work but friction. Exhaust loss by dissipation in eddies and includes any preasure-d~oplosses between the last who(r1 exit and the exhaust flange- The magnitude of the Ovor= all loss depends upon the particular turbine desi~"l steam flow, and exhawt vacuum. a typical memhant propulsion turbine total exhaust loss (E.L.) at full power with rated ~ x h a u ~ ( ~ vacuum is about 12 Btu/lb and at partia1 powers vari"* flow I*' approximately ae the Square of the ratio of the absolute exhaust PressureThe enthalpy of s t a m at the "state line end point "
'*
point" gives a reasonable approximation to the c~onditioncurve. The nature of the deviation from an ~UJ~urate Curve is indicated by the dashed line in Fig. 6 ltlld is caused by the inability to obtain an average stage nlliciency in the first and last stages of the turbine. Il'ig~re6 also indicates the trend of the condition ~ u r v eat partial powers. Note that a t very low powers, if the initial temperature remains constant, the exhaust Irluby be superheated. 1.7 Exhaust Loss. Among the factors which deter1lliflethe efficiencyand size of a turbine, the exhaust loss ifi one of the more important. Machines designed to uporate economically at high vacuums are inherently Ifir~l;er,but more efficient. For this reason the designer Illuat strike a balance between the required economy and llllo weight and size of the unit. In general, units (Ifl~iWed to operate during a large portion of their life at lli~herpowers should have ample exhaust areas, whereas rlllits which generally operate a t reduced power may be tlo~ignedwith smaller exhausts, because the exhaust loss reduces rapidly as the load decreases. An understanding of the effects of high volumetric lollding per unit of exhaust amulus area is useful. To Illustrate, assume that the exhaust pressure of a typical hrbine is reduced in a series of steps. Three conditions B1.u encountered, as illustrated in Fig. 7. In condition I, the back pressure decreases, the steam velocity (D2) bh the throat of the last row increases until it equals the b(R)ustic velocity corresponding to the steam conditions Lb this point. The steam jet leaves a t the blade exit Bll~le(7). With a further decrease in back pressure, r~~resented by condition 11, the throat pressure and vO1ocit~remain constant, but expansion now takes pluce beyond the throat causing an increase in the efflux Vcrlocity(02") and a deflectionof the jet to angle r 8". Qsudition 111is reached with a further decrease in back ProSSure, when the axial component of the efflux velocity peaches the acoustic velocity. Any further decrease in bnak pressure will result in expansion in the exhaust Or at the condenser inlet, but there d l be no ol''mge in the conditions a t the exhaust annulus; and, UlOmfore~ the steam rate d l not be affected. Thk ee~lditionis sometimes referred to as "choking. Turbine last-stage annulus areas ordinarily are sized lo 6000 to 8000 lb/hr of s b m per square fa* annulus area at 1.5 in. Hg abs; for other back t)mssures, this range vary inversely as the bbnolutepressure. The lower of these values ~epreeents mohineswhich have a very low exhaust loss and are dfJni@edfor normal operation a t full load. The higher Vrlue represents machines that, for economy in weight and size, are designed with relatively small exhaust hoods ,nd are expected to operate at reduced load for the major portion of their life. Equivalent Nonextraction Sleam Rate. WhenIvm is extracted from a turbine, there is a
+
( A ~ D =, RATED STEAM
ASTERN
100
100 AHEAD
%AHEAD RATED R.P.M.
ASTERN fig.
8 Turbine br4ue-ve~n-rpmrekatio&pr 1
reduction in the exhaust flow as compared with nonextraction operation at the same power; therefore, there is a corresponding decrease in exhaust loss. hi^ improvement in p&omance is reflected in the so-called * ~ u i v a l e n nonextrmtion t steam rate, which may be 1-2 percent less than the unonextractionsteamrate,, at the same power. The equivalent nonextraction steam rate, therefore, is based upon specific extraction steam quantitim required by the pafiioular steamcycle and power for each application and should be used in heatbalance ca!culations when extraction is involved. 1.9 Toyue and Speed Characteristics. ~h~ inherent ability of steam turbines to maneuver rapidly is due to their speed-brque chmacteristics. Curves of turbine output torque for typical ahead and astern turbines are plotted against speed in Pig. 8 at rated ahead steam flow, The curves for the astern turbine show that the torque available to decelerate the unit while it is still rotating in the ahead direction increases as the ahead speed from which the maneuver stads is increased. A similar relationship for the ahead turbine torque assists in decelerating the unit from astern rotation. 1.10 Machinery Arrangements. The steam turbine is essentially a -speed machine, whereas the propeller
MARINE E,NGlNEERlNG
is most efficient a t low rprns. In the early part of this century, before the development of speed-reducing devices, a direct drive was necessary. This inefficient compromise produced large, heavy turbines operating far below their most efficient speeds, while propellers of small diameter turning at turbine speeds gave a poor propulsive efficiency. Hydraulic reduction was considered in the early days but has had only limited application, chiefly because of the low efficiency when compared to other forms of speed reduction. Electric drive has enjoyed much wider application than hydraulic drive, usually for special annlications, but reduction gearing is the common choice -r r for higher-powered installations. Steam turbines, reduction gears, shafting, and propellers form a closely related system; and if an optimum overall system design is to be achieved, it is important that the total system requirements be considered in the development of each component. This relates not only to physical arrangement and choice of speeds, but also to such things as the vibratory characteristics of the overall system. The most common geared steam turbine arrangement is the compound unit consisting of a high-pressure turbine and a low-pressure turbine driving a single fixed-pitch propeller through reduction gears and shafting. A complete astern turbine generally is provided in the l~w-~ressure turbine casing. I n lieu of the compound arr'angement, all of the ahead and astern blading may be provided in a single casing at a small sacrifice of 1-2 percent in efficiency. This arrangement is suitable for rated outputs up to about 20,000-22,000 shp and offers a number of advantages
including decreases in space, weight, and cost, reductions in oil and steam piping, and simpler foundations. At least three arrangements of the low-pressure turbine and condenser are in use. I n one, the low-pressure turbine is supported by longitudinal girders forming an integral part of its lower casing; the girders are supported by foundation structure at the forward end and by the gear casing at the aft end. This arrangement permits the condenser to be hung from and located below the turbine and has the advantage that thermal expansion of the condenser does not affect the turbinegear alignment. As an alternative, the turbine may be supported by the condenser. I n this case, thermal expansion of the condenser will raise the turbine centerline with respect to the pinion, and this must be considered in the design of the flexible couplings between the turbine and pinion. I n a third arrangement, the condenser is located forward of the low-pressure turbine such that the turbine exhausts axially into the condenser. This arrangement has the advantage of reduced overall height but the disadvantage of increased machinery length. The detail design of foundations for machinery is tho responsibility of the ship designer; however, it is in order for the machinery supplier to review and comment upon the machinery foundation drawings to ensure that proper support is afforded the equipment and that no undesirable restraints are imposed. Emphasis must be given to the provision of adequate foundation rigidity to avoid vibration conditions. This is particularly important with respect to periodic variations in propeller thrust which may excite longitudinal vibrations in tho propulsion system. Reference [lo] gives a more complete discussion of foundation principles and problema.
Section 2 Reheat Main Propulsion Turbines 1.1
Reheat Principles. Marine steam propulsion
plants generally are restricted to steam temperatures of approximately 950 F because of boiler slagging and corrosion problems arising from fuel impurities, and this imposes a constraint upon the turbine efficiency. Steam temperatures higher than 950 F can be permitted with the use of cIeaner, but more expensive, fuels or by providing special fuel treatment systems. However, it is difficult to justify economically either of these two methods; consequently, if the efficiency of a steam turbine is to be improved, other approaches must be considered. The reheat cycle is the best means available to achieve higher turbine efficiencies and better fuel rates. I n the reheat cycle, steam is withdrawn from the turbine after partial expansion and is passed through a h e exchanger (reheater) where its temperature is raised; it is then readmitted to the turbine for expansion to condenser pressure. The increase in cycle efficiency is due, primarily, to an increase in the mean effective
STEAM TURBINES
temperature at which the heat is added since the reheal. part of the cycle increases the quantity of heat added nl higher temperatures. A secondary but important effeoi is the reduced formation of moisture in the last stages of the turbine which improves the efficiency of these stago#. Referring to a typical condition curve for a reheat cyclcr, Fig. 9, it can be seen that reheat allows the use of high(w pressures without the problems of higher temperatur~. If the same moisture content were to be realized in r straight-through cycle, the initial temperature wouLl have to be increased from T Ito TI'. 2.2 Turbine Performance. The considerations entoring into the selection of initial steam conditions a ~ a l vacuum for reheat plants follow those for nonrehnab units. Marine reheat plants probably will continue to have more modest steam conditions than land-ba~c~tl applications because of lower power ratings and tlln fact that the safety of a ship is dependent upon tl10 reliability of its power plant.
In the case of nonreheat marine power plants, turbine piwformance is specified in terms of the nonextraction sbnnm rate. This criterion adequately meets the require~rlonts of plant designers and purchasing activities; IIII-thermore,it is easily demonstrated on ship trials, and i tln variations are reflected correctly in that portion of the I,otal fuel burned for main-propulsion purposes. When l,lio reheat principle is used, nonextraction steam rate is no longer a proper criterion of turbine performance, since ii, does not recognix the addition of heat in the reheater. 'I'he steam rate varies with the reheat pressure that is mlscted, decreasing as the prewre is dropped; and nvon if, in addition to defined steam conditions and vaouum, a specific reheat pressure is associated with each alstlm rate, the result could not be used for comparison of competitive turbine designs. This may be seen by c~onaiderationof the reheat turbine condition curve, Fig. 1). Ipor example, two turbine designs could be developed rrwh that each has the same initial and reheat ateam cr)nditions, the same condenser vacuum, and the same rlosm rate, but the corresponding plant efficiencies and f ~ rates d could differ. The turbine having a higher afliciency in the stages prior to the reheater and a lower nlficiency following this point would require a greater amount of heat to be added in the reheater at the expense of increased fuel to the boiler. It becomes necessary, therefore, to use some form of b a t rate as the criterion of reheat turbine performance. Ileat rate may be defined in several ways, but a stantlnrdized method for marine units has been suggested ( I I]. I n general, the turbine heat rate in Btu/shp-hr 1~ expressed as
_
Turbine heat rate Heat added to turbine cycle by boiler (Btu/hr) Power Output (shp)
i
I S
The heat added is defined as QT@T
i Where
h *
i
(4)
&T
+
- ~ R W ) Q R ~ H H-RHCR)
(5)
= throttle flow, lb/hr
reheater flow, lb/hr H T = throttle enthalpy, Btu/lb h g = ~ final feedwater enthalpy, Btu/lb HER -- = enthalpy leaving reheater., BtuAb - -,-HCA= enthalpy entering reheater, Btu/lb
Q R A ~=
!
heat rate may be defined on a nonextraction 1 bnisTurbine which is relatively simple in that other components 1
/
Of the cycle are not involved. The demonstration of Ao~lextraction heat rate requires a special test with blaoders closed. Alternatively, an extracting turbine heat rate may be defined in a manner similar to land pracblue Ill] wherein the feed heating arrangement must be mmpletely specified. This type of heat rate can be demonstrated during shipboard trials in the course of regular economy runs but requires additional measuremants, more complex calculations, and more involved
I ENTROPY Fig. 9 Turbine condition c u n a e h e a t cycle
corrective procedures. Typical extracting turbine heat rates are given in Fig. 10 for a five-feed-heater cycle. 2.3 G a s Reheat. Reheat may be accomplished by returning the steam to a special section of the boiler, called the reheater, where its temperature is increased by flue gases. With a gas reheater, the steam can be reheated to thelinitial temperature and maximum cycle efficiency can be realized within the temperature limitation. The improvement in plant performance offered by the gas reheat cycle is equivalent to an increase of about 125 deg F in the initial steam temperature of a nonreheat plant. I n other words, a reheat plant with an initial
MARINE ENGINEERING
STEAM TURBINES BASED UPON REHEAT CYCLES WITH THE FOLLOWING CHARACTERISTICS: THROlTLE PRESSURE, PSlG FEED TEMP. TO BOILER, F THROTTLE TEMPERATURE, F REHEAT TEMPERATURE, F EXHAUST PRESSURE, 1" Hg abs. NO. OF FEED HEATERS
-850 426
- 5 -
1050 1450 18M) 446 419 502 950 850 1.5
MARINE ENGINEERING
STEAM TURBINES
Section 3 Main P ~ p ~ l s i oTnrbines-N~~clear n Cycle
171
172
STEAM TURBINES
MARINE ENGINEERING
FLOW-
Fig. 13 Moisture baffles
Experience has shown that it is desirable to use a corrosion-resistant material for all surfaces subject to moisture impingement. Rotors and casings may be machined from stainless steel, or carbon steel casings may be faced in critical areas with a welded inlay of stainless steel. Although proper choice of materials alleviates the corrosion-erosion problem it does nothing for the loss in efficiency, and therefore steps must be taken to remove as much moi&,ure as possible from the steam path. In general, two approaches are possible: (1) internal separation, which is used for both nuclear and nonnuclear cycles and is described in section 7.4; and (2) external separation, which is treated in the following. 3.5 External Moisture Separution. Moisture may be extracted by removing the steam from the turbine and treating it in an external separator. This method is applicable to the cro~compoundtype of unit where the pressure in the high-pressure turbine exhaust is about 45 psia and the moisture may reach 12 percent at full power as illustrated by Fig. 12. If this moisture were not removed, it would increase still further in expanding through the low-pressure turbine and cause serious erosion problems and a loss in turbine efficiency. The ideal separator should be designed for maximum moisture removal coupled with minimum pressure drop and minimum space requirements. The power loss to the overall plant is about 1 percent for every 1-psi pressure dmp, compared with a power loss of 0.6 percent for every 1percent of moisture entering the low-pressure turbine. Thus, if a net gain in efficiencyis to be realized, the sum of moisture and pressure-drop losses with a separator must be less than the moisture loss with no separator. Many types and arrangements have been developed, but in general external separators consist of a pressure vessel having inlet and outlet connections to the crossover piping, internals arranged to remove moisture, and a drainage system. The separator may be located forward of the high-pressure turbine and supported on the same
longitudinal girders, or arranged at the side of the 11111t and supported by ship's structure. The separator internals may include centrifug~~ I, baffle, or wire-mesh devices. Centrifugal devices rotrllc* the steam flow, which tends to drive the heavier wnl,c&~ droplets to the outer diameter where they are drained olI The baffle t v ~ e function s by collecting moisture on tllcQlt surfaces a n i then allowingVitto drain off. Baffles vrwv from simple chevron styles to more elaborate arranKc! ments having hooks and partially sheltered passages f o l drainage as shown in Fig. 13. Wire-mesh types ~ I I I I ( * tion similarly by collecting moisture through surfric.13 contact. The design of moisture separators is not an ex!ic:l science but rather the result of a great deal of testing r i t ~ ~ l experience. It has been found that the steam ve1oc;il.u configuration in the crossover pipe is complex ILII(I includes secondary ROWS induced by the exhaust e l l ~ ~ w and by the turbine-exhaust hood. This uneven veloc:iI(v distribution may overload some sections of the separd,or A large portion of the moisture in the high-pressl~w turbine exhaust, up to 80 percent, has been found i n 1 consist of water running along the surface of the ~ ) i ~ n wall. This suggests the use of a skimmer at the inl(l1. the separator to avoid overloading the inlet portionr of the separator. Reentrainment of separated moisture may occur W I I ~ ~ I L water on an internal surface is swept back into the V I L ~ M I I stream before it can be drained, and must be avoitltttl because droplets may be produced which are too sn1111l to respond to subsequent separating efforts, or they sair be formed in a late stage and thus escape. Large droplets may split into smaller ones, a procB(!aa called "droplet fracture" that makes separation moln difficult. Such fracture occurs when there is a ltug* relative difference in velocity between the vapor and i.ho droplet, such as may occur when large drops are awol~l' from a trailing edge. In addition, mechanical fraclliw, or split-up, may occur by collision of a large drop wit11Ill1 obstacle. Testing and performance verification of m~isl~ltrr separators involve practical problems. The perforts ance of separators is highly sensitive to the moislr~re particle size. Since both the measurement of drop rim and the artificial creation of moisture truly represent,rl.iv~ of that found in actual turbine exhausts are difficall., I[ not impossible, it has been found necessary to sr~lq~lv saturated steam to a high-pressure turbine coupled l r ~o power absorption device, then lead the exhaust conl,~ri~r ing the required moisture to the separator on t80nl After passing through the separator, the steam is tlac~l. tled through a valve to a low back pressure at which l.Ile steam becomes superheated, and its temperature mtly 11. used to calculate thqresidual moisture at the sepnn~lnlr outlet.
173
Section 4 Camhined Steam and Gas h r l ~ i n eMain Propulsion Cycles
1
'
4.1 COSAG Cycle. In some shipboard applications tliesel engines, gas turbines, and steam turbines can be omployed effectively in various combinations. The prime movers may be combined either mechanically or thermodynamically, or both. The brief discussion which follows is concerned chiefly with the steam turbine ss an element of a combined cycle and is limited to ciombined cycles that significantlyaffect the size or other trharacteristics of the steam turbines. For additional iliscussions regarding combined cycles, see Chapters 1 and 6. Insofar as steam turbine applications are concerned, two combined cycles are commonly considered; these are Ithe combined steam turbine and gas turbine (COSAG) oycle and the combined gas turbine and steam turbine (CdGAS) cycle. In the COSAG cycle the steam and gas turbines are connected to a common reduction gear but u *thermodynamically independent. The chief application of the COSAG principle has been in high-speed 11ava1 vessels of the destroyer type. Such vessels ~~orrnally have a service profile which requires operation J speeds above one-half power for only a very small percentage of their operating life, generally less than one . Accordingly, if this boost power were furnished simple-cycle, aircraft-type gas turbines having a relatively limited life, while the basic ahead md astern power is supplied by a conventional long-life but relatively heavier steam plant, there would be a dgnificant reduction in machinery weight. The weight lrsvad may be used in a number of ways; if it is used to w r y additional fuel, an increase in cruising range of Db-40 percent is possible as compared with a convenMona1 steam plant. Rteam conditions suitable for a conventional steam plant of the same power may be used for the COSAG $yule and, therefore, there are no significant differences (w the design of the steam turbines. Boveral classes of naval vessels, including some British h~troyers,have been built with COSAG machinery 1181, bnd a shore-based prototype plant has been tested by the V~li@dStates Navy [19]. 4.2 COGAS Cycle. In the combined gas turbine and ~ C h mturbine (COGAS) cycle, both mechanical and kenn no dynamic interconnections exist between the cyales. The principle advantage gained by a thermodynamic interconnection lies in the potential for Improved overall efficiency and for savings in space and Wight. This is the result of the ability of the gas turbine cycle to accept heat at a relatively higher
temperature level and the steam cycle's ability M reject heat a t a lower temperature level. The disadvantage of the high-temperature heat rejection in the gas turbine exhaust is minimized by the partial transfer of this heat to the steam cycle. If the exhaust of a simple (nonregenerative) cycle gas turbine is supplied to an unfired wasteheat boiler, the available steam pressure and temperature is low when compared with a conventional steam plant. For example, a gas turbine having an inlet temperature of 1500 F may be expected to have an exhaust temperature of 700-750 F. Heat recovery is dependent upon the boiler design and the amount of heat-transfer surface, but in practical cases the turbine inlet steam conditions are limited to 200-300 psig and 600-625 F. The efficiency of the unfired COGAS cycle increases with an increase in the gas turbine inlet temperature but is relatively insensitive to the gas turbine pressure ratio and to the steam pressure. The steam portion of the plant is relatively simple and there are no high-temperature problems. The reduced heat drop associated with the steam conditions permits a corresponding reduction in the total turbine blade speed that is required and, therefore, fewer stages are needed; this being the case, it is easily possible to use a single-casing steam turbine. Both the steam and gas turbine outputs are supplied to a common reduction gear. Overall propulsion fuel rates of 0.40 lb/shp-hr, which are comparable to the fuel rates of modern steam reheat cycles, are possible but consideration must be given to the cost of the more expensive fuels required by the gas turbine. The waste-heat boiler may also be fired, since 75 percent of the available oxygen is left in the exhaust gases. In a sense, the exhaust-fired combined cycle, therefore, takes the conventional steam cycle, relocates some of its heat-exchange surface, and replaces the forced-draft fans with a gas turbine. Due to the high initial gas temperature, the power derived from the gas ' turbine, and the heat recovery in the boiler (which reduces the hot exhaust gas loss), a net overall reduction in fuel rate of 4-5 percent when compared with a reheat steam cycle can be attained. High-pressure, high-temperature steam conditions, comparable to the normal nonreheat or reheat steam cycles, may be selected for the steam turbines. Consequently, the design and arrangement of the steam turbines suitable for fired COGAS cycles may be generally s i e l a r to units normally supplied for nonreheat and reheat steam cycles. -"
MARINE ENGINEERING
MOISTURE
Section 5 Tllrbine Speed, Number of Sbges, nimensions 'I'heOretically there are by which steam may be expanded in two basic methods a turbine, namely by impulse dges, where the entire stationary nozzle, or by pressure drop occurs in the drop is usually maction stages, where the moving blades, In divided equally thh between and generally does not sharpfixed distinction actual eficiency of pure impulse stages can be exist since the improved by the use of pressure h o p in the moving blades (the pressure hop variesfrom about 5-10 percent in the high-pressure stagesto 3&40 Permnt in the lowpressure stages). impulse stagesare more efficientthan reaction stages at pressures above 4-00 psig, but equal eEciencies may be realbd in the intermediate and low-pressure r e ~ o n s . Indeed, there is little differenoe ih the blade profiles, heights, and angles between the exhauseend blading of mmparable impulse and reaction turbines. E~~~the so-called reaction turbines generally use an impulse desip for their first or contml stage. 5.1
and fieaction.
in the There are Pmnounwd difIemnws, Of number of Stages and the constructional impulse and reaction turbines, as may be noted Figs. 14-17'. The impu15e type is charactehd by 'learanoes individual wheels and dhphrag-1 large between the blades and the caaasing, and a between the diaphragm Packings and the diameter at the high-pressure psking of relatively first stage (see Figs. l4 and 15). AS shown by Figs. 16 and 171 reaction turbines usual1y in employ a d n m - t y ~ erotor) Stationary blade-tip and the casing, close radial or which acts sa a large-diameter h i g h - ~ r e s ~packing e Of dummy piston to balanw theaial the moving in the pressure drops such as shown in Fig. 17) which double-flow turbines, are inherently balanoed). turbines 5.2 Variable S~.d* Marine operate through a wide speed range) and the
PACKING BOX
OIL DEFLECTOR
VALVE LIFT ROD
VALVE STEM LEAKOFF NOZZLE DIAPHRAGM
JOURNAL BEARING THRUST BEARING
-
AGES
--
PACKING BOX
OIL DEFLECTOR JOURNAL BRG. JOURNAL ERG.
AHEAD STEAM INLET
Ol L DEFLECTOR
175
STEAM TURBINES
BLADE RING
1
IMPULSE STAQE ROTATING BLADE
r
IMPULSESTAGE N O n L E BLOCK
178
MARINE ENGINEERING
STEAM TURBINES
80
E0
a W L
I
5
E0
*
K K W
w
5 m
4o
's
f
20 IMPULSE- 3 ROW
0
I
0
I
I
1
.2
.4
.8
I
.8
1.O
VELOCITY RATIO, UIC- BLADE SPEED THEOR€JICAL STEAM VELOCITY Fig. 18
Velodty ratio versus blade dflciency
called a Curtis stage) has a lower peak diagram efficiency, as may be noted from Fig. 18. Frequently, a two-row impulse wheel is used for the first or "control" stage. Theoretically it has the energyabsorbing capacity of four single-row wheels and requires less space. In addition, it is useful for control because it permits the use of a lower first-stage exhaust pressure and temperature which reduces leakage and rotation losses. Because of this reduction in losses, there is very little difference in the overall stage efficiency of a Curtis control stage and the equivalent Rateau stages at the design point. The overall efficiency of a Curtis control stage at part load exceeds that of an equivalent Rateau stage. In some astern turbines, three-row velocity-compounded wheels are used. For this type, the maximum efficiency is reached theoretically when the velocity ratio equals approximately 0.16, as may be seen from Fig. 18. The peak dciency ie less than that for a two-row wheel, but it has the energy-absorbing capacity of nine singlerow wheels. Experience indicates that two- and threerow wheels reach their peak efficiencies when the velocity ratios are somewhat higher than the theoretical values. 5.5 Single-Cylinder Turbine. It is possible to contain all of the ahead and astern turbine stages within a single casing. While such an arrangement could be built for any power output, single-cylinder turbines generally are not considered for powers above 20,000-
25,000 shp. At powers below this range, the singlecylinder turbine has some definite advantages, such as reduced initial, installation, and maintenance costs and more simple gland sealing, gland exhaust, lubricating oil supply and drain systems, overspeed protection system, and machinery foundations. The cross section of a typical single-cylinder turbine is shown in Fig. 19. T h i n impulse stages are provided in the ahead turbine; these consist of a two-row wheel followed by twelve single-row wheels. The asten1 turbine consists of two impulse stages, a two-row wheel, followed by a single-row wheel. The steam rate for a single-casing turbine is approximately one percent higher than a comparable two-casing or cross-compound turbine. This higher steam rate is due to several factors; namely (1) the total blade speed i~ limited by the number of stages which can be accommodated on a single rotor of practical length and also by the maximum rotor speed for which the ahead exhaust stages can be designed, (2) the increased rotor length requires a larger-diameter shaft and consequently interstage leakage losses are greater, and (3) some compromise is necessary with respect to the blade height to diameter ratio selection. 5.6 Cross-Compound Turbine. Historically, marine turbines have been built with as many as four casinm, but these units were directly connected to'the main shafts and operated at propeller speeds. The moder~i C
I 80
STEAM TURBINES
MARINE ENGINEERING
marine turbine, freed from the speed limitations of the propeller by mechanical or electric transmission, normally does not require more than two cylinders. A crosscompound turbine consists of a high-pressure and a lowpressure cylinder arranged so that the ahead steam flow passes through both cylinders in series. The complete astern turbine is incorporated in the exhaust end of the low-pressure turbine. The improved efficiency of a cross-compound unit, when compared to a single-cylinder unit, is due to the ability to provide suflicient stages to achieve an optimum total blade speed. In addition, the high-pressure portion of the turbine can be made smaller, lighter, and more efficient by running it a t higher speed than the lowpressure turbine. 5.7 Double-Flow Turbines. As the rated power and steam flow of a cross-compound turbine are increased, the required diameter a t the exhaust increases correspondingly; and a point is reached where the size becomes objectionable from both an arrangement and a manufacturing viewpoint. To provide additional exhaust area, the low-pressure turbines of high-powered cross-com~oundsets may be designed to have a doubleflow exha\& as &own in Fig. 15. With this arrangemerit, the steam flow is divided and flows through two equal-capacity low-pressure elements to the condenser. Since the same total exhaust area can be provided a t a
reduced diameter, this construction permits the lowpressure turbine to operate a t higher rpm for the samtb stress, which in turn makes possible a reduction in weigh I,. It is customary to provide an astern turbine a t each end of a double-flow rotor. The astern steam flow iti controlled by a single throttle and the flow is dividctl equally between the two turbines. The symmetricr~l arrangement and equal division of flow results in tho same pressure a t each ahead exhaust and therefore 11o pressure differential across the ahead blading. 5.8 Design Selection. In trying to arrive a t the bo~l, overall turbine design for a given set of conditions, tlrtr designer is faced with the selection of proper values for rr s e a t manv variables, including, -. but not limited to, t,htl number of casings, revolutions per minute for etrcll rotor, number of stages, and the nozzle and blade heigl~b for each stage. An optimum design could be arrived rrl, by an interative process consisting of a comparison of rr series of turbine designs in which each of the principr~l variables, one a t a time, is tested through an approprith(rr range. In the evaluation of the results, proper considoration should be given to weight, size, and cost as well nM efficiency. However, a complex study of this typc in seldom necessary because experience and comparirrorl with similar designs aided by the judgement of thn turbine designer make short-cut procedures possible.
.
I
/L
181 HAND CONTROL VALVES
STEAM PRESSURE: AFTER THROTTLEc'MAX. POWER
y:'
-
r A F T E R THROTTLE LOW POWER
4VAILABLE ENERGY PER POUND OF STEAM FULL POWERa ab LOWPOWER= cd
-
u
Section 6 T~~rbine Control 6.1 Power and Speed. Means must be provided to vary the flow of steam through the turbine so that its power output and speed can be controlled. Steam flow may be varied by: A throttle valve w A throttle valve plus hand control valves Bar-lift valves and cam-lift valves Bypass valves Variable boiler pressure 6.2 Throttle Valve Control. The most simple method of regulating steam flow is by a throttle valve in the steam supply to the turbine, as illustrated by Fig. 20. If properly sized, the valve will have little pressure drop when wide open; therefore, a t maximum power, practically full boiler pressure will exist a t the inlet to the first-stage nozzles. As the valve is closed to reduce the rate of steam flow, its pressure drop increases; consequently, a throttling or constant enthalpy process occurs a t the valve and causes a thermodynamic loss since there is a decrease in the available energy per pound of steam. Figure 21 illustrates the reduction in available energy as a result of throttling. Because of throttling losses a t . lower powers, as illustrated by curve A A of Fig. 22, throttle valve control
STEAM PRESSURE: BEFORE THROTTLE
ENTROPY Fig. 21
Efiect of thrMe governing
One nozzle group normally has about one half of the total nozzle area and is controlled only by the throttle valve, while each of the remaining groups is controlled by the throttle valve and a hand control valve. Thus, if the throttle is wide open, the nozzle area and therefore the steam flow may be varied in a series of steps by opening each hand control valve in proper sequence. When the throttle valve and all hand control valves in nervice a t a particular point are wide open, throttling losses will be a t a minimum. The smooth curve AC in Pig. 22 would result if it were possible to have an infinite lumber of hand control valves.
Fig. 23
Throttb valve plus hand control~valves
With the limited number of valves that it is practical to use, there are powers which cannot be obtained by having combbtions of valves wide open or shut. Two modes of operation are possible for these intermediate powers. One procedure is to fully open as many control valves as can be utilized and then partially open one additional control valve to get the exact power desired. The throttling loss of the partially opened valve produces the scalloped effect (or valve loops, as they are
alone is not satisfactory for ahead turbines, but it i~ generally used for astern turbines where high efficiciltiy a t part load is not necessary. 6.3 'ThroWle Valve Plus Hand Control Valvmr. Throttling losses a t reduced powers can be minimizetl if the first stage is of the impulse type and its total noesla area is divided into groups as illustrated by Fig. 21. Y I S ~ STAGE NOZZLES
THROTTLE
Fig. 22
Typical efficiency nnver
fw various types of control
70 25 Fig. 2 0
Throttle valve control
1
I
50
75
THROlTLE FLOW. PER CENT
I 100
MARINE ENGINEERING
STEAM TURBINES
STEAM TURBJNES
MARINE ENGINEERING
184
form. the Pilot valve B closed, the pree dotted area equals the inlet presswe, and the pilot and main are held tkhtly down upon their mats' men the stem is lzted, the Pilot Valve Opens firat,and the premure within the dotted area drops. It is to mntml the pressure in the balance by limiting the lift of the Pilot valve SO that a sufficient force is exerted downward on the disk to prevent chattering, which can be caused by instability Of the flow as the main valve starb to open This of valve is widely used with marine turbines. It is possible to substitute an extmwd h a n d - o ~ a t e d Rlre in the
Alternatively a centrifugal pump may be which case the discharge pressure t3Cts against a Piston and an adjustable spring. The overspeed setti% is obtained by an extrapolation of speed Versus spring s e t t ~for within the operating range- A centr3ugal pump tends to act as a centrifuge, and small air bubbles present in the lubricating oil tend to collect at the center of the impeller. Unless this air i8 vent,,d, the pressurespeed relationship will be affected. ~t has generally not been considered necessary to (D) BALANCED VALVE provide overspeed protection while operating asten'; however, when the astern throttle is power operated for Fig. 29 Vario- typer of valver remote-control purposes, it is a relatively simple matter to include astern overspeed protection. A continuous supply of lubricating oil is essential for Without the resulting complications in subeystem controls and the safe operation of a turbine of lubricating oil, turbine besings may adequate operating procedures do not justify its use. 6.7 Ovenpeed and Low oil Pnssun on. fail in a matter of minutes. If the lubricatix-oil a safe a springOverspeed protection is desirable for every turbine that System Pressure can reach a dangerous speed upon a loss of load. lnthe loaded piston will actuate the Pilot will of gesRd propulsion turbines this can occur if a to the servomotor, and the ahead Operating is broken, and periodic racing c l o ~ . Oil failure may Occur with the propeller is lost or a can occur in heavy weather when the propeller inter- a high ahead speed, and it is important that the vePml pit,&- be available to stop rotation of the shaft rnittently emerges from the water due to the ing. speed-limiting governor is best suited for this is dead in the water. If this is not done, the coastilly " purpose since it will prevent an excessive speed while still period of most large vessels is so long that the governor emergency lubricating oil Willbe used up as the propellnr allowins continuous operationat the setting (usually to 15 percentabove the is dragged through the and the rotati''' continuous rating). ~ rdevices, i ~ which shut off team may damage the gem and turbine bearin6s' 6.8 governor^ A governor generally is prOvidn'l for the primary flow completely, aR notmachinery. overspeed of propulsion pmhfor the control of a turbineelectric prnPulsion unit. Tl'' 'perate througl' tion is Standard for merchant turbines but is not fitted governor is adjustable 80 that it surface ship propu~s~on units where wide range Of speed and is designed so that it to naval simplification and reduced wek,t are important and approximately constant turbine 'pm irrespective Of 'Irn where experience has indicated that, with but few load requirements a t the selected speed setting 'I' exceptions over many years, the risk of dawerous principle, Such a governor is similar to the 'peed contn)lr seater overspeeding is relatively slkhtht. Overspeed protection of turbine-generator sets, but it has a sre furnid"'d is an ereential requirement when an electric drive is used, of speed adjustment. Governors except p o s s i b ~ ythat the generator for geared propulsion turbine since there is the overspeed-limiting protection. m y lose its electrical load. A typical overspeed system is shown in 6.9 Valve Des*n. Three types Of flow to a turbil'n F ~z8. ~ . .The speed eemors me small pOsitive-displa~ commody used to control the 29(a) is siml"n' pumps each driven by its corresponding turbine The single-seated Valve shown in because Of ''l' rotor and supplied from the -in lubriOatingngOil system. but it requira a large lifting The pump discharge pressures at any dvensped may be unbalanced presswe across the valve when in the clOM"' varied by adjustment of the variable Orifices. The position. ail'yl* To reduce the force required to open the disohsrge pRwures operate a pilot valve by acting pilot vniv@ agsiost a spring-loaded piston. ~h~ pilot valve in turn seated, balanced valves hafing an F i u r e m(b) shows this desknin diafTanl are w d . power oil to the proper side of a hyhadic
*
'
'"'
185
bypass valve for the internal pilot, as in ~ i29(,,1~. . Although it can be more nearly balanced because the main valve disk is bed to the valve stem, and is not susceptible to chattering, a hand-operated bypass requires a @parah control, and it is pomible that the operator may foget to it. The dOuble-eabd balanced valve can be used to minimbe the force required to open the valve, butit is seldom used with high-temperature, high-pressure marine turbines a t locations where both mustbe tight under all conditions .Of operation. A valve of this design is shown in diagrammatic form in ~ i29(d). ~ ,
Rotors and Blades The lexth of each row of turbine is governed by the volume flow, the mean diameter Of the flow path, the velocity Of the steam, and the active arc through which flow takes Plwe. Blade lengths gnerauy increase the high-pressure to the low-pressure end the turbine, and the length of the Lasestage is determined by the selected level of exhaust lea* loss- Toward the exhaust end of the as the Of increases, it is necessary to the Outlet angles in order to obtain sufficient flow however, this results in a decrease in efficiency if too far. sinceea~hbladeextends radially from the the pitch is geater a t the tip than a t the To keep this 'preading from causing great a lorn in efficiency due to shape of the flow p s p the length Of having a uniform Cross section their len@h is generally limited to 20 percent Of the diameter Of the flow Path. This limitation can be circumvented by the use of tapered and t*kd having vmY from the root to the tip ae tosuit the steam and blade velocities at each 'long the length. It is usually possible to reduce the cross section from the mot to the tip, which decreases the at the and permits a longer for the same limiting stress. The length of tapered and may 25 Percent to as much as 33 prcent Of the mean diameter of the flow 7'1 'ladeDesign'
When the angles and crowsectional shape of a blade are established, a blade width is such that the calculated bending streseesare weeptable. Bending stresses are caused by axial and tangential forces exerted by the steam and by blade vibration. when the centers of gavity of =tionsst radii do not fall on a radial line, bending stre69es ala, will be introduced by ppnWUgal forces. The centrifugal fore due to the mass of the blade and its rotation causes a tensile stresein the blade that with 10% blades is significant at the blade This varies with the blade len@h, the of the blade speed, and for blades of uniform section it is independent of the section shape or width. T~ minimi% the centaugal foroes in the long blades at the lowpressure end of a tub'me, the blsdes are often &ntrzugal forces and stressesare relatively steady in nature and do not cause vibration or fatigue failures; however, stresm due to centrifugal forces me limited to one half of the yield strength of the Blading must be d e s b d to kthand bendiw stresses under the worstconditiom of lo&%. In the caee of the control or first the highest loadings are experienced at reduced powers due to inCrebsed hest dmps and velocities. F~~ turbines having bypaas valves, the stages preceding the pointwhere the bypass steam is ,readmitted their maximum load just
MARINE ENGINEERING
FRACTION OF RATED SPEED
ng.30
Typical Campbell diagram
STEAM TURBINES
1 87
200 -
MARINE ENGINEERING 11.6 Securing fhe Turbine. The following procedure Traces of water should be removed
traced to its source. by operation of the purifier. At regular intervals oil should be followed when securing a turbine: sam~lesshould be analyzed to check Ph, viscosity, (1) Close all turbine control valves and valves in the -main steam line to the turbine. additives, water ccontent,and other properties. (2) Open all turbine drains. In the event that the lubricating-oil pressure is lost (3) Engage and start the turning gear. This allows for any reason, the low oil pressure trip will shut off the ahead steam. If the vessel is underway ahead, it will the turbine rotors to cool uniformly while the oil circulacontinue to coast for some time. Due to the hydro- tion enables the heat transmitted through the shafts to dynamic action of the water on the propeller, the be carried away from the bearings and thus avoid possible propeller will continue to turn in the ahead direction and damage to the babbitt lining. will rotate the engine. To avoid bearing failures, it is (4) Secure the gland sealing and exhaust systems. (5) Keep the condenser circulating and condensate extremely important that shaft rotation be stopped by the use of astern steam until the vessel stops or oil pumps in operation at minimum speeds until the turbines are drained, then secure. pressure is restored. The inlet steam conditions should be periodically (6) Secure the first-stage air ejector jets but leave monitored. If an abnormally high inlet steam temper* the second-stage jets in service for a few hours to draw ture is permitted over an extended period, damage may air through the turbines. This should be repeated every result. If the inlet temperature is too low, then two or three days to keep the turbines dry. (7) When the turbines have cooled sufficiently to moisture erosion will increase in the last stages of the avoid bowing, secure all associated equipment. low-pressure turbine. (8) Circulate oil and operate the turning gear every The operator should be constantly alert for any abnormal change in noise level, for unusual sounds, and two or three days in port, covering all applicable parts tor indication. of increased vibration, particularly with lubricant to prevent rusting. during maneuvering. If such are noted, slow down 11.7 Emergency Operation. If either turbine of a until the noise or vibration disappears Operate for 10 cross-compound unit is damaged to the extent that it to 15 minutes at this reduced speed, then slowly increase cannot be operated, the other turbine can be run on speed, taking a t least another 15 minutes to reach high-pressure steam by rearranging the steam and operating power. exhaust connections as necessary (see also Chapter 18 If the rotor becomes temporarily bent due to thermal for additional discussion on this subject). The damaged conditions and rubs on the packing strips, heat will be turbine is disconnected from the reduction gear and generated at the shaft surface on a small segment of its remains idle. circumference. This will increase the shaft distortion When operating with the high-pressure turbine alone, and cause a harder rub, which will generate additional a special pipe is provided to exhaust directly to the heat such that the rub becomes progressively worse, condenser. If the astern turbine is confined to the lowpossibly resulting in a severe casualty. Hence, it is pressure turbine, and this is usually the case, no astern necessary to slow down, allow time for temperatures in operation is possible and the astern throttle should be the shaft to equalize, and thus permit the shaft to wired shut to prevent its being opened by mistake. straighten. When the high-pressure turbine is out of service, high11.5 Prolonged Astern Operation. Main propulsion pressure steam may be admitted directly to the inlet of steam turbines designed for merchant ships generally are the low-pressure turbine and controlled by a valve in tho capable of continuous astern operation at 70 percent of supply line which serves temporarily as a throttle. An the ahead speed for one hour without danger of rotation orifice is generally fitted after the valve to limit the steam losses causing overheating of the idle ahead blading. flow to an allowable value. This performance is contingent upon the exhaust vacuum The power output is reduced not only by the decreased being at or near the design value. In addition, the& turbine efficiency but also by consideration of the gear must be no steam leakage into the ahead turbine through loadings when operating with a single turbine. Gear8 the ahead throttle or extraction valves. If temperatures driven by a single turbine are loaded to design torquo in the crossover pipe and high-pressure turbine exceed values when the propeller speed is about 70 percent and allowable values, the speed should be reduced. corresponding power about 35 percent of the normal It should be noted that if the inlet steam temperature the ahead rating. It is generally recommended that tho is constant, the astern exhaust temperature will rise with emergency speed should not exceed 70-75 percent of tho a drop in speed since the exhaust is superheated and the normal ahead rating. turbine efflciency decreases. -
STEAM TURBINES
20 1
Section 12 A~~xiliaryTurbines 12.1 Introduction. For ateam power plants, steam shown in Fig. 43. The performance that may be h~~rbines are also commonly selected as prime movers for expected from properly designed multistage turbines is wxiliaries such as electric generators, feed pumps, and indicated in Fig. 44. tho cargo oil pumps of tankers. Many of the basic Accurate control of speed is essential to maintain 1)rinoiples of steam turbine design and const~ction constant and correct frequency in an a-c electrical system. cwtlined in previous sections apply generally to the Speed control is accomplished by regulating steam flow mmnaller units, but the design criteria may be modified to the unit as directed by a control system utilizing the h a u s e of the reduced power output and because of input from a speed sensor. Although there are several ooanomic considerations. Some of the more important types of sensors and systems, in each case the flow aunsiderations in this regard are discussed in the follow- regulation is achieved by the operation of nozzle control leg sections. valves supplying steam to the first stage of the turbine. 12.2 Ship Service Turbine Generators. The ship A simple mechanical system is shown in Fig. 45(a) in arvice turbine generator (SSTG) provides electric power which a flyweight assembly senses shaft speed. Two for the operation of motors, lighting, communications, flyweights are mounted on a plate which turns about a urcl hotel services. The electrical generating capacity vertical axis driven by the turbine shaft through a worm mquired for a particular veml depends upon its type, and gear. Centrifugal force throws weights outward it,* dze, and its propulsion power, but in most cases the and then compresses the stationary spring, thus lifting elnotric power requirements can be met by the selection the vertical rod and moving the linkage and control valve of a unit from a series of standard ratings which range until an equilibrium position is reached that corresponds !rum 500 to 2500 kw as follows: 500,600,750,1000,1250, to the speed. The speed setting may be changed by 1600, 2000, and 2500 kw. These ratings have been adjustment of the speed changer. This simple form of shown for standrtrdization purposes, as it minimizes the governor is used for small mechanical-drive turbines but srlmber of frame &s required to be offered by rnanu- does not have sufficient force to operate large steamfwturers and thus reduces development costs. I t is control valves. To overcome this difficulty, a pilot valve pmible, of course, to design and build nonstandard and servomotor may be added as shown in Fig. 45(b) to ul~itsof any size that may be required, both above and form a mechanical-hydraulic system. The vertical rod blow this standard range. now operates the pilot valve to admit (or drain) highA typical SSTG consists of a high-speed, multistage pressure oil to (or from) the spring-loaded servomotor gundensing turbine driving a generator through a single- cylinder. As the servomotor piston respqnds, it tends to nduotion gear. These components are mounted on a restore the pilot valve to the neutral position. In some bdplate together with the turbine drain, lubricating oil, cases the pilot valve is double ported and high-pressure 8lrrrhd seal, and gland exhaust systems to form an oil is directed to either the top or bottom of the servoIntegral unit. The turbine may exhaust to the main motor piston as required. wndenser or to an auxiliary condenser. When the An analysis will show that, with any of the preceding wuxiliary condenser is supported by the bedplate, the arrangements, speed will vary slightly with load. The turbine-gear-generator-condenser assembly is called a difference in speed between rated load and no load "paokaged unit. " divided by the rated speed is called the "regulation" or Rteam and vacuum conditions for the SSTG normally "speed droop" and is usually about 3-4 percent. The &rethe same as for the propulsion plant. However, to amount of friction in the mechanism is important, and mrluoe the initial cost of the auxiliary turbine when steam the speed change above and below a mean required to btrlnperatures are 950 F or above, it has been proposed produce corrective action is termed the "dead band" and Miat, while at sea, steam be extracted from an early stage ia a measure of the "sensitivity. " When a sudden change of illc main propulsion turbine a t a reduced pressure and in load occurs and the governor overcorrects followed by blnperature, with an automatic shitover to a source of undercorrection, perhaps continuing for several oscilwprheated steam when at low powers and while in port. lations, the action is called "hunting. " A certain Iaviqs in initial coats also may be realized by the amount of regulation is essential to minimize hunting. Inmidlation of a multistage turbine for normal senice As both regulation and friction are reduced, the sensiwm~d B single-stage, back-pressure-type turbine for tivity is increased; however, the stability is decreased, atn~rdbyservice in lieu of two multistage unifs. thus a compromise is usually necessary, with the regulaA single-cylinder, multistage condensing turbine tion being kept as small as stability or freedom-fromplmrator has a lower efficiency than the main unit hunting considerations will allow. When a-c generator ~wilnarilybecause its rated output is much smsller. An sets are operated in parallel, it is necessary that each kwla turbine generally consiste of five to eight impulse speed governor be adjusted for the same speed regulation nlayos and operates at 8000 to 12000 rpm. A cross if each set is to take an equal share of the load regardless lion of a typical multistsge condensing turbine is of the load variation.
STEAM TURBINES MARINE ENGINEERING
RATED LOAD Fig. 44
- KW
Turbine-generatw steam rate
I n addition to the flyweight or mechanical type of peed sensor, an hydraulic pump driven from the turbine shaft may be used in a n hydraulic system. This pump may be either a positive-displacement type or a centrifugal type and the system may be similar to the speedlimiting governor for main propulsion units described in Section 6.7. A third basic type of governor is available wherein npeed or frequency control may be accomplished by a combination of electric and hydraulic components (see Fig. 46). The speed signal is obtained from the frequency of a small permanent-magnet alternator driven by the turbine rotor; its LGC voltage impulses are oonverted into a d-c voltage which is proportional to ~peed. A reference d-c voltage of opposite polarity, which is representative of the desired operating speed, is a~tablishedby manual adjustment of a speed-setting potentiometer. These two voltages are connected to the input of an electronic amplifier. If the two voltages are equal and opposite; as occurs during steady-state operation, they cancel and there is no voltage input to the amplifier and therefore no change in its output voltage.' The amplifier output voltage drives an electrohydraulic transducer, which directs the flow of oil to a norvomotor that adjusts the governor steam valves to maintain the turbines a t the speed corresponding to the position of the speed-setting potentiometer. If the Ourbine speed changes, the speed signal frequency and Cherefore the voltage supplied to the amplifier change.
The difference between this voltage and the reference voltage is supplied to the amplifier. The amplifier then supplies an output voltage to the electro-hydraulic transducer which causes the steam valves to increase or decrease the steam flow to return the turbine speed to the set value. Stability is achieved by a time delay in the negative feedback around the amplifier. Since there is only one speed a t which the speed signal and reference voltages are equal and opposite, this type of control is "isochronous"; that is, i t maintains the same turbine speed regardless of load variation. A load sensor measuring current in each lead of the generator is utilized to anticipate speed changes and thus improve the dynamic response of the control system. The load sensors of several similar units operating in parallel may be interconnected to ensure equal load sharing with isochronous operation. Operation in parallel with an infinite bus or dissimilar governors is possible by the use of electronic components which introduce droop characteristics as required. 12.3 Single-Stage Auxiliary Turbines. Single-stage turbines, sometimes called mechanical-drive or generalpurpose turbines, may be used to drive pumps, fans, blowers, and standby generating sets. The need for small turbines has resulted in standardized sizes up to 1500 hp with wheel diameters from 12 to 36 in. Rotational speeds vary from 600 to 7200 rpm; the lower speeds apply to the larger wheel sizes used with directconnected turbines and the higher speeds to smaller
MARINE ENGINEERING
tion is generally 5-6 percent. Mechanical-drive turbines
$?$
are designed as complete units arranged for coupling to
SEED CHANGER
the driven unit. The close-coupled, integral type of turbine-driven pump, consisting of a single-stage steam turbine and a single- or two-stage centrifugal pump mounted on the same shaft, has achieved wide application for boiler feed, fire, and tank cleaning services. Packaged units of this type are supplied with a forced-feed lubricating system, speed controls for either constant or differential pressure regulation, speed limiting governor, and back' pressure trip.
PERMANENT MAGNET ALTERNATOR
FREQUENCY SENSOR fRECTIFIER1
I
SHAFT
SETTING WTENTIOMETER
References
CONTROL VALVE
(a) Mechanical
SPEED CHANGER
Fig. 45
Governing
wheels associahd with geared units. The efficiency generally improves with incmadng blade speed, as shown is usually a large energy drop which in ~ i 47.~ There . can be best handled in a twerowstage udng nozzles of i d valves ~ ~ l may be the expanding type. ~ ~ d i ~hand and claing of nos&s to provided to p e r ~ the t acoommodak major c,aoges in load. The speed on the turbine shaft and acts governor is often directly through levers to a c t ~ t the e inlet valve, usually balancsd a ~ e - s e a t e dthrottle valve. Speed ngula-
1 John F. Lee, Theory and Design of Steam and Gas Turbines, McGraw-Hill Book Company, Inc., New York, 1954. 2 Terrell Croft, Steam-Turbine Principles and k Inc., New York, Practice, McGraw-Hill B o ~ Company, 1940. 3 C. B. Biezeno and R. Grammel, Engineerin(/ Dvnamics, Steam Turbines, Vol. 111, Blackie & So11 ~ f m i k d London , and Glasgow, 1954. 4 E. F. Church, Jr., Steam Turbines, McGraw-Hill ~ o o Company, k Inc., New York, 1950. 5 L. E. Newman, Modem Turbines, John Wiley Sons, Inc., New York, 1944. 6 J. K. Salisbury, Steam Turbines and Their Cycl~9, John Wiley & Sons, I ~ c .New , York, 1950. 7 B. G. A. Skrotzki and W. A. Vopat, Steam anfd Gas Turbines, McGraw-Hill Book Company, Inca, New York, 1950. 8 A. Stodola, Steam and Gas Turbines, Peter Smith, New York, 1945. g "Recommended Practices for Preparing Marinn Steam Power Plant Heat Balances," Technical all(l Research Bulletin NO. 3-11? SNAME. 10 R. M. Cashman, "Design of Marine Machincry Foundations, " Trans. S N m E , 196211 C. W. Stott, "Marine Reheat Cycles and S~stemfl Evaluation," Man'ne Technology, ~ 0 1 .71 no. 31 ~ J ~ I Y Marine Installation," Trans. SNAME, 1941. C- K 13 R. p. Giblon, W. I. Signell, N. A. Smith, Spears, and C. W. Stott, "A Modern Steam Reh(illb Power Plant, " New York Section, SNAME, Octoh'r 1965. 14 R. H. "&gey, "High Pressure Steam for Marill0 Propulsion," Trans- SNAME1 1943Worthen, "The Or' 15 H. F. Robinson and ECarrier 8.5- b w e , " Trans. SNAME, 16 A. W. Davis, ''The Application of the ltnB"l"' Cycle to Marine P r o ~ u l i o nwith special Referenm Id' the C.P.R. Beaver Class TurbeElectric Cargo Lir~!r'l" Trans. North East Coast Institution of Engineera l ~ ' " Shipbuilders, 1946-47. and R' .' Mn'''t'' 17 C. H. Grow Jr-, J. T. "A Modern 26,500 SHP stearn Tanker Power l'bu't
SENSOR
L
I
Q T RANSFORMERS
Ro. 46 Electdc gwe-
the Reheat Cycle and a C ~ ~ t ~ ~Pitch l l ~ b l ~ l'rO~ller," The Society of Marhe Port EMneers, N~~ York, September 1965.
24 R. 711. Nolan, "Vibration of Marbe-Turbine Blading," Trans. SNAME, 1949. 25 'warJuok gwedieh Marine Tubine and (-jear Jje~elo~ment, " SNAME, Spring M e e t h , 196b. 26 R*Coats, "Pametrada standard TwbineS, present and Future Outloak, " Trans. IME, 196.5.
NOTE: REDUCE THE STEAM RATE GIVEN BY THE CHART 0.8% FOR EVERY 10°F SUPERHEAT AT THE
PRESSURE R A ~ = O EXHAUST PRESSURE (PSIA) INLETPRESSURE (PSIA)
47 Turbhr.
r a f~ single-stag. (2.row) auriliarl ,,,,h
GAS TURBINES
CHAPTER V I
C,
- LOW
PRESS.'COME
C2- HIGH PRESS. COME
A- 0. white
I
- BURNER - HIGH 'PRESS. TURB. 12 - LOW PRESS. TURB. B
Gas Turbines
TI
FT- FREE POWER TURB. r
8
R (A) SIMPLE
- REGENERATOR
I -INTERCOOLER
CYCLE
Section 1 Basic Considerations 1.1 Introduction. The gas turbine has developed since World War I1 to join the steam turbine and the diesel engine as alternative prime movers for various shipboard applications. Each year its development leads to improved performance and more attractive costs of installation and operation to the point that it gains on the other two prime movers as the economic selection for main propulsion and certain auxiliary drive machinery. This is caused by the fact that the gas turbine inherently profits more than the other two from component improvements and cycle improvements allowed by aerodynamic, heat-transfer, and metallurgical advances. Also, the gas turbine is attractive in that it is inherently subject to ' "package" construction and installation and to automatic control. Therefore, its very numerous variations should be given serious consideration in the selection of a prime mover for the main propulsion plant and the larger auxiliiry machinery. All gas turbine cycles are outgrowths of the Brayton thermodynamic cycle. The Brayton cycle is an ideal cycle in which the working fluid is a perfect gas (atmospheric air in most cases) which is compressed isentropically by a compressor, heated a t a constant elevated pressure in a combustion chamber, then allowed to flow through a turbine expanding isentropically back - to.the compressor suction pressure. The power produced by the turbine is greater than the power required by the compressor. The excess power is used to drive the ship's propeller or some other a d i r y . In gas turbines the efficiency of the components is extremely important since the compressor power is very high compared to its counterpart in competitive thermodynamic cycles. For example, a typical marine propulsion gas turbine rated a t 20,000 shp might require a 30,000-shp compressor and, therefore, 50,000 shp in turbine power to balance the cycle. Thus, with 80,000 shp of machinery involved, a 1-percent improvement in the component efficiencies would improve the cycle by 800 hp, which is 4 percent of the overall performance of the 20,000-shp cycle. The corresponding steam turbine cycle would have been improved by only about 1.05 percent if the counterpart component efficiencies were improved by 1 percent. This example illustrates why year-+year developments result in such marked improvements in the performance of gas turbines as compared with alternative prime movers.
At moderate turbine inlet pressures and temperatures and with the component efficiencies attainable when gas turbines were first developed, the simple open-cycle gas turbine operating with atmospheric air as the working fluid and burning light distillate fuel was limited in output and specific fuel consumption (cycle efficiency). However during the subsequent stages of progressive development, the cycle efficiency has been greatly improved by the following changes : Higher compressor pressure ratios. Higher turbine inlet temperatures which were permitted by metallurgical and cooling developments. Improved compressor and turbine stage efficiencies. Increased compressor pressure loading per stage. Improved combustion efficiency. The introduction of intercooling in the process of compression. The introduction of reheating (a second comb us ti or^ chamber between the compressor turbine and power turbine). The introduction of regeneration (recovery of wasto heat from the turbine exhaust and subsequent addition to the compressor discharge air flow before it enters the combustion chamber). Further waste-heat recovery.
.
Different designs have used various combinations of the foregoing to provide vast improvements in the cycb efficiency and specific air consumption. These combinations modify but do not change the basic concept^ of the Brayton cycle. At the same time they introducu complications into the arrangements. Particular consideration must be given to the gaH turbine cycle selected (i.e., the simplest Brayton cyclo or the more complicated variations). The basic advantages of the gas turbine for marine applications are ita simplicity and light weight. As an internal combustiorl engine, it is a self-contained power plant in one packa~a with a minimum number of large supporting auxiliarien. The advantages this confers in space, weight, and reducorl maintenance are very significant. A reduction in fuol consumption is always desirable; but machinery cod,rr must also be considered, and this too will vary with tlrtr cycle and the arrangement. 1.2 Cycle Performance. The considerations ia-
I
( 6 ) REGENERATIVE
(
1
~ OUT 3
CYCLE
CIR
4
OUT
(C) REGENERATIVE CYCLE WlTH REHEAT
OUT
(€1 REGENERATIVE CYCLE WlTH REHEAT AND INTERCOOLING
Rg. 1 Cyde arrangomonh
volved in the selection of the design pressures and
temperatures and the various components in the cycle combustion chamber (burner), from which it expands nre best illustrated by a n example. For discussion purposes, consider a gas turbine cycle in which the turbine tjxpansion takes place in a "two-shaft " turbine, one shaft baing the free turbine shaft which drives the propeller lhrough a set of double-reduction gears. huther, unsume a plant rating of 20,000 shp using a fuel with a lower heating value of 18,400 Btu/lb. Ambient condillions are taken as 14.7 psia and 75 F. The common r~llowancesmade for duct losses are 4-in. H 2 0 in the irilet duct and 6-in. H 2 0 in the exhaust duct. The gear IOM, about 2 1 percent, is allowed for in the performance hut no allowance is made for miscellaneous hotel services and small non-engine-driven auxiliary power rnquirements. Figure l(a) shows the basic components of the simple ~rycle. Atmospheric air is drawn into the mmpremr, where it is compressed, then heated under pressure in the
back to atmospheric pressure through a high-pressure turbine (to drive the compressor), and free power turbine (to drive the geared propeller). Figure 2 shows the performance attainable with this simple cycle within the limitations just discussed. It should be noted that the values shown are not to be taken as absolute eince the compressor, combustion, turbine, and other efficiencies vary with particular designs and their state of development. Note that a t a turbine inlet temperature of 1600 F, a fuel rate of 0.45 Ib/shp-hr is attainable a t a high optimum compression ratio of 15 to 1and that even lower values a n be attained a t higher temperatures and pressure ratios. The performance indicated by Fig. 2 is very good considering such a simple thermodynamic cycle. It is nearly attainable with engines derived from aircraft jet engines in which the jet's exhaust nozzle is replaced by a specially designed power turbine. The jet engine's
GAS TURBINES MARINE ENGINEERING
208
SIGN COMPRESSION RATIO fig. 2
Smple cyde pdormance
4 PdOnnance 0f a regenerative cyck wifh reheat
GAS TURBINES 210
21 1
MARINE ENGINEERING
AMBIENT TEMPERATURE. O F
k
Fig. 7 Typical effect of ambient temperature on rpecitlc fuel rate, shaft horsepower, and air fkw
gas turbine's overall performance. A change of only 10 deg F in ambient temperature will change the power capacity of the unit by approximately 5 percent and there are also significant changes in specific fuel consumption and air flow. Figure 7 shows a typical ambient te'mperature correction curve for variations in these three variables. Note that if a unit designed for a 75 F atnbient is operated on a 10 F day, the engine output can be increased 28 percent, the air flow can be increased 14 percent, and the specific fuel consumption is reduced about 8 percent, provided the engine is designed to hold up structurally and nozzled to pass the fuel flow a t the increased power. This variation with ambient conditions can vary in a small way between designs and types of cycles. However, Fig. 7 is generally valid for simple cycles and regenerative cycles of any design to a reasonable degree of accuracy provided the machine is not limited mechanically. Both output and efficiency are very sensitive to pressure drops anywhere in the cycle, but those in the inlet and exhaust system are the only ones which the marine engineer can control. The inlet pressure drop is the more critical one since it not only introduces an efficiency loss into the cycle, but it also reduces the weight flow of air. A pressure drop of 1 percent (4 in. water) in the inlet reduces the net output by 2 to 23 percent and increases the specific fuel consumption by 1 to la percent while a pressure drop of 1 percent (4 in. water) a t the exhaust reduces the output by 1 to 13 percent and increases the specific fuel consumption by 1 to 1 i percent. The increase in fuel rate corresponds, of course, to the increase in heat rate and reduction in thermal efficiency. Marine units frequently are rated on the basis of 4-in. H 2 0 inlet duct pressure loss and 6-in. Hz0 exhaust duct pressure loss.
Typical variations in the weight of gas turbine machinery and "all purpose" fuel consumption are given by Figs. 10 and 12 of Chapter 1; these data are only typical, and actual values for any specific application vary not only with the type of unit considered but also with the manufacturer and with progress in the state of the art. Reductions in weight and improvements in fuel consumption may be confidently expected in the future. 1.4 Combined Cycles. The gas turbine is a very flexible power plant and can be applied not only alone but also in combination with other prime movers. Various combinations have been proposed and some of them have been applied successfully. Some possible combinations include: combined diesel and gas turbine plants, abbreviated CODAG; combined steam and gas turbine plants, COSAG; and combined gas turbine and gas turbine plants, COGAG. In these cycles gas turbines and other engines or gas turbines of two different sizes or types are combined in one plant to give optimum performance over a very wide range of power and speed. In addition, combined diesel or gas turbine plants, CODOG, or even combined gas turbine or gas turbine plants, COGOG, (where one plant is a diesel or a small gas turbine, respectively, for use a t low or cruising powers, and the other a large gas turbine which operates alone a t high ~owers)are also possibilities 11-31.' - he gas turbine can also be combined with a steam turbine plant in various ways. The designation COSAG normally implies a cycle in which the steam plant and the gas turbine are essentially independent, but they may be geared to the mme propeller shaft [ P 5 ] . However, -
'Numbera in bracketa designate References at end of chapter.
GAS TURBINES
MARINE ENGINEERING
1.9 Operation and Maintenance. A further feature of the gas turbine is its low manning requirements and ready adaptability to automation. As indicated in the wction on controls, the gas turbine inherently requires builtrin automatic control systems to protect it during ~tartingand operation, since manual operation cannot respond fast enough to meet the requirements. The extension of the control system to provide fully automatic control of all systems, including auxiliaries, is therefore quite simple and is commonly provided on gas turbine installations of all sizes and types. The relative simplicity of the gas turbine has enabled it to attain outstanding records of reliability and maintainability when used for aircraft propulsion and in industrial service. The same level of reliability and maintainability can be expected in marine service if the unit is properly applied and installed.
213
Marine units derived from aircraft engines usually have the gas generator section, comprising the compressor and its turbine, arranged to be removed and replaced as a unit. Maintenance on the power turbine, which usually has the smallest part of the total maintenance requirements, is performed aboard ship. Because of their light weight, small gas turbines used for auxiliary power, or the propulsion of small boats, can also be readily removed for maintenance. Units designed specifically for marine use and those derived from industrial gas turbines are usually designed for maintenance and overhaul in place. Since they are somewhat larger and heavier than the aviation-type units, removal and replacement are not as readily accomplished. For this reason, they usually have split casings and other provisions for easy access and maintenance. The work ca.n be performed by the usual ship repair forces.
Section 2 An'angernent and Structural Details Genemi Amngemeni* In addition to the classificatiOn dependfig upon the choice of cycle, gas turbines can be acmrdhg to the mechanical arrangement. speaking, f 3 S turbines are either "single-shaft" units in which the compressor and turbine are attached to a single shaft, which in turn is connected to the load; or multishaft units in which the twbine~and sometimes the compressor, is divided into sections, eachwith its own shaft which can run at different and Weeds. Each comPreSmr must be driven by a section of the turbine, and the load can be driven by One of thesections or by its own independent turbine. Where the Compressor is divided into two successive and sections with similar coaxial turbines in series, it is a "twin-s~ool" machine. When a flingle ComPreSsor is used and the turbine is divided into t~~ One which drives the compressor and the other the load at speeds, it is known as a "two-shaft " machine. Both single-shaft and two-shaft machines can be used in service- The single-shaft units are most oOmmonly used for generator drives, either a-c or d-C. the a-c generator units are used only as auxiliary power units where they run a t constant fn3q~encyand 'peed. ~ropulsionunits, where the propeller must Operate Over a very wide speed range, a controllable and reversible-~it~hpropeller or some eq~ivalent varisble-s~eed transmiasion, such aS an electric drive, be used with a single-shaft machine because of its limited range and poor acceleration characteristics. For main ~ ~ o ~ u l s units, i o n a multishaft machine is 'lormally used) the arrangement bekg a two-shaft unit with an variable-speed power turbine.
-*
With this arrangement the power turbine and propeller can be stopped if neceymry, and the gas producer kept in operation for rapid load pickup. The use of variablearea nozzles on the power turbine increases fledbility by enabling the compressor to be &tained at or nw full speed and air flow, down to low-power turbine speeh. This makes nearly full power available by adding fuel without the necessity of waiting for the to accelerate and increase the air flow. where low-load economy is of importance, the controls be arranged to reduce the compressor speed a t low loah the maximum turbine inlet and/or exhaust temperature for best efficiency. Since a gas turbine inherently has a poor part-load fuel rate performance, this variablearea nozzle feature can be very advantageous. ~i~~~~ 8 illustrates a typical comparative of part-loa,j performance curves for regenerative and simplecycle units. The dotted curve indicates the min possible a t reduced load by the use of a variablearea nozzle control for a regenerative cycle. The phisiml a m g e m e n t of the variouscomponents (i.e., compressors, combustion systems, and turbines) that make up the gas turbine is influencedby the thermQ dynamic factors (i.e., the turbine connected to a cornpressor must develop enough power to drive it), by mechanical considemtions (i.e., shafts must have adequate bearings, seals, etc.), and also by the neoessity to conduct the very high ah and gas flows to and from the various components with minimum pressure lossas. 2.2 Air Inlet Armngemenk. ~h~ provisionsfor handling the large inlet and exhaust volume flows are particularly important. Not only must the ductwork to and from the unit be accommodated within the con-
2 14
MARINE ENGINEERING
PERCENT
SHAFT
HORSEPOWER
a.large plenum can be used, the compressor inlet can be inserted in one wall of the plenum so that the air flows axially into the compressor annulus. Model tests of the inlet configuration are often conducted to identify undesirable flow distortions and resonant conditions. If tests are not conducted, a good rule of thumb is that the engine inlet should be a t least two engine diameters away from the bullhead or far side of the plenum. An arrangement which is very satisfactory, both aerodynamically and structurally, is show11 in Fig. 9. Air enters the passage radially from the inlet hood, which is large enough-that is, has low enough velocities-to insure equal distribution and essentially uniform radial velocities around the circumference. The air is then turned in a n axial direction, and accelerated by the reduction in passage area, to the plane of the inlet guide vanes. The squeeze and resulting acceleration suppresg vortices and smooth the flow a t the inlet guide vanes. Struts tie the two sides of the casing together. 2.3 Exhaust Collectors or Hoods. Due to the higher tmnneratures. the exhaust volume flow is larger than that ----- Iof the inlet. Maintaining velocities that provido reasonable pressure drops is, therefore, somewhat morc! difficult. The high temperature also introduces expansion problems since the movement of the ducting due to thermal growth must be accommodated without introducing high forces into the turbine structure. Tlai allowable forces and moments depend somewhat up011 the size of the unit and must be specified for the partiaular gas turbine used. The exhaust collector, or exhaust hood, is that part ol' the turbine casing that collects the gases leaving the! last-stage turbine wheel and conducts them to tllo connection with the heat-recovery equipment or exhaunC duct. It' usually, however, serves as more than n collector. Due to the limited annulus area that can bo provided a t the normal gas turbine exhaust temperaturcn,
GAS TURBINES
215 ACCESSORY DRIVE GEARS
POWER TURBINE
--
Fig. 8
Gas turbine part-load performance characteristics
fines of the ship structure, but the total system pressure drop must be held to a minimum. I n addition, the actual configuration of the air inlet to the compressor is important. Unequal air flow into the annulus or flow into the annulus a t varying -- . inlet angles around the circumference (velocity and pressure distortion) can reduce the efficiency by the effect on the firstrrow entrance angles. Also they can cause blade vibrations that can lead to early blade failure. Ideally the compressor inlet should pull from an infinite plenum, as with an aircraft jet engine in flight. Practically, the engine must pull the air through a duct system; therefore, some form of air inlet housing or hood must be used. If
Fig. 9
Large axial-flow compressor and turbine, regenerative-cycle,two-shaft industrial design
'
AIR INLET ASSEMBLY
Fig. 10 Axial-Row tompressor and turbine, simple-cycle, two-haft design
the axial velocity and the corresponding absolute velocity leaving the last-stage turbine blading is usually high, 400 fps or more. The energy represented by this velocity, known as the leaving loss, is a rather high percentage of the turbine energy, and so for good shaft efficiency a s much of it must be recovered as possible. To recover this velocity energy, some form of diffuser is employed. Since a good diffuser takes up considerable space, a judicious compromise usually must be made between the space used and the energy that can be recovered by the pressure rise in the diffuser. A typical straight diffuser is shown in Fig. 9, while a curved diffuser is shown in Fig. 10. Most authorities agree that good diffusion cannot be obtained in a turn, so that the diffuser efficiency of curved diffusers like Fig. 10 is probably low. However, they obviously take considerably less axial length, which usually is an important advantage. I n addition*to its aerodynamic function as a diffuser and collector, the exhaust hood must frequently act a s a structural member and carry loads and bending moments through the gas path. This is the case of the arrangement shown in Fig. 9 where the power turbine bearing
housing is supported by struts through the gas path. I n some cases, particularly with single-shaft machines, the rotor bearings may not require support through the exhaust hood; but the inner wall of the diffuser usually must be supported by struts through the gas path. Such struts should be located as far down the diffuser as possible, as any obstruction in a diffuser, even when streamlined, markedly reduces its recovery efficiency ; therefore, the lower the velocity region a t which this occurs, the better. I n addition, provision must be made for the thermal expansion (particularly during starting) of the struts, which are completely immersed in the hot gas stream. I n small units, distortion of the casings a t the attachment points may be sufficient; but in large units some radial flexibility is generally allowed. Tangential struts or a radii1 strut with a tangential spring member a t the outer end are possible arrangements. Air-cooled struts attached to cylindrical inner and outer members that can distort sufficiently to allow for some expansion can also be used. The exhaust hood, or collector, is usually fabricated from relatively thin material with suitable stiffening ribs formed in the material or welded on externally. This
216
MARINE ENGINEERING
serves to stiffen the structure to prevent drumming or resonances and also enables the structure to withstand the internal pressure due t o the back pressure imposed on the unit when heat-recovery equipment is used. Large flat surfaces should be designed to withstand a t least 20 in. of water without bulging. Materials for exhaust hoods are frequently stainless steels, particularly for units with high exhaust temperatures. For large units, where thicker stock is required to give adequate stiffness, carbon or low-alloy steels can be used. In any case the exhaust system is usually covered with some form of insulation (blanket, block, plastic, etc.) to reduce the temperature of the exposed surface. 2.4 Structural Arrangements. Structurally, the stationary gas turbine parts, compressor casings, comburc tion casings, turbine casings or shells, and the related structural supports for the mtor bearings must withstand not only the internal pressure forces but also the external forces imposed on the unit from its own weight and the reactions from engine torque and external connections. The casings must be designed for the internal pressure forces and must also be checked as a beam, under the reactions due to the weight of the components, plus whatever "g" loading may be imposed. Rotor and stator weights must be considered, and the supports are frequently located so as to mini& the bending moments in the structure. The calculation of the bearing housings and supports cannot be based upon the weights of the rotors alone. To insure the integrity of the unit in the event of a blade or bucket failure, they should be able to carry the centrifugal loads imposed by the loss of some credible combination of bucket or blades within the tensile strength of the members. The loss of two adjacent vane sections or one complete bucket and dovetail is considered a reasonable assumption. Supports for gas turbines can take many forms. They must support the unit and maintain it in line with the driven equipment, while allowing for the axial and radial thermal growth of the unit from cold to normal operating temperatures. Several methods of support are shown in the various illustrations, and it will be seen that small units frequently use three points of support with one centering key or gib, while larger units usually use a t least four supports; more are used if the whole unit is divided into several casings, as in compound cycle units. 2.5 Mounting in the Hull. In marine applications, the gas turbine usually cannot be mounted rigidly to the ship's structure. Normal movement and distortions of the hull when underway would cause distortions and misalignment in the turbine. This could cause internal . rubs or even bearing or structural failure. The turbine components can be mounted on a subbase, as shown in Fig. 9, which is built up of structural sections of sufficient rigidity to maintain the gas turbine alignment when properly supported by the ship's hull. I n cases where aircraft gas turbines have been adapted to marine use, some form of tubular structure may be used, but the
purpose is the same. A rigid structural subbase also provides a convenient mounting for many of the gas turbine auxiliaries, particularly the lube oil tank and other components of the lubrication system. When properly applied, a three-point support of such a s u b base will prevent ships' structural distortions from rnisaliiing the various components. 2.6 Regenerators and Recuperators. The recovery of heat from the gas turbine exhaust and its return to the cycle to improve the overall efficiency are accomplished with a regenerator or recuperator. Both terms are used more or less interchangeably although the term "regenerator" is sometimes considered to be limited to rotary heat exchangers in which a heat storage matrix is alternately exposed to the hot exhaust gases, and then to the compressor discharge air stream, transferring heat from the former to the latter. The term recuperator is then reserved for fixed-surface heat exhangers in which the hot exhaust is on one side of a wall and the air on the other, the heat being transferred through the wall by conduction. Both types have been used successfully with gas turbines although the fixed-surface type is far more common. Considerable effort, however, has gone into the development of various forms of rotary regenerators for small engines because of their advantages of small size, light weight, and high effectiveness. On the other hand, work on fixed-surface types has also resulted in reduced weight and space, so that both types arc competitive. Rotary regenerators have been built with the heatexchange element (or matrix) either in the form of a flat disk or as a hollow cylinder. The choice of one or the other is determined primarily by the geometry of tho installation as related to the gas turbine components, although considerations of the seal design between the cold high-pressure chamber and the hot low-pressure chamber can affect the choice. Seal leakage and the "letdown" which occurs when a section of the rotary matrix passes from the high-pressuru region to the exhaust or low-pressure region offset tho high effectiveness that can be realized in the matrix, a, that the overall cycle efficiency is comparable to fixedsurface types. The matrix itself can be metal or ceramic, in the form of wire, strip, plates, pebbles, etc. The influence of seal leakage, let down, and pressure drop* and their relations to the geometry of the regenerator arc! given in references [13] and [12]. Fixed-surface recuperators were originally of convontional shell and tube construction, in some cases wit11 in. dia when used for mobilo tubes as small as applications where weight and space were important. More common constructions with tubes M to 1 in. diu were large and bulky, and considerable trouble wan experienced in some cases with cracking of the tubn sheets due to thermal stresses set up between the cold shell and the hot tubes across the outer periphery of thla tube sheet. Present designs are usually of the "plate-fin" can-
GAS TURBINES
struction as illustrated in Fig. 11. Here the compressor discharge air is between adjacent plates with the turbine exhaust gas in the finned passages. The pressure load is then carried by the fins acting as columns and transmitted to the outer casing where it is restricted by "strongbacks" and ~iamied into the end structure. Details of course vary between different manufacturers; a typical arrangement and its development are described in references [13, 141. The materials used in fixed-surface heat exchangers depend upon the temperature range in which they must work. Where the maximum turbine exhaust tempera-
I
SS
i
8
f
&
CORRUGATED CENTERS
used, but the metal thicknems chosen should provide an adequate allowance for minor corrosion. For turbine exhaust temperatures over 1000 F. or where the. design has been optimized for minimum weight with resulting the astern rotor blade sections to reduce the rotatioh thin-gage materials, corrosion-resistant materials such 10-s -----. as stainless steel or one of the Inconels are necessary. Lacking an internal reversing method, marine gas 247 Reduction Gearing and Reversing Consideraturbine installations must be reversed by an external tions. The gas turbine is a high-speed qachine with output shaft speeds ranging from about 3600 rpm for means. Electric drives offer ready reversing but are large machines up to 100,000rpmfor very small machines. usually ruled out on the basis of weight, cost, and to some extent efficiency, except for special applications Approximately 25,000 rpm is an upper limit for units (Chapter 10). From a practical standpoint there are suitable for the propulsion of small boats. With these two alternatives, a reversing gear or a controllable and output speeds, a reduction gear is necessary to reduce the reversible pitch (CRP) propeller. Both have been used speed to the range suitable for a propeller. Smaller successfully in gas-turbine-driven ships. Reversing units suitable for boats or driving auxiliary unit&, such as gears haye been commonly applied to diesel-propelled generators in larger vessels, frequently have a reduction ships up $0 several thousand horsepower and have also gear built integral with the unit. Larger units n~rmally been used in some gas turbine applications [I, 16, 17, 181. require a separate reduction gear, usually of the double- CRP propellers likewise are quite common in smaller or triple-reduction type. sizes, and are finding increasing applications in higherThe gearing itself can be of any arrangement. S q l l e r horsepower qliesel and gas turbbe ships [19, 20, 211. units with built-in gears frequently use a or One important consideration in choosing a reversing star gear arrangement. Larger units use double-helical means for a gas turbine is whether it is a aingleshaft or gears. Any of the gear types and arrangements d e scribed in Chapter 9 can be used with a gas turbine to two-shaft machine. The singleshaft gas turbine has a very limited speed range, in some cases only from 75 to suit the rating, speed ratio required, and arrangement of 100 percent speed, while the load turbine of a two-shaft the machinery in the vessel. machine can be stopped by the application of sufficient A gas turbine, in common with all turbine machinery, while the gas generator continues to run. This is not inherently reversible. Steam turbines can provide torque means that a direct-geared, single-shaft gas turbine can separate reversing elements built into the LP casing, but only be applied in conjunction with a CRP propeller, in this is not practical in a gas turbine as the rotation loss order to be able to properly maneuver the ship, unless it is of the astern elements rotating in the ahead direction a t applied only as a boost engine where its limited speed atmospheric pressure would be very high. The resulting range matches the requirements. A two-shaft -chinc temperatures and losses would be unacceptable. can be applied equally well with a reversing gear or a Design studies have been made of a unique arrangement in which the astern bucket's vane section takes the CRP propeller. A reversing gear also provides a means place of the shank of a long-shank bucket and which of disconnecting the shafting from the turbine. Such a means should be provided if a single-shaft machine is has its own variable-angle nozzle that can be closed used so that it can be started and checked out without off when going ahead [15]. This arrangement shields rotating the propeller and shafting. 0--
t
217
I
21 8
GAS TURBINES
MARINE ENGINEERING
Sectiun 3 Accessories ~
~
turbine, pumps ~ i andl Drbes, i ~ The~gas ~
,kle bsially a complete, self-contained power plant, for its operation. Fuel requires cerbin pumps (on liquid-fueled units) and lube-oil pumps are always needed. A positive fuel pressure must be supplied to the engine during all operating conditions, including sta+up, These pumps can be independently driven by electric motorsl but are ususlly driven through a reduction gar from the m i n turbine shaft. The takes many forms; spur, bevel, worm, a,essorr or spiral gearing has been depending upon the configuration of the turbine and, in some cases, installation requirements in the form of length or space limitaThe Bimplest gearing amrngement that will drive tions. (these include governors, the required tachometer generators, speed switches, etc., as well as the fuel and lube oil pumps) is usually best. In any case, the garing must be designed for the duty and l i e ,quired of the min unit. They are usually provided by the gas turbine manufacturer as standard engine 3.2 Statiing ~ ~ , ,A igas ~ turbine, ~ ~ . like other inter,l com~ustionengines, is not self-starting, and
Since the starting device is normally required Only to about 50 percent speed, it is usually connected to the turbine through some form of special clutch which be disconnected during normal Operation. The it simplest, and probably most satisfactory, form Of is a Simple jaw clutch, magnetically Or ~~~~~~~~~~~~y be engaged and Spring disengaged. Provisions made for rotating the starter slowly during engagement to make sure the jaws are fully engaged before is applied; otherwise, severe damage Alternatively, Some form of overrunning can be used, but Unless the clutch is of some slf-s~nchmnizing form such as illustrated in Chapter 91 it is susceptible to damage or failure if the starting device is energized load the gas turbine is still decelerating. The under such conditions can be very severe, and few Overto running Clutch de&ns have the torque disengage stand such Shock. The clutch should Operating completely a t Some speed below the speed of the turbine shaft to avoid excessive wear On mechanism. 3.3 Inlet Air Filters. Additional items and considered as accessories are inlet air filters and exhaust SilenCerS. Pressure drops are Of -jar imporand exhaust tance in gas turbine operation; therefore1 equipment mist be designed with for the economic balance of size and Pressure Gas turbines require clean air, as compressor will eventually become coated with a layer
external means mustbe utilized to bring it up to the selfsustainingspeed. ~ h i ais the speed from which the rotors can be accelerated by the ad&tion of fuel alone, asdstance, and it is usually about 30 to lvithout 50 percentof the gas-producer's full speed. On units sbftaonly one starting aith two or more device can be used, although in certain industrial-type gas turbines each shaft utilizes its own s*ting device. Starting devices in common US include electric motors, stearn turbines, air motors operating on stored comprewd air, and small diesel engines, which must, in turn, have their own starting systems. Other starting systems include hydraulic motors fed from x%h-pressure pumps or accumulator systems, and special r 0 t a r y - t ~ ~ ~ starting motors fed from high-pressure air supplies, sorne EXHAUST of which include the combustion of fuel to furnish energy. For large units a ~lnallindependent gas turbine m y be used to furnish the starting power. In any case, it is important that the starting device have adequate power to bring the Unit to the selfsustaining speed without requiring the addition of energy from the main fuel supply sufficient to cause excessive temperatures at the turbine inlet. That is, the fuel/air ratio during the shrting cycle should be held close to normal limits. This requires a relatively large energy input from the starting device. Since the operating time is of relatively short duration, the starting device a n be highly overloaded or peak-load rated, particularly if it is an electric motor. Starting times range from seconds on a very small gas turbine to 1 to 2 minutes On large aircraft-type engines and 15 to 20 minutes on the ~ i12~ Gas . turbine installation on P G boats showing air inlet and exhaul) arrangements largest industrial-type units.
dirt, which reduces its capacity and efficiency and sound power level of a given design of gas turbine is a results in a degadation of the entire engine or may even function of its size or power and is approximately b u s a ComPressor blade failure due to stall. TO insure proportional to its rating. ~~~tof the noise is generated air, most, n ~ n m r i n egas turbine installations are aerodynamicallyand is related to blade passing frequency equipped with an air filter or clmer. In marine and, therefore, is in the high-frequency range (221. a~~limtionsi however, the nI0st important requirement The m&jor sources of the noise radiated to the suris to keep particles and water, n'hether in drops or as roundings are the inlet openings, exhaust openings, and wsterl from entering the compre%~Or. For this gears. However, the entire machine radiates noise; the reanon air should be placed as high above the water sound intensity is related to the easing thickness or, more as possible and must be equipped with effectivebaffles or exactly, to the casing mss. Noise radiated from the eliminators to Prevent the entry of solid water. Behind casings is usually confined to the engine room, and its the eliminators a demister should be installed to intercept effect h n be reduced and controlled by appropriate water droplets. The demister can consist of an inertial sound treatment. ~ p a * ~or,r alternatively, of pads (similar to filter The airborne noise, in both the inlet and exhaust, can pads) of metal or synthetic fibers of controlled size and be attenuated to almost any required level by the use 'pacing to the size of droplets passed. of suitable silencers. I n general, the greater the decibel If the demister pads are lvet, even with Sea~~ater,reduction in noise level required, the more expensive the are also effective in stopping the ingestion of salt silencer and the greater its pressure drop, so a noise and. other foreign particles and thereby serve as a filter reduction greater than n e c e w v should notbe used. medium' A typieal arrangement having an Specifying silencer performance, it is important that the inertial-t~~e separator is shown in Fig. 12. The type of sound-pressurelevel a t the turbine be given in each of the under consideration and its above-water profile and octave bands and that the reduction to give the required height above the waterline will dictate in large part the decibels at a predetermined radius be also specified for tYpe of mist eliminator to be used. the same octave bands. If the engine room is used as a plenum, oil Vapors from Since the sound attenuation in the surroundings Other can also adversely affect engine per- be ,yomewhat directional, the configuration of the inlet formance. and exhaust openings and their orientation should be 3A Inletand Exhaust The gas turbine, carefully chosen. Of course the sound levels required being a -chine, generates a relatively large also depend upon the service (e.g., eargo, amount of noise of a w*de frequency spectrum. The passenger, or naval). Of
Section 4 C0ntr0ls The control system of a gas turbine perform several functions that are vital to its It must control the speed of the shafts that make up the *m~lete unit, schedule the fuel flow during starting and qther transient mnditions, prevent Overtemperatures in the combustion and turbine system, and prevent a dsngerous overspeed under any conditions. The system to perform these functions is made up Of a number separate devices corn'bed into various Systems; but the trend is for all the to be inteerated into a single system, usually Of the type, that controls all the operations of the unit. 4*2 'peed Measunhent and Control. A nonintegrated system will mnsid of a speed governor, Usuall~ the or flyweight type, which through a relay system 6-e., a pilot valve and piston) Operates the main fuel valve to regulate the fuel flow to the Power output and the speed Of the unit- For a single-shaft mchkre, this is that is neces~rY1and a governor unit,
such as shown in l?k. 13, will meet all The governor characteristics may be of the isochronous or droop type as required. For machines nith two or more shafts, additional control usually must be provided with a twolhaft turbine having fixed nozzle areas, the speed of the gas producer section is normally controlled by one speed governor and the output is a function primarily of that speed. The power turbine speed, however, is independent and is a function of the power turbine and its loading characteristics (i,e,, horsepower or torque vs. rpm relationship of the propeller). T~ prevent overspeeding on loss of load, a supplemental speed governor, wmetimes called a "topping governor,?, mustbe driven by the power turbine which will override the min governor and reduce the fuel supply should the power turbine speed exceed the topping governor ~ ~ ~ t n- ~ ~ i ~ When a t ~ o - ~ h gas ~ f t turbine is equipped wit. variable-area power turbine nozzles, an additional control element is introduced which gives additional flexibility in the control and charactefistics of the unit.
GAS TURBINES
MARINE ENGINEERING
Not Only is the poser turbine speed independent of the gas producer speed, but the speed can be contfolled over A a wide range of load and load the hydraulic relay system is normally used to load turbine nozzle and thus setthe energy distribution betNreenthe power turbine and the gas producer turbine. The control function an be set up in several ways; normally however, the gas producer speed, through its the nozzle position to maintain the set speed and corresponding air flow. The power turbine speed, through a separate governing system, sets the adjusts the outputto match the speed fuel ,flow and This, provides the maximum laad response, the ah flow constant, the load can be changed since ,pidly by merely varying the .fuel flow without risk of overtemperature or compressor instability. Alternais desired and tively, maximum parbload slower load pickup be accepted, the controls a n be if
an
arranged to vary the air flow with load by varying the gas producer speed. This makes it possible to maintail' the maximum turbine inlet temperature Over a wid' range of load within limits set by the exhaust temperature control. liar m u l t i ~ l ~ units - ~ ~the a ~same ~ princiby the need ples apply, but the problem is to control more shafts and1in some a s e s l more than On' combustion system (each having independent temperatures). A description of such 'ystems 's given references [231 241. the Instead of a governor Nith Or signal can be taken from a tachometer converted to a magnetic speed Pickup, and the and modulated Signal which an corresponding speed and output. Such systems an' particularly advantageous for generating sets where load signal can be fed into the electronic 'ystem to givs regu1ation as '* load anticipation and thereby
'''
'
MARINE ENGINEERING
difficult to meet, particularly when pumping low lubricity fuels such as JP-4 and especially JP-5. Only pumps specially designed for the service will satisfactorily and reliably meet the requirements. One additional requirement of a fuel control system is to divide the fuel evenly when there is more than one fuel nozzle or combustion chamber. This is particularly important with multiple combustion chambers, since even individual nozzle arcs can only reliably meter the air flow to within 3 to 4-percent accuracy, so that a 5-percent difference in the fuel nozzle flow can result in an 8 to 9-percent variation in temperature rise. On a 1000 deg F rise this variation is 80 to 90 deg P, an amount large enough to affect the life of the hot-gas-path One way of assuring essentially equal flows is to use a piston-type pump with one cylinder delivering flow to each fuel nozzle. This gives a pulsating flow to each nozzle, but provides an equal distribution between them. An alternative method is to meter the flow to each nozzle through a small gear pump element with all elements being moupted on a common shaft and thus running a t the same speed. The speed-control governor then varies the fuel flow from the shaft-driven fuel Pump- The speed of the gear elements is pmportional to the fuel flow1 and each element passes the same fuel flow regardless of pressure variations due to differences in the individual
The control of the fuel flow during the startiIrg cycle is an important function of the fuel control system to temperatures at prevent the development of low speeds. ~~~~l~~ controls with means of metering the fuel during the lighboff and starting cycle are usually required. These controls can incorporate a bias from the compressor discharge pressure, which is a
convenient measure of air flow, to set the fuel schedule during acceleration, and in some applications a temperature override from the temperature control system. The latter, if the system has adequate response times, approaches an ideal way to set the fuel flow, since it responds to the most important condition-temperature. Maximum fuel flow and minimum fuel flow stops are also normally incorporated in the control to prevent overloading the unit under low ambient conditions and flameout due to cutback of the fuel flow on sudden reductions in load. 4.6 Overspeed Protection. The governor and speed control normally control the shaft speeds within prescribed limits. However, as with most turbine machinery, a backup to prevent dangerous overspeeding must be ~rovidedin the form of an overspeed shutdown. Such a device should be applied to every shaft. Upon reaching a speed of about 110 percent of rated rpm, the overspeed trip mechanism shuts off all fuel to the unit. In the case of two-shaft units, this is in addition to a pre-emergency or topping governor driven from the load turbine, \vhich acts to limit fuel flow to that corresponding to about 10a-percent speed. ~ h overspeed , trip can be mechanical in the form of a centrifugal mechanism which, upon reaching the trip speed, dumps oil from the fuel control valves or opens a set of electrical contacts that similarly shuts off all fuel flow. Alter~lativelya tachometer geneator or magnetic speed pi~liupscan be used to generate an electrical signal which interrupts the control circuits and shuts off the fuel as the shaft reaches the set overspeed. Electrical devices must be arranged to "fail safe" so that Once Opens the started, a Zero speed signal or lack of signal control circuits and shuts off
Section 5 Centrifugal Compressor Design
GAS TURBINES
373 ---
imparts kinetic energy to the fluid, the diffuser and scroll, which are the major stationary parts, must convert the kinetic energy into potential energy in the form of a static pressure rise. Therefore, the higher the pressure ratio required, the more important is the design and corresponding performance of the diffuser. The forward section of the impeller, usually called the inducer, may be separate from or integral with the rest of the impeller. I n any case, however, the aerodynamic shape of the inducer is very ifiportant insofar as the overall efficiency, capacity, pressure rise, and surge performance of the machine are concerned. 5.2 Centrifugal Impeller Design. To achieve good performance and high efficiency, the detailed aerodynamic design of the imfieller (rotor) and diiuser is necessarily based on experience and usually a long development period. Many approaches have been taken in designing centrifugal compressor impellers. The usual method employs a combination of velocity
I
Various detail factors are of importance in the final design. Shrouded impellers are sometimes used where Fig. 14 Centrifugal compreswr and radial-flow turbine, simple-cycle, maximum efficiency is important, and stress conditions single-shaft unit with single combustor and gearbox will allow them, although quoted gains of 3 to 5 points in efficiencyare debatable. Axial thrust generated by the rotor can be minimized by balance holes to the back face of the impeller or abscissa may be expressed as radial vanes or blades on the back face to create an opposing pressure gradient. The thrust of double flow, i.e., double entry, impellers is inherently balanced. The use of pre-whirl, with rotation, will reduce the inlet relative Mach number, but also reduce the work input for a given tip speed. However, higher tip speeds W = weight flow, lb/sec are then permissible, and so higher pressure ratios can be P = inlet pressure, psia obtained in this manner. I Ti = inlet temperature, deg R 5.3 Diffuser Design. As stated earlier, the diffuser A plot using a similar flow parameter is shown in Fig. 15. used with a centrifugal compresmr plays an important A commonly used is &/ND31which is essentially part in the determination of the overall performance and and therefore applies to all geometrically efficiencyof the compressor. However, the design of the diffusing section of the machine is usually dictated by One important characteristic of all C O ~ ~ ~ ~ U O U Sspace - f l o wand configuration limitations as well as by aerocomPressors is the surge line. This is the line represent- dynamic considerations. ing the between Pressure rise and volume flow Diffusers can be classified as annular, vaned above which is unstable. Under these con- annular, or scroll. These fundamental types, in additions the air flow surges or pulsates, often with dition to being used separately, may be used also in destructive effects on the unit. This limit usually must combination; that is, an annular diffuser may have a be determined test and is nearly always shown on the considerable vaneless section preceding or following a plot of the unit's characteristic performance curve. vaned section, and a scroll may be used as a secondary the is simple in mn- diffuser after either of the annular types, or the cornstructionl usually comprising only a single rotor in a bination. casing, the achievement of high Pressure ratios and good Scroll diffusers are used, either alone or in efficiencies is Since the rotor Or impeller with radial diffusers, where the air leaving the
*
GAS TURBINES
-Ua P, Pa
-
* ROTOR TIP
0
SPEED, FPS
ROTOR INLET AREA,
IN.^
INLET PRESSURE. PSIA
I
DISCHARGE PRESSURE. P
I
1
RADIUS, IN. fig.
IS Perfonnonce mop f a centrifugal canprerur Fig. 16 Turbine wheel rtresaes
The oentrifugal force of the Vanes is introduced a t Shrouded impellers a n also be oalcuhted by an exten'pecific of the disk, which is divided into sion of the method. In this case the shroud is cctlcula&d ringa and strips. Thermal stresses due to a radial as an unsupported ring and then influencecoefficienbare tem~erat&e gradient are computed by assigning a calculated for the interaction between the blades and each ring, based on tests or calculations. the shroud and disk to arrive at the final stress distribuA typiml plot of wheel stresses is shown in Fig. 16. tion.
Section 6
GAS TUR'BINES
MARINE ENGINEERING A-
\
P'
VANE
VANE
(a) Dovetail root
(b) Cylindrical raot
WHEEL
Fig. 18
F ; ~ .17
cascade performance of the individual blade rows. That dkgramfor a specific blade row (rotor or is, the stator) gives the entrancecondition to the row; and the leaving velocities and data give the ,gles, which combined with the e& velocity diagram give the absolute leaving conditions for the row. An important consideation in the design of axial blading is that of blade loading. %rly data on low-speed caaades [291were extended by later NACA progams, including information on blade at higher ~~~h numbers and higher blade,tting angles [3*34. From the early data, Lieblein and others developed a digusion parameter that is as a blade loading limit [331. frequently
Compressw rotor blade attachments
performance map for axial-flow WPr-
6.3 Blading Design. The both and a common stator, are always of C-4 Or C-7' Thu being NACA-65 section or the at that length is set to give the deskn lof3ltion. For manufacturing reasons blades (rotor) a e often of constant tip diameter, which permits a simple casing design and 'lso maintains a high tip speed and minimum number of stages' The 'Oat a minimum at diameter comes~ondinglyvaries with inlet to a xim mum at the disehargel with seveml constant taper, as in Fig- lo,and of the high-pressure stages having a constant rOO'' diameter, as in Fig. 9. The latter tends to reduce the leaving velocity a t the discharge, and the accom-
panying leaving loss. Com~ressor~ are also designed with a constant root diameter and varying tip dlmeter, and in other cases both diameters are varied. The thiclcness of the airfoil sections is chosen to meet vibration criteria while maintaining good aerodynamic performance. he blade chord must then be sufficient to maintain the blade bending stresses at an acceptable level to insure freedom from fatigue failures. Sample are commonly made up and their natural frequencies checked for all the lower modes and, if necessary, t ~ ~ n ebyd changing the thickness to avoid resonances a t running speeds as shown by a Campbell diagram (see Fig. 30 of Chapter 5 for a typical Campbell dimgram). Even this cannot give positive assurance that all resonances are avoided, particularly in variable speed units. Even constant-speed units must pass through resonant speeds when starting up and shutting down. he allowable bending stress chosen depends upon the Inaterial and the designer's experience; but a good rule for deel is 10,000 psi in the outermost fiber based on the root section modulus, or 25 Percent of the fatigue limit at lo8 cycles- The first- and last-shge design stresses are frequently reduced to a b u t one half of these values to allow for the increased loading that occurs a t off-design conditions. Much higher stresses are used in aircrafttype units. Titanium blades are frequently used in gas turbines of the aircraft type. The blade must be attached to the rotor by some means; this is accomplished by an enlarged section, or root, in the form of a dovetail, as in liig. 18(a) cylindrical roots as in Fig. 18(b)*and bolted attachments have been used. The dovetail section must be made and is usually a good fit in the slot in the rotor; axial slots are commonly broached but circumferential grooves are machined in a lathe.
The stresses in the dovetail are shearing stresses (across section A-A in li'ig. 18) and crushing or compressive stresses (on surf ces B and B). The total load on the dovetail is determined by adding the centrifugal force due to the vane and that of the dovetail itself. The shear and crushing stresses are then calculated from this total load and the geometrimlly determined areas. Wheel dovetail neck shear stresses a t maximum speed should be less than 30 percent of the minimum yield stress and the tensile stresses less than 20 percent of the minimum ultimate tensile strength. In the layout of the vane sections, the sections a t each radius are usually stacked so that the line through the centroids of the sections is a straight line, either radial or with a slight inclination to compensate for the aerodynamic bending moments. 6.4 Rotor Design. Most axial compressors have the rotor made up of a series of individual disks, each usually carrying one row of rotor blades. The individual disks are held together by a central bolt as in Fig. 10 or a row of bolts a t an intermediate radius as in li'ig. g. In either case, the Golts are prestressed a t assembly so that the total bolt load is sufficient to keep the disks from separating under the highest bending moment that is likely to be imposed on the rotor. The individual disks are usually to give an approximately constant radial stress in the wheel. Various methods may be used for the calculation of the wheel stresses, including the centrifugal loading of the blades, but the one most commonly used is the "Manson method" [27, 281. The average tangential stress is of particular significance, since, for the ductile materials usually chosen, it is generally considered that the *heel bursting speed is reached wrhen this stress exceeds a value in the range of 85 to 100 percent of the ultimate
MARINE ENGINEERING
229
GAS TURBINES
of freedom from distortion of the casing, due to its symmetry,'so that closer tip clearances over the blading can be maintained in operation. I t is also somewhat lighter. Larger units commonly have the compressor casing split on the horizontal centerline, each half containing half the stator blades. The two halves are bolted together with a horizontal flange as shown in Fig. 9. The horizontal flanges must be designed to transmit the circumferential hoop stress, due to the internal pressure, across the joint without leaking. The joint itself is always made with a metal-to-metal contact, the joint
being carefully finished to avoid leakage. The best proportions for the flange and bolting are usually based on experiment as well as a stress analysis [34]. 6.6
Other Compressor Types and Combinations.
Other types of compressors have, from time to time, been proposed for application to a gas turbine cycle. Of the several possible types, only a variation of the centrifugal type (the so-called "mixed-flow" compressor) and a combination of one or more axial stages followed by a single centrifugal stage (for small volume flow, high pressure ratio machines) have had practical application.
Section 7 Turbine Design and Construction
of the msterbl. Typiml stress distribution separate blade rings by T-slot attachments, Or brazed the rings. The rings are made in two Or more curves for compressor wheel are given in Fig. 19. into segments which are then slid into paves machined hi^ is a wheel ,with a central hole, and the tangential the casingsstress a t the bore is quite high, approaching twice the The inner ends of the stator blades can be either center stress in a solid disk. lo, Or unc~~~~~~~~ have been built with drum rotors, similar shrouded, as in the forward stages of into If shmuded, they are riveted Or shrouded. to some steam turbine rotors, and solid rotors have been applications are rare. S t r e r ~ s the shroud .ring, which carries a labyrinth packing to designed, although in drum rotors are usually based on unsupported ring minimize e a hfinished 3 along to theanrotor. If unsh*udedl the blade the endsL are accurate diameter to theory, and are appreciably higher than for individual at give the minimum allowable running disks for the same peripheral speeds. Consequently they are used only in compressors with low blade speeds and a blade tipsSome compreaior designs U S variableangle large number of stages for a given overall pressure ratio. blades in one or more rows so that the blade angles call Such designs can be classed as very conservative. 6.5 ibtor ~ ~Statori blades~ are not ~ subject . be adjusted to match the operating conditions, partichavn that rotor blades experience; ularly during starting. In such eases, the blades to the centrifu&al thrOug" therefore, theb mechanical design and aerodynamic cylindrical shanks that extend radially outward to the design are not as limited. Nevertheless, bending holes in the casing. Levers are attached in On'' ends, which are linked together, so that loadings must be calculated stressesdue to rows can be adjusted simultaneously by a" or more and kept, to consenrative values to allow for unknown stresses. Blade vibration modes and fre- external control mechanism. tho The compressor easing must be designed to for a preliminary design, and quencies must, be stator blades or stator blade rings and contain tho checked for sample blades, so as to avoid lower-order various stages. Small0r rewasnoes within the running speed range. pressures developed in the a casil'g, machines are frequently made due to centrifugal ~h~~~is, of course, no the rotor and stator blade rings being assembled externs1 forces as there is for rotor blades. ~h~ method of attachment is usually similar to that to the easing and then assembled endwise into the casil* for mbrblades, although a simple T-slot mot attach- as a unit, as in Fig. 10. The casing is then merit is frequently used. Blades may be mounted an air inlet easing and a discharge diffuserthat locate a'"1 This has the advantag(' -hined in the fasing, mounted in restrain the blade rings endwise. d.,.tlY b
qs
= n -fl H5/4
(2)
! 1
I
*
I
MARINE ENGINEERING
GAS TURBINES
23 1
U = WHEEL SPEED C I = NOZZLE EXIT V E LOCITY D l = BUCKET RELATIVE ENTRANCE VELOCITY D~
= BUCKET RELATIVE
Ce
= BUCKET ABSOLUTE
E X l T VELOCITY EXlT VELOCITY
I
C2
I (A) IMPULSE DIAGRAM
I (
V E L O C I T Y RATIO. U/C,
fig. 20 Typical turbine dtldender vs. speciflc spwd
specific speed, although the exact level of efficiency for each type is dependent upon the detail designs and geometry as well. 7.2 Aerodynamic Design. Since radial-outflow turbines are rarely used, discussion will be limited to radialc2 inflow designs. While the radial-flow turbine is basically similar to a centrifugal compressor operating in reverse, there are differences, as can be seen from Fig. 14. ( B ) 'OX REACT'0N 'IAGRAM Principally, a simple scroll inlet is usually used, and the considerable radial depth of the diffuser is replaced by a pig. 21 AX~~I-now turbine velocity diagrams series of short guide vanes or nozzles, which direct the incoming stream into the wheel. Also, the design is somewhat less critical from an aerodynamic point of view since the flow through the passages is accelerating A typical diagram for the so-called impulse stage (i.e., i n s u d of diffusing. This results in easier control and no pressure drop or corresponding conversion to velocity . A less build-up of the boundary layer, so that passage energy in the bucket passage) is shown in ~ i g21(a). shapes are not as critical. similar diagram for a reaction stage, in which p r t of the The best specific speed 9 of a radial-flow turbine lies pressure drop and conversion to velocity energy does in the range 2.5 to 4, as shown in Fig. 20. It should be occur in the bucket, is shown in Fig. 21(b). A stage will noted that different turbines are not only affected by vary in the amount of reaction with the velocity ratio, changes in specific speeds, but also by influences of over- U/C1, i.e., the ratio of the bucket circumferential all she (i.e., Reynolds number, manufacturing con- velocity to the theoretical spouting velocity for the total available energy to the stage.
Rg. 22
Axial-flw turbine stage impad dikimcy
pressure ratio as shown by the typical curves of Fig. 22. Impact efficiency as used in Fig. 22 is defined as the ratio of the work developed by the turbine to the thearetical work available when expanding from the initial impact pressure to the final impact pressure. The impact pressure equals the static pressure plus the impact head corresponding to the absolute gas
The following possible 'sources of loss probably cannot be evaluated directly. 1. N o n s b d y state effects which may produce a nozzle bucket "interferencen loss. 2. Nonuniformity of nozzle exit stream within the noz&lepit,&. 3. &nificant flow separation from nozzle or bucket profiles due to either a poor profile design or an exces-
1. Friction losses on all gss path surfaces. This may be assessed by use of an "effectiven drag or surface friction coefficient. 2- Flow mprtration and mixing losses as occasioned by excessive edge thickness, diffusing interstage passages, and divergence of sidemlls in nozzles and buckets. 3. Leabage losses such as nozzle seal leakages, wheelspace flows, and bucket tip clearance. 4. Shock and attendant high Mach number losses such as may occur from the nozzle throat up to the bucket entrance. 5.' h s a due to nonuniformity of available energy, which is a loss more or less unique to a directly fired gas turbine caused by the large initial temperature non-
turbines with radial entry, the nozzles must turn the stream from the radial direction to the proper angle, and thus look more like a conventional cascade. Nozzles are constructed frequently of two side. plates of appropriate inner and outer diameters with the nozzle vanes fastened between. The vanes can be pinned or riveted in place, or the side plates can be punched with openings the shape of the vanes which are then brazed or welded in place. For smaller units, precision cast nozzles with vanes and sidewalls intern1 are used with advantage. The material of the nozzles can be any of the commonly used high-temperature alloys, depending upon the design inlet temperature and, to some extent, on the fuel used. For fabricated construction M-252, S-816, A-286, or the
232
GAS TURBINES
MARINE ENGINEERING
Fig. 23
Crou section of large axial-flow gar turbine with drum-type rator, rotors welded up from several forgings
Fig. 24 Variable-angle load turbine nozzle arrangement, fadory assembly of one half the nozzle showing partitions, levers, and operating linkage in turbine shell
various Inconels such as 713 are suitable, while for cast construction X-40 or similar alloys are commonly used. In an axial-flow turbine, the nozzles are formed into an annular ring and serve to turn the flow from an essentially axial direction to that required by the bucket entrance conditions. The actual mechanical arrangement varies. Turbines using rotors built up from i~idividualwheels bolted together use different nozzle arrangements (particularly for the stages following the first) than those units using a drum type of rotor construction. The former arrangement is illustrated by Figs. 9 and 10 while the latter is shown in Fig. 23. The nozzle must be centered in the turbine casing and held against the pressure forces due to the pressure drop across the nozzle; a t the same time it must be free to expand with temperature changes. In large units operating at high inlet temperatures, the nozzle assembly may expand over an inch on the diameter between cold conditions and full load. To allow for this expansion and still keep the
carries some form of multitooth packing running against the drums or wheel rims to restrict the leakage flow to a controlled amount. A typical example of this type of construction is shown in Fig. 23. In the case of two-shaft machines, which are usually constructed with the gas producer turbine and the power turbine separate but with a common axis of rotation and in the same turbine casing, the nozzle for the power turbine can be of either fixed or variable area. The fixed-area nozzle is constructed the same as other intermediate nozzles, except that the diaphragm portion is usually solid as it normally does not encircle a shaft. For greater operating flexibility, the load turbine, if single stage, can be built with variable-angle partitions to give a variable area; in this way the energy distribution and hence the relative speeds between the gas producer turbine and the power turbine can be varied by tho control system. The advantages of this arrangement are discussed in Section 4. A variable nozzle area arrangement is shown in Fig. 24. Nozzle ,haterials selected depend upon the operating temperature, and consideration must be given to thcr fuel used, since some elements in the fuel can cause rapid corrosion or intergranular attack of otherwise suitablo
!
materials. The Nimonic materials have been frequently used but more recently other materials such as 5816 and M-252 have been used in forged and machined partitions. At lower temperature levels, A-286 or Type 310 stainless steel are suitable. Suitable materials for precision cast parts are Haynes Stellite's casting alloys X-40 (HS-31), or X-45 and FSX-414, which were developed from X-40 for improved characteristics. FSX-414 material has good strength, resistance to thermal fatigue, resistance to oxidation, and resistance to sulfurous corrosion; it has been used for turbine nozzles a t firing temperatures up to 1700 F. For long-term operation a t high inlet temperatures (i.e., 1500 to 2000 F or more) some form of nozzle cooling is required to keep metal temperatures below the peak gas temperatures so as to attain lives of 50,000 hr or more. The simplest and most commonly used method is air cooling by means of internal passages cast into the partitions. A typical arrangement is one in which the compressor air passes between a fabricated core and the cast shell forming the partitions, and is bled into the gas stream through holes near the trailing edge to keep the thin edge cool. 7.4 Rotor Design. The design of the radial turbine rotor is basically the same as that of a centrifugal compressor rotor, except that to lower the hub stresses the disk is generally walloped between vanes. Also, the rotor is frequently made in two parts, held together by the center bolt. The outlet portion or exducer corresponds to the inducer of a centrifugal compressor. This is advantageous since the vanes in the larger diameter, highly stressed portion of the rotor can be straight and radial, while the portion curved to give the required exit angle is in the smaller-diameter, lower-stressed exducer mtion. This also enables a composite construction with a forged and machined wheel and cast exducer to be used. Stresses in the rotor can be calculated by use of the Manson method, or modifications of it, as described for aompressor rotors. However, due to the high temperature levels, and the high temperature gradients from the outer diameter to the hub, the temperature distribution must be accurately assessed and the thermal stresses taken into consideration. The gas loading on the blades is negligible (although difficulty is sometimes encountered due to blade vibrations) and stress variation in an axial direction is neglected. Stresses can be calculated for various operating oonditions; but two combinations that have been used are:
1. At a speed corresponding to the overspeed trip setting (usually 110 percent) and temperatures correnponding to full power. 2. At temperatures corresponding to the overtemperature trip and lwpercent ~peed. The calculated stresses for a typical radial turbine wheel of A-286 material are shown in Fig. 16. The mmponding values of 10,000-hr rupture stress a t the
233
estimated wheel temperature at the location are also shown. For marine service where continuous operation a t rated load is expected, the material stresa and corresponding factors of safety would be based on long-time rupture, probably 100,000 hr. It can be seen that overstressing occurs a t the bore of the disk in both cases. This is typical of many rotors with a central hole. Because of this, and as a test of disk quality, rotors are usually overspeeded to a speed that will cause plastic yielding in the bore a t room temperature. This prestresses the bore, so that the disk is then stable and experiences little or no further plastic deformation a t normal running speeds. With the profile of the rotor disk determined to give the desired factors of safety, the major mechanical design of the rotor is complete. It is usual to build a sample rotpr and excite it with a vibrator to determine the vibration modes of the vanes. The modes determined are tuned by minor changes in thickness or profile to avoid resonance with possible exciting forces, such as nozzle passing frequency a t normal running speeds. One point that must be borne in mind, is that the rotor will expand due to centrifugal stresses. An allowance must be made for the corresponding strains and the thermal growth due to the operating temperature when establish'ing the cold clearance between the rotor qnd the nozzle and the vanes and the casing in order to provide adequate running clearance under operating conditions. For good efficiency, these clearances must be as small as possible; therefore, careful assessment of the various growths is necessary. I n an axial-flow turbine, the buckets or blades are usually attached to the wheel or rotor mechanically by what is commonly known as a "fir tree" attachment (or dovetail). But sometimes on smaller units the buckets are attached to the rotor by welding, or even cast integrally with the wheel. Good bucket design is a difficult and complex process. Not only must the aerodynamic conditions be satisfied as to entrance and exit angles and passage shapes, but equally important the vane section must have centrifugal and bending stresses below the allowable limits for the material and service involved. There must also be no vibration modes, of any significance, that are resonant to stimuli in the operating speed range. The bucket design is by necelqsity a process of trial and error. The final design of the bucket and its attachment entails plotting the calculated and/or measured vibration frequencies on a Campbell diagram to determine possible resonances [35]. It is generally considered advisable to keep the three fundamental frequencies from coinciding with passing frequencies of combustion chambers, nozzles, struts, etc. above 50-percent speed. First- to third-order frequencies are also to be avoided a t running speed. Even a t best all stimuli cannot be avoided and so some manufacturers have found it advisable to build in dampening in the form of tie wires in long buckets or special root dampening devices in short ones. Interlocking, integral covers, which are practical with
234
1
1
MARINE ENGINEERING
precision cast buckets, can also be used to control vibration. With the vane sections determined, their properties a t various radii can be calculated (i.e., section areas, section moduli, and centers of gravity). The areas can then be used to calculate the centrifugal stresses along the vahe section, and the section moduli in combination with the gas bending forces are used to determine the bending stresses. Usually the sections are "stacked" with respect to their centers of gravity so that the centrifugal forces create a moment which offsets the gas bending moment and reduces the stresses in the leading and trailing edges. The combined stress is then compared with the allowable stress a t that radius. The allowable stress is determiged by the long-time stress rupture properties of the material, since in marine service long and continuous operation a t or near full load is required. The criteria vary of course with the manufacturer and the expected service requirements, and the allowable stress criteria used range from 50 percent of the 100,000-hr value to perhaps 75 percent of the 50,000-. hr value. For large units intended for the main propulsion of large seagoing vessels, the minimum calculated rupture life should be not less than 70,000 hr. The root attachment, frequently called the dovetail, must carry the centrifugal loading of the vane section plus that of the platform and the dovetail itself into the wheel. This must be accomplished without exceeding the allowable stresses in the dovetail or the adjacent wheel rim sections. Frequently, the bucket material and the wheel material are not the same, so that the allowable stresses in the two parts differ even though the temperatures are generally assumed the same. Stresses in both parts must therefore be checked. An extended or "long-shank" bucket design may be used to reduce the temperature a t the dovetail, by the temperature dmp in the shank. A device used to attain the required aerodynamic shape within allowable stress limits and vibration characteristics is the hollow bucket. The whole vane section may be hollow, but usually just the tip section is hollowed for pahaps ysthe length of the vane. Inforged and machined buckets, the hollow can be formed by spark-discharge machining, while with precision cast buckets the avity is cast. In either w, the effective tapar ratio of the vane is increased, which reduces the root stresses for a given length and pitch line velocity. There is a wide range of bucket materials available, the choice again depending upon the temperature level and consideration of the effect of the fuel used. Both nickel-based and cobalt-based alloys are used for both cast and forged buckets. Diffused aluminum coatings are often used to increase the hot corrosion life of nickelbased alloys. At lower temperatures, as in last stages, A-286 is suitable, while 422 material has been very s u c ~ ~ w fwithin ul its temperature limitations. It should also be pointed out that most gas turbine buckets are unshrouded; i.e., they have no shrouds or covers as are frequently used in steam turbine practice.
b
Leakage past the tip of the bucket is limited by running closely controlled clearances between the tip of the bucket and a stationary member in the turbine casing. There is evidence that unshrouded buckets have tip losses equal to, or in some cases, less than shrouded .buckets in which only the axial clearance is controlled. The addition of several radial seals on the cover of a shrouded bucket will improve the efficiency. Smaller rotors are usually held together by a single central bolt, as in Fig.10, with the angular location of the individual wheels maintained by dowels or "curvic" couplings. With large rotors, a single central bolt becomes a rather formidable proposition, and so a multiplicity of smaller bolts a t an intermediate radius is usually used. I n either case, the bolt or bolts are stressed to a point where the resulting preload will positively prevent the wheels from separating under all normal operating conditions. Dynamically, the rotor then acts as a solid or one-piece rotor, and calculations such as that for critical speed can be made on the basis of the eection inertia of the rotor a t the contacting points, or lands, between the wheels. The tensile load in the bolts is usually checked a t assembly by measuring the actual stretch or elongation of the bolts as the nuts are torqued up. For very large rotors, the bolts may be stretched hydraulically, the nuts seated, and the elongation checked after the hydraulic load is released. The centrifugal load of the complete bucket (vane, platform, ahank, and dovetail), plus the interrupted portion of the wheel rim between the dovetail slots, is carried by the wheel disk. This loading can be represented as a distributed load around the wheel circumference. The stresses in the wheel disk are usually calculated by the Manson method, the same as for the compressor disks. However, in turbine wheels the thermal stresses are of considerably greater magnitude and must be assessed as accurately as possible. Methods of calculating heat transfer from a rotating disk are given by Kreith and Taylor in reference [36] and a method for calculating stress in disks subjected to creep is given by Wahl in reference IJI]. It should be ~ o i n t e dout that for turbines used in marine service, stress levels should be based on long-time opention and creep may need to be considered as a factor. The large thermal ~ o w t of h the nozzles due to their and elaatic growth of h k h temperature and the the wheels make it difticult to ensure the desired overlap of the nozzle and bucket a t the root of the gas path under all oonditions. A uniform overlap around the circumference isparticularly ditfcult to maintain. of the rotors must be critical calculated to avoid resomnce with normal rullning speeds. ~h~~~ calculations are -lly based on the traditional byleigh method, taking bearing flexibilities into account. While exact bearing flexibility is rarely known, a range of values usually cttn be estimated from past experience so that meaningful values of actual running critical speeds can be determined. Critical speeds calculated on the assumption of rigid bearings are of little value.
GAS TURBINES
Section 8 Combustion Systems
1
I
8.1 Combustion Chamber Conlgurations. The term A variation of the basic arrangement of individual 'gas turbine" is a contraction of the more explicit term chambers is one in which the chambers are arranged for "combustion gas turbine," and the process of combustion "reverse flowU; that is, the air from the compressor is an important part of the gas turbine system. The gas enters the downstream end of the casing and flows turbine is basically an internal combustion engine as the between the casing and the liner toward the h a d end. combustion, which releases the energy in the cycle, takes The air enters the liner alow its length and reverses place inside the machine. In gas turbines, the com- direction to flow back to the turbine nozzle as a product bustion is also a continuous process, as contrasted to the of combustion, i.e., hot gas. intermittent or cyclic process that takes place in a A modificati~n of the individosl chamber concept, reciprocating engine, whether the Otto or diesel cycle. called the "cannular" arrangement, is one in which a The combustion also takes place in a very confined space, multiplicity of individual liners (up to 16) is lorated in an as contrasted to most other continuous-process com- annular space between an inner and outer combustor bustion systems, such as the furnace of a conventional casing. Two variations of this firrangement are possible. boiler. The resulti~g continuous high rates of heat I n one the air is introduced a t the head or dome end of release make good combustion and cooling of the combustion chamber major problems-problems which have, however, been successfully solved in a number of ways. Combustion chamber design is as much an art as science and, perhaps for this reason, a wide variety of configurations is used for the combustion systems of gas turbines. In some cases the design is dictated more by the experience and practice of the manufacturer than the inherent ad vantages of a particular design. Combustion systems may be divided into two broad classes : 1. Those separate from the compressor and turbine that are mounted adjacent to, but not on, the unit. 2. Those built as an integral part of the unit and combined structurally with the compressor and turbine. European manufacturers more frequently use type 1, which is illustrated by Fig. 25, while American manufacturers commonly use type 2, which is illustrated by Figs. 9 and 10. Combustion chambers separate from the compressor and turbine generally take the form of a single large chamber or pressure housing with an internal liner or liners that contain the combustion products and protect the vessel from the high gas temperatures. Such systems can also be designed with separate smaller combustion chambers grouped within the pressure vessel and manifolded together a t the chamber exits t o provide only one or two gas ducts to the turbine. Some single combustion chambers are separate from, but mounted directly on, the unit while sml1 machines may have a single combustion chamber built into the compressor and turbine structure as in Fig. 14. Larger gas turbines with integral combustion systems LEGEND employ a variety of arrangements. A number of individual chambers may be mounted around the axis of A - COMBUSTION AIR NOZZLE the machine and be fed from a common compressor B- DISCHARGE TO TURBINE C SIGHT PORT AND QlLUTlON HOLES discharge; the chambers then feed into individual arcs D- LINER SUPPORT POINTS of the turbine first-stage nozzle. This design can also be E-FUEL NOZZLE adapted to a regenerative cycle by arranging the indiF COMBUSTION AIR REGISTER G-IGNITOR vidual chambers to be fed with air returning from the regenerator by one or more headers as in Fig. 9. Fig. 25 Large single canbultor f a mounting separate from gas turbk. 9
-
-
- . 236
GAS TURBINES
MARINE ENGINEERING
237
Smokefree combustion is generally in conflict with leanlimit stability. Similarly, low pressure loss and short flame length are difficult to attain simultaneously. The aerodynamic processes in a combustion system design present a most difficult set of problems. The provisions for the interaction of fuel and air by culation in the primary zone to provide the correct spatial arrangement of the reactants, the necessary flame stability, the necessary mixing and dilution downstream to give the required temperature distribution a t the combustor exit, the maintenance of metal parts at Proper operating temperatures, and the low Pressure drop requirements combine to present a complex design challenge. 8-4 Mechanical Details and Construction. Structurally, the gas turbine combustion system is rather ~imple,but the design is complicated by the varied and rather severe conditions it must meet. The design must 1. Flame temperatures of 2900 to 3300 F or more. 2. Cyclic temperature variations. 3. Metal temperatures during operation of 900 to 1300 F with peaks to 1700 F. 4. Metal temperature gradients of 540 deg F/in. 5. Exciting frequencies, mechanical or aerodymmic,
over a wide frequency range and with varying amplitudes. 6. Collapsing pressure differentials across the liner of
on suspension pins or fittings should normally be limited to 250-500 psi. Sometimes surface treatment of the
as atmospheres to the 1.8 power, i.e. I = Btu/hr-cu ft-atm1.8
I' = Btu/hr-sq ft-atm1.8 Values per cubic foot per atmosphere will range from 1 to efficiency apprecisbly. 10 X 10' while, if the pressure is taken as atm1.8, 8.3 Design Objectives. The physical configuration values will range from 0.5 or less to 5 X 108. Typical of representative combustion systems and the mo* values per squale foot of flow area inside the liner will important parameters used in their design have been run 4 X 106 Btu/hr per atmosphere. described, but specific design objectives have not been In small chambers, the radiant heat flux will be of discussed. I n addition to the obviaus primary objective the order of 80,000 Btu/hr-sq ft, which increases with of burning the required amount of fuel, the objectives in the size of the chamber due to the greater thickness combustion chamber design, not necessarily in order of of the body of radiating flame. It also increases with the importance, may be liste'd as: luminosity of the flame. Liquid fuels, in general, give a 1. Effective release of chemical energy by the cornmore luminous flame than gaseous fuels; the heavy oils1 bustion of fuel within highly confined spaces. particularly the residuals known as Bunker C1 are very 2. Stable operation over a wide range of fuel/air bad in this respect. The result is that large single ratios. combustors intended to burn heavy oils are designed for 3. ~ ipressure drop ~ compatible i with ~ the other~ rather low rates of heat release compared to the small requirements. individual chambers used in some designs. 4. A controlled temperature distribution a t the Most combustion chambers are designed so that a turbine inlet. small amount of primary air is admitted, U S U ~ ~with ~Y 5, short flame length to prevent discharging flames some swirl velocity, a t the upstream end of the ~ b m b e r into the turbine. in what is usually called the dome. In the dome the air 6 , clean burning, i.e., negligible smoke and no is mixed with the fuel, which then burns nearly carbon formation under all operating conditions. chiometrically witl; temperatures of 3000 to 3500 F. The b bto give l ~reliable , starting. 7. ~ ~ ~ ~easy~ignition quantity of air admitted in the primary zone may give 8. Durable components with low maintenance. up to 20-30 percent excess air a t that point. Of the Some of these requirements are mutually conflicting. remaining air, part is used for cooling the liner (perhaps
Combustion system components can fail in one or more of the following ways: 3. Structural failure due to pressure forces. Di8tortion and due to temperature 5. Overheating, resulting in local metal failure. 6. High-temperature corrosion.
~
Pressure loads, tending to collapse the liner and axid forms due to the PrefjSure dzerence over the dome arm1 are important, particularly in the case of annular designs- The critical pressure difference that the liner depends upon the pmportions (length/hmeter and thickness/diameter), the degree of and stiffening, manufacturing eccentricities, discontinuities, and Young's modulus a t the operating bm~eratures- Collapse is usually due to yielding a t locd eccentricities rather than elastic instability. In the days of gas turbines, very thick liner rnaterial was sometimes used in an attempt to obtain longer life and reliability, but experience has shown that the thinner the liner) the better, and except in large nirigle chambers, thicknesses of more than 0.050 in. to 0.060 in. are rarely required. The liner must be s u ~ ~ o r in h dthe combustor casing being allowed to expand freely. Bearing pressures
Liner materials can be any one of a number of hightemper&ture, c~rrosion-re,gisting alloys. ~h~ ~ ~ i ~ i ~ commonly use Nimonic 75 below 1300 F and Nimonic go, which is less ductile and more difficult to fabricate, for higher temperatures and larger diameters. In the U.S., stainless steels 18-8 or 25-20 and Inconel are the usual materials, and they give satisfactory results where cooling is carefully controlled. Occasionally, combustor parts are attacked by corrosion*dueto constituents of the fuel. Attack by H*S is particularly virulent, but is likely only in the zone, or overrich pockets. ~t is below temperatures of 1100 li. ~ t t a ~ k also occur from vanadium pentoxide, which may be formed from the vanadium that is frequently present in residual fuels. 8.5 Fuel Nozzles. The fuel nozzles, or fuel injectors, must introduce the fuel into the combustion chamber over the entire range from lightoff to maximum load in a way that is compatible with the basic objectives listed in the foregoing. Fuel noazles can have a marked effecton the performance of the combustion chamber, and must be considered as a n integral part of combu&ion systems. Gas turbines are well suited for the propu~sion of
238
GAS TURBINES
MARINE ENGINEERING
liquefied petroleum gas carriers as they can readily burn the gas boil-off [38]; however, aside from this application, gakous fuels are not available for most marine applications. Liquid fuels commonly used in marine installations are JP-4, JP-5, distillate fuels, and residual fuels. The problems with each one vary, and few noz~l&or fuel, systems will provide satisfactory operation over the whole range of fuels. The nozzle must atomize the liquid fuel into droplets sinall enough to insure completk combustion of the fuel ih the residence time avawble, and provide a spray pattern that will insure intimate mixing of the fuel with the available oxygen under the full- and part-load aerodynamic conditions existing in the primary zone. The spray must not strike the walls of the chamber, and must not cause excessive temperature variations. Five types of liquid-fuel nozzles have been applied in gas turbine combustors. Each has advantages and disadvantages as follows: 1. Pmaure-Atomiaing Nozzles (a) Simplex (or single orifice) Nozzles. The simplest type but not suitable for the wide flow fange usually required. (b) Duplex (or dual orifice) Noszles. With either intern1 or external flow dividers, these give a much wider operating range and are usually satisfactory for the lower viscosity fuels. 2. Ai Blast (Air Atomizing) Nozzles. These nozzles u t h e the combustion-chamber pressure drop to create an air s t r a m that is used to assist in ato&lng the fuel. The air flow around the nozzle body breaks up the fuel spray, so the fuel distribution is determined mainly by
the air flow pattern. Combustion is characterized by a blue flame of low luminosity, cool walls, and minimum smoke. I t has, however, poor "lean blowout" and poor atomization at starting. 3. Air Assist (Air Atomizing) Nozzles. These are single-orifice nozeles, with small quantities of air introduced internally via separate passages from a separate external compressor. At low fuel flows and pressures, the air atomizes the fuel, while at high flows the air can be cut off and fuel pressure relied upon for atomization. For h a v y residual fuels, air assist nozzles can be designed with a separate air compressor taking air from the main compressor discharge, and providing a pressure ratio of about 2: 1 over the entire operating range; this gives efficient atomization (a Sauter Mean Diameter of the fuel droplets of 60-80 microns) over the whole firing range. The complication and weight of the extra atomizing air compressor penalize this arrangement, but it is the only one suitable for burning Bunker C in high-intensity combustion systems. They may also be used in distillate burning machines to assure smokeless combustion throughout the load range. 4. Vaporizing Fuel Nozzles. In vaporizing nozzles the fuel is introduced in a tube or passage and vaporized by heat from the combustion chamber. The advantages are: (a) the vaporizing tubes contain both fuel and air and very rich mixtures are avoided, (b) a blue flame is produced and problems of smoke and radiation are diminished, and (c) dispersal of the fuel is dictated by the airflow pattern and is sensibly independent of fuel flow. The outlet temperature traverse is, therefore, not sensitive to fuel flow. The system has the dissdvantages that it will not operate at weak mixture ratios and a
SPARK RATE MONITOR OUTPUT
INPUT VOLTAGE:
110-120 VOLTS-50-60
INPUT CURRENT:
7 AMPS, MAX.
STORED ENERGY RANGE:
1-20 JOULES, lNCREMENTS OF I JOULE
SPARK RATE RAmE:
0-100 SPARKSISEC AT I JOULE 0 - 3 SPARKSISEC AT 20 JOULES
GAP VOLTAGE:
3000-3100 VOLTS
SPARK RATE MONITOR OUTPUT
30 VOLTS PEAK
Fig. 27 "Higlen&
CYCLES
ignition s y s h l circuit
given design will only operate sstisfactorily with minimal variation in fuel properties. 5. Centrifugal Atomiaers. These employ the centrifugal force of a rotating cup to atomize the fuel. This arrangement gives very uniform temperature distribution (f25 deg F in a radial direction and f 7 5 deg F circumferentially). I t is adaptable to a wide range of fuels, from liquid propane to SAE 30 oil, with no apparent trouble from combustion chamber deposits. It is, however, only suitable for annular combustors with a rather restricted configuration. All of these systemshave been used in gm turbines, and the choice depends upon the fuel to be used, the arrangement of the combustion system chosen, the range df fuel flows over which satisfactory operation must be obtained, and the ambient conditions (particularly the temperature at which reliable light-off must be obtained). 8.6 Ignition System. Normally the combustion in the system is self-sustaining and continuous, but it must be initiated by an external means. This is the function performed by the ignition system. Electric ignition is almost universally used, the system consisting of a spark plug or plugs in the cornbustion chamber and a source of high voltage to create the spark a t the plug electrodes. The voltage used ranges from 3500 volts for a low-voltage system to 18,000 volts in a high-voltage system. The energy required
239
varies from 0.2 to 4.0 joules per spark and up to 12 joules can be released in about 100 microseconds, i.e., 100 kw at peak spark discharge in a high-energy system. However, only about 7 percent of the total heat energy appears at the plug face. The energy can be furnished by a high-voltage trans former, sdmetimes with a capacitor to store the energy, or in some cases a magneto driven by the engine or starting device is used. The electrical circuit for a high-energy system is shown in Fig. 27. The spark plug can be of several types. A plug with a central electrode in an insulator, discharging to another electrode on the body (similar to an automobile sparkplug), can be used. Far large chambers, where it is difficult to insure the presence of a combustible mixture a t the gap of a plug mounted in the combustor case, a retractable plug is used. The plug can be spring-injected and retracted by compressor discharge pressure acting on the piston, or it can be pressure-injected and springreturned. In either case, the electro&s can be inserted far enough into the chamber to insure ignition without danger of burning them during operation. For high-energy systems a aurface-discharge plug can be used. This type is less susceptible to fouling or burning. I t furnishes a high-energy murce to ignite the fuel but can be used only with a high-energy, capacitordischarge electrical system.
Section 9 Bearings, Seals, and lubrication 9.1 Bearing Types. Gas turbine bearings fall into two classes, (i) antifriction or rolling-contact bearings and (ii) sleeve or journal bearings and the corresponding oil-film thrust bearings. Gas turbines adapted from aircraft designs generally use antifriction bearings because of their high capacity, low space requirement, and lighter overall engine weight, although in small ongines this difference is negligible. Gas turbines adapted from industrial practice or designed specifically for marine service generally use sleeve or oil-film bearings, particularly in the larger sizes. 9.2 Antifriction Bearings. Antifriction bearings, whether ball or roller, can and do provide bng, reliable ~ r v i c ebut ; by their very nature they have a definitely limited life. For a given load and speed the bearing manufacturers specify a B-TO life, which 90 percent of bearings operating with proper lubrication and temperatures will meet or exceed. For antifriction baring applications, each rotor is trupported by two or more bearings. One bearing is frequently a cylindrical roller bearing to provide for axial movement of the rotor, and the other bearing, which locates the rotor and takes the axial thrust, is nome form of ball bearing. I n the arrangement shown
by Fig. 10, the third bearing, which supports the turbine, is also a cylindrical roller bearing that permits axial expansion between the compressor and turbine. Roller bearings are frequently used adjacent to the turbine wheels, since they are usually considered better able to withstand high operating temperatures and higher soakback temperatures after shutting down. In addition, they have somewhat greater overload capacity, which may prevent complete bearing failure in the event of damage and loss of balance in the rotor. While antifriction bearings require only small amounts of oil for lubrication, means must be provided to keep them cool, particularly in locations where heat from the turbine can flow into them through the shaft or housing. A carefully controlled amount of cooling oil is required to maintain their internal clearances and, at the same time, avoid an excess that can be trapped between the rolling elements (particularly rollers) and the raceways with resulting hydrauiic lock and noise. The oil jets should be directed a t the shaft adjacent to the inner race and around the housing, to keep the bearing cool without excess oil being forced into the bearing. 9.3 Sleeve Bearings. Properly designed oil-film bearings, appropriately applied and supplied with an
MARINE ENGINEERING
240
GAS TURBINES
-
*' Fig. 28
Labyrinth bearing sealing arrangment
adequate quantity of clean oil of suitable viscosity, will run almost forever. Gas turbine sleeve bearings usually follow normal bearing practice. I n large machines, the journal bearings consist of babbitted shells (frequently with spherical, self-aligning seats) held in bearing housings of cast or fabricated construction. Bearing loadings are usually below 200 psi of projected bearing area. The bearing bore is frequently elliptical, and where very lightly loaded it may incorporate special pressure pockets or other means to ensure stability. Since the rotor weight goes down as the cube of the scale while bearing area reduces only as the square, bearing loadings are very low in small units even though L/d ratios less than 0.4 are frequently adopted. Lightly loaded bearings are subject to oil whip and other instabilities, so special measures must sometimes be taken to provide stable, satisfactory operation. Three-lobed bearings, floating bushes, and even pivoted-shoe journal bearings have been employed in various units. 9.4 Thrust Bearings. With antifriction bearings, one of the ball-bearing assemblies is generally arranged to take the net thrust. The net thrust is the difference between the com~ressorthrust and the 'turbine thrust --. in the case of &s producers, or the turbine thrust plus or minus any gear or coupling thrust in the case of power turbines. With oil-film bearings, a separate thrust e i n g is usually provided for each rotor. This bearing normally comprises a thrust collar firmly attached to the shaft with thrust washers or thrust plates on either side. For lightly loaded thrust bearings, these can be simple babbitted flat plates with radial grooves for oil distribution and flow. However, it is preferred practice to use tapered-land thrust shoes in which each land has a slight circumferential taper. While tapered-land thrust bearings can carry very
high loadings (above 750 psi), they are not inherently self-aligning and require careful manufacture as the taper required is very slight. Consequently, for highly loaded thrust bearings, a multiple, pivoted-shoe thrust bearing is often adopted. Since the thrust is usually greater in one direction than the other, the more lightly loaded side is sometimes made with fewer pads or shoes than the loaded side. Sleeve bearings and thrust bearings require an ample supply of clean cool oil for both lubrication and cooling. These bearings operate with a hydrodynamic film; and various factors, such as an interruption in the oil supply, high oil temperatures or low oil viscosity, can result in bearing wear and damage or even failure (by unduly reducing the oil film thickness). Small gas turbines frequently run with light oils (SAE 10 or even lighter) and rather high temperatures. The aircraftcderivative gas turbine is designed to operate with a synthetic oil, while the heavy industrial type of turbine is usually designed for heavier oils (at least turbine oil, medium). The oil temperature to the bearings of industrial-derivative turbines is about 130 F with a 30 to 40 deg F rise through the journal and thrust bearings. 9.5 Shaft Seals. Shaft seals serve the purpose of preventing or controlling fluid leakage along a shaft where it passes through a wall or diaphragm that separates regions a t two different pressure levels or contains two different fluids. Shaft seals are used whero shafts enter a bearing housing, where they enter e compressor or turbine casing, and between individual stages of a compressor or turbine. Shaft seals can be divided into two general classss: contact seals and labyrinth seals. Contact seals usually consist of a carbon or graphite ring with a flat face that L
held by a spring in contact with a face or sealing ring on the shaft, which has been lapped almost absolutely flat. Contact is maintained between the two faces to prevent leakage; hence the name. The metal face is hardened, and the carbon ring material selected to give minimum friction and wear. These seals, which are used in a variety of other applications besides gas turbines, will operate with essentially no lubrication, although they are most frequently applied where a liquid is present on a t least one aide of the seal. They are usually proprietary items. A labyrinth seal works on the principle of a series of throttlings, produced by a series of teeth on the stationary member, the shaft, or sometimes both. The teeth break down the total pressure difference between the fluid on the two sides of the seal into a series of steps to control the Bow through the clearance space between the tip of the tooth and the mating member. The velocity created in the clearance by the pressure drop is a t least partially dissipated in turbulence in the volume between adjacent teeth, thereby minimizing the flow. See Chapter 2 for a discussion regarding labyrinth seals. The clearance that can be maintained between the shaft and the stationary member depends upon the apecific machine configurationused, particularly the location of the seal with respect to the bearings and the clearance in the bearings. For small machines with small shafts and the seals located immediately adjacent to the bearings, a total clearance of 1.3 to 1.5 mils per inch of shaft diameter is usually eatisfactory. For large machines with a considerable distance between a bearing and the seal, a radial clearance of 2 mils per foot of shaft span may be necessary. A variety of materials can be used for labyrinth seals. Simple bearing housing seals, as shown in Fig. 28, that are intended primarily to prevent leakage of oil out of the bearing housing, can have the stationary member solid and be made of brass or aluminum. For seals located remote from bearings, such as in turbine diaphragms, the stationary member is frequently segmented and spring-supported so as to limit the contact pressure in the event of contact with the shaft. At low temperatures, these can also be made of brass, leaded bronze, or even plastic. At high temperatures, however, ferritic or even austenitic materials with a chrome-moly steel shaft are used; such a seal construction is shown in Fig. 29. It is also common practice to put the teeth on the shaft, as shown in Fig. 30, to minimize the heating and resulting bowing of the shaft in the event of contact. A third type, which involves a combination of the two principles, is the carbon ring seal. In this case a carbon or graphite ring, which can be either solid or segmented and held together by a garter spring, is bored to have a close clearance to the shaft, so as to control the leakage by laminar flow through the clearance space. The ring h a t s on the shaft and is free to turn in a groove in the housing, but the pressure difference holds it against one aide of the groove and seals off leakage by that path. This type of seal is shown in Fig. 10. It is quite a
24 1
compact seal and is sometimes used in smaller machines where length is important. This type of seal has also frequently been used in steam turbine practice. 9.6 Lubrication Systems. Proper lubrication is vital to the operation of gas turbines, whether equipped with rolling-contact or fluid-film bearings. The continuous supply of the proper grade of lubricant a t the proper pressure and temperature is so important that most gas turbines are equipped with their own integral lubrication system. Where aircraft jet engines have been adapted to industrial or marine use, the jet engine used as a gas generator frequently has different lubrication requirements from the power turbine and power transmission system and, therefore, usually retains its own independent lubrication system. Engines equipped with rolling-contact bearings require less oil, and usually a different grade, than those with fluid-film bearings. Many antifriction-bearing engines are designed to operate with synthetic lubricants. It is important that synthetic lubricants be used only in engines equipped with suitable gaskets, O-rings, seals, etc., as synthetic lubricants will attack and cause rapid deterioration and failure of many common gasket and O-ring materirtls. Basically, the lubrication system consists of an oil reservoir, a pump or pumps (for pumping oil from the reservoirs to the bearings, gears, and control systems), pressure regulators (to control the supply pressure to the various components), an oil cooler or coolers (to control the oil temperature), and a filter or filters (to assure clean oil). A typical lube oil system suitable for large units is shown in Fig. 31. The lube oil flow is determined by the quantity required to absorb the losses of the bearings, plus heat pickup from the surroundings, within the allowable temperature rises in the bearings. The oil required by all control devices such as governors, hydraulic actuating cylinders, etc., must also be provided. With large units, the pump size may be determined by the oil flow requirements of hydraulic cylinders used to actuate variable-angle nozzles and similar devices. The oil pump is always sized to deliver more than the calculated requirements, the excess being returned to the oil tank (reservoir) by a pressure-regulating valve. The main lube oil pumps are generally of the gear type and a t lejtst one is almost always driven directly by the main gas turbine shaft to ensure that the pump is driven as long as the turbine shaft rotates. Shaft-driven centrifugal pumps have also been used for the main lube oil pump, and,,centrifugal pumps, usually driven by electric motors, are frequently used for auxiliary pumps where the capacity required warrants it. I n the smaller sizes, motor-driven gear or vane-type pumps are used for the auxiliary or emergency supply. Sleeve bearing units should always have a supply of oil to the bearings before the unit is started, and some designs also require control oil pressure before starting. This is the purpose of the auxiliary pump, which is usually under control of a pressure switch so that it runs
11
111
MARINE ENGlNEERING
242 /
TO HYDRAULIC CONTROL
is owrated for a period of time to circulate oil to the bearings in order 6 remove the heat that flows into them
Y
STATOR PART
Fig. 29
GAS TURBINES
Typical high-IW labyrinlh seal with atathary to&
whenever the control system is energized and no main pump pressure is available. The auxiliary pump will also start on loss of oil pressure while the unit is operating and thereby furnish lubrication until the unit can be shut down. Large units are also usually equipped with an additional emergency pump that is supplied from a reliable, separate power source. This will supply sufficientlubrication for the bearings to bring the unit to rest in the event of failure of the main and auxiliary lube oil supply. The unit is shut down on loss of bearing header pressure, and the emergency pump is started to protect the unit from damage. Sometimes the auxiliary oil pump, or a supplementary small "cool down" pump,
from the This hot keeps parts, the particularly wheels ofbelow the turbine. babbitt in the the bearings 250 F and prevents the damage that would otherwise result from -exposure to higher temperatures while the machine cools down. Antifriction bearing machines generally .do not require pre- or post-lubrication for the protection of the bearings, since rolling-contact bearings adjacent to hot parts of the machine are usually stabilized a t temperatures of 350 F to 450 F. However, lube oil deteriorates at the temperatures which are frequently reached after shutdown. Therefore, in some cases, posblubrication after shutdown is used to remove the heat and keep the bearing area cool to prevent varnishing and carbonizing of the oil. The heat generated in the bearings and gear meshes, plus the heat flowing in from the hot parts of the machine and absorbed by the oil, must be removed by the oil cooler. The latter source can amount to f i of the total heat absorption. For certain installations oil coolers are direct oil-to-air radiators, but for marine installations water-cooled heat exchangers are the logical choice. The heat exchanger must be capable of rejecting all the heat absorbed by the oil and provide a proper oil-cooler discharge temperature (usually 130 F). These heat exchangers are generally of the shell-and-tube type with the tubes readily accessible for cleaning. Frequently, the oil coolers are in duplicate, with quick change-over valves, so that the machine can operate with either cooler whiie the other is being cleaned.
OPTIONAL RECYCUNQ OF LUBE OIL
r - - - - - - -)------
3
PC MAIN L.O. PUMPS TO USED L.O. TANK
SCAYENGlNQ PUMP
NOTES: I. OIL FLOWS AND TANK CAPACITY TO BE SPECIFIED. 2.01L FLOW TO THRUST BEARING AND MAIN REDUCTION GEAR SHALL BE SPECIFIED FOR EACH APPLICATION.
LEGEND PC -PURCHASER'S CONNECTION -SUPPLLED BT PURCHASER -SUPPUED WITH GAS TURBINE -MOTOR DRIVEN PUMP
---
00
@ -ACCESSORY @EAR DRIVEN PUMP fig. 31
Typical lube-oil system for gas turbine propulsion unit
Oil filters are almost always installed, since it is generally considered good practice to do so. The filter must have adequate capacity for the full oil flow within the manufacturer's pressure drop limitations. Frequently dual filters are installed with quick transfer (four-way) valves so that one filter can be cleaned while the other is in service. The filtration system should be chosen with consideration for the minimum clearances in the machine. On large machines which may not require very fine (below 10 micron) filtration for the bearing oil, an additional finer filter (down to 2 microns) may be added in the circuit to the governor and other hydraulic devices to protect their very close clearances and fine finishes. Filter by-passes, particularly internal by-passes, are not recommended even on full-flow filters. When the filter is plugged with dirt and the by-pass opens, large quantities of contaminants may go through the by-pass and into the bearing system. I t is preferable to monitor the pressure drop across the filter and provide dual filters if it is necessary to assure continuous operation under all conditions. Pressure regulators are usually simple spring-loaded relief valves as close regulation of the pressure level is not important. Some systems use two pressure levels, one for the control functions and the other a t a lower
pressure for lubrication. The control circuit is usually arranged to have top priority on the oil supply, since operation of the controls is vital to the operation of the unit, and the lubrication supply can be reduced or even cut off for the fraction of a second it takes the controls to operate. The oil reservoir, or lube oil tank, is usually located below the unit although with positively scavenged (drained) systems it can be located anywhere. With jet engines it is frequently fastened to the side of the compressor casing. With gravity drain systems, it must be located a sufficient distance below the bearings to allow apsitive slope of a t least in. per foot to the drain lines under all conditions of pitch and roll. On small, compact units this is not difficult to achieve; but on large multishaft, multicasing units more than one tank may be required to provide adequate drainage without exceeding a reasonable suction lift on the shaft-driven oil pumps, which are generally mounted on or close to the unit. The capacity of the tank is usually basedon the main lube oil pump flow. Where possible, a capacity of four times the oil pump capacity in gpm should be used. Thii gives what is known as a four-minute supply; i.e., in the event of failure of the drain system, the tank will provide a four-minute supply of lubricant. Smaller
244
MARINE EN
units, particularly those mounted in a package, such as shipboard generating sets, may have smaller tanks. Due to space limitations, these may be as small as a two-minute supply or less. Adequate deaeration of the oil is diacult in tanks this small, although the carry-over of mist out the vent can be minimized by properly locatr ing baffling, by locating the oil drains and the tank vent connection as far apart as possible, and by providing a deaeration tray. Pressurized outer shaft seals, in which compressor bleed air is introduced between two seal sections and flows through one side into the bearing housing (to prevent oil leakage along the shaft), introduce extra air into the oil and make satisfactory deaeration more difficult. The tank should be provided with a bottom that slopes both ways to a drain connection. The tank should have a removable cover or access door of sufficient size to enable every part of the tank to be reached for cleaning. An oil level gage and/or sight glass should also be provided aloqg with a low-level and sometimes a high-level alarm. Supply and drain piping is preferably made of seamless tubing. Stainless steel tubing is frequently used in the smaller sizes. AN-type flexible hose connections are also suitable in the smaller sizes (below about 1.5 in.). In large units, it is good practice to run the pressure feed lines inside the drain lines as far as possible. The drain line then acts as a guard line in the event of a leak or failure of the feed line, which otherwise could spray oil onto hot parts of the machine and cause a fire. In order to avoid fire hazards, the number of pipe joints should also be minimized. Where joints are necessary, welded fianged connections are preferred, with an SAE four-bolt split fiange connection being the second choice. Compression-type fittings are satisfactory with the smaller stainless steel lines. Flexible lines usually are provided with standard AN-type fittings. Threaded pipe joints, and particularly pipe nipples, should be used only where unavoidable; and then extraheavy or double extraheavy-schedule pipe should be used to ensure adequate wall thickness under the threads to avoid fatigue failures from originating in the threads. Pipe sizes, both feed and drain, should be sized for low velocities a t full flow. A velocity of not more than 6 fps in feed lines and 2 fps in drain lines will keep system pressure drops to reasonable values and provide free and complete drainage from bearings and gear housings. Poor drainage can give trouble with oil leakage along shafts and extra losses and heating in gear systems. References
I
1 W. A. Brockett, G. L. Graves, Jr., M. R. Hauschildt, and J. W. Sawyer, "U. S. Navy's Marine Gas Turbines, " ASME Paper 66-GT/M-28. 2 R. C. Case, "Marine Gas Turbine Growth in the U. S. Coast Guard, * ASME Paper 66GT/M-36. 3 E. B. Good, "Gas Turbine Installation Design for Naval Ships," ASME Paper 66-GT/M-34. 4 G. M. Boatwright and E. P. Winert, "Combined
GAS TURBINES
Power Plants, " SNAME Philadelphia Section, March 1964. 5 R. G. Mills, "The Combined Steam Turbine-Gas Turbine Plant for Marine Use," ASME Paper 55-A-154. 6 G. C. Swensson and E. P. Winert, "Laboratory Test Experience with a Combined Steam Turbine and Gas Turbine Unit," SNAME New England Section, March 1963. 7 J. L. Mangan and R. C. Petitt, "A Highly Efficient Steam Turbine-Gas Turbine Cycle," Presented a t the ASME Aviation and Space, Hydraulic and Gas Turbine Conference and Products Show, March 1963. 8 A. 0. White, "The Combined Gas Turbine-Steam Turbine Cycle with Supercharged Boiler and Its Fuels," ASME Paper 57-A-264. 9 W. P. Gorzegns and R. J. Zoschak, "The Supercharged Steam Generator. Some Aspects of Design and Pressure Level Selection,"ASME Paper 66-GT/CMC-68. 10 "Installation Design Criteria for Gas Turbine Applications in Naval Vessels," Navships Technical Manual 0941-038-7010. 11 D. B. Harper and W. H. Rohsenow, "Effect of Rotary Regenerator Performance on Gas Turbine Plant Performance," ASME Paper 62-A-149. 12 A. T. Bowden and H. Hryniszak, "The Rotary Regenerative Air Preheater for Gas Turbines," ASME Paper 52-A-74. 13 I. Howitt and R. P. Thurner, "Gas Turbine, Extended Surface, Heat Exchanger; Modern Design and Performance," ASME Paper 64-GTP-18. 14 R. F. Caughill, "Design Considerations and Operating Experience of Regenerators for Industrial Gas Turbines, ASME Paper 61-GTP-12. 15 R. P. Allen and E. A. Butler, "An Axial Flow Reversing Gas Turbine for Marine Propulsion," ASME Paper 66 GT/M-21. 16 D. L. Caldera, C. E. Hoch, and G. C. Swensson, "Gas Turbine Propulsion Machinery for the MSTS Roll-On/Roll-Off Ship," SNAME New York Metropolitan Section, April 1967. 17 C. Zeien, H. F. Smith, and F. W. Hirst, "The Gas Turbine Ship Callaghan's First Two Years of Operation, " Trans. SNAME, V O ~ .77, 1969. 18 W. S. Richardson, "The Friction Clutch ReverseReduction Gears In the GTS," ASME Paper No. 69-GT-5. 19 P. K. Wennburg, "The Design of the Main Propulsion Machinery Plant Installed in the USCGC Hamilton (WPG-715)," Trans. SNAME, vol. 74, 1966. 20 K. H. Kurzak and H. Reuhr, "Propulsion Machinery of the Koeln Class Escort Frigates with Special Consideration of Gas Turbine Propulsion," ASME Paper 65-GTP-11. 21 L. A. Gunsteren, "Hydrodynamics of Controllable Pitch Propellers," SNAME New York Metropolitan Section, March 1970. 22 M. J. T. Smith and M. E. House, "Internally Generated Noise from Gas Turbine Engines," ASME Paper 66 GT/N43.
23 B. Wichstrom and H. Ohauist, "Startine and Control of a Large Gas ~urbihe,"'ASME Paper 64-GTP-7. 24 A. I?. McLean, "Control Design and Development for the Ford 704-705 Series Gas Turbine Engines," ASME Paper 64 WA/GTP-5. 25 D. A. O'Neil, "Governing Gas Turbine Engines for Marine Propulsion-Power vs Speed Governing," ASME Paper No. 69-GT-54. 26 W. B. Brown and G. R. Bradshaw, "Design and Performance of a Family of Diffusing Scrolls with Mixed Flow Impeller and Vaneless Diffuser," NACA Report Ann
.,,.I\
YJO-lY4Y.
27 S. 5. Manson, "The Determination of Elastic Stresses in Gas Turbine Disks," NACA TN 1279, 1947. 28 M. B. Millenson and S. S. Manson, "Determination of Stresses in Gas Turbine Disks Subject to Plastic Flow and Creep, " NACA TN 1636, 1948. 29 L. J. Herric, J. C. Emery, and J. R. Erwin, "Systematic Two-Dimensional Cascade Test of NACA &Series Compressor Blades a t Lol!~ Speeds," NACA TN 3916, 1957. 30 J. C. Emery, "Low Cascade Investigation of Thin Low Camber NACA 65-Series Blade Sections At High Inlet Angles," NACA RM L57E03, 1957.
f
f
245
31 J. C. Dunavant, J. C. Emery, H. C. Walch, and W. R. Westphal, "High Speed Cascade Tests of the NACA 65-(12Alo) 10 and NACA 65-(12AsIs) 10 Compressor Blade Sections," NACA RM L55108, 1955. 32 J. C. Emery and J. C. Dunavant, "Two Dimen) sional Cascade Tests of NACA ~ ~ : ( C I O A ~10O Blade Sections a t Typical Compressor Hub Conditions for Speeds up to Choking," NACA RM L57H05, 1957. 33 S. Lieblein, F. C. Schwenk, and R. L. Broderick, "Diffusion Factor for Estimating Losses and Limiting Blade Loadings in Axial-Flow Compressor Blade Elements," NACA RM E53D01, June 1953. 34 P. N. Bright, "Structml Design Problems in Gas Turbine Engines, " ASME Paper 54-A-152. 35 R. W. Nolan, 'Tibration of Marine Turbine Blading," Trans. SNAME, vol. 57, 1949. 36 F. Kreith and J. H. Taylor, Jr., "Heat Transfer from a Rotating Disk in Turbulent Flow," ASME Paper 65-A-146. 37 A. M. Wahl, "Stress Distributions in Rotating Disks Subjected to Creep Including Effects of Variable Thickness and Temperature," ASME Paper 56-A-162. 38 H. F. Smith, "Gas Turbine Propulsion of LNG Tankers,'' ASME Paper No. 69-GT-47.
MEDIUM AND HIGH-SPEED DIESEL ENGINES
bhp = brake horsepower, hp N = revolutions per minute, rpm
Laskar Wechsler
I
The '"piston speed" is the average speed of the piston during its stroke. It is usually expressed in feet per minute and determined from the expression: V, = ~ ~ 7 Piston speed is a useful yardstick for comparing the inertia loading and cylinder component wear characteristics of generally similar engines, 1.3 Types of Diesel Engines. Diesel engines are divided into various types for descriptive purposes. These descriptions are used to specify exactly the kind of engine wanted for a given application. The descriptive divisions include cycle (two-stroke or four-stroke),
overlap period and is usually expressed in degrees of crankshaft rotation. The intalre strolre is followed by the compression stroke to repeat the complete cycle. It can be seen that four strokes of the piston were required for the cycle. The two-stroke cycle is dominant for large bore engines; in fact, there are no four-stroke engines on the market with a bore exceeding 21 inches. elo ow this bore size, each cycle has its advocates. The two-stroke engine, by virtue of the greater number of power strokes per revolution, can develop equal output to a four-stroke-cycle engine a t lower mean cylinder pressures. On the other hand, it is necessary for piston rings to traverse ports in a two-stroke-cycle engine cylinder, which generally requires this type of engine to run at lower mean viston s ~ e e dthan a four-stroke6cycle . engine. The net effect -of these factors has been that over the years the two types of engines have been quite competitive in weight, size and performance, with one or the other sometimes ahead for brief periods due to a new invention or breakthrough. Engines are either liquid- or air-cooled. A comiderable amount of heat is generated in the cylinders and the temperature of the cylinder boundaries must be controlled to prevent them from exceeding safe limits.
depending on the number of piston strokes to complete one full cycle of operation, Fig. 1. In the two-stroke cycle, air is compressed in the cylinder during the compression stroke, fuel is injected, and burning takes place during the power (or expansion) stroke. Before the piston reachw its bottom dead center position, the gases are exhausted through ports or valves. Scavenging of the spent gases takes place during the period around bottom dead center, and then the fresh charge is compressed to start the new cycle. It can be seen that the entire cycle is completed in two strokes, one compression and one expansion; hence the name two-stroke cycle. In the four-stroke-cycle engine, the cycle also begins with a compression stroke, followed by fuel injection near top dead center, then by the power stroke. It is here that the cycles differ. Just before bottom dead center, the exhaust valves open and the gases start to discharge from the cylinder, the exhaust process continuing during the next stroke of the piston. At top dead center, the clearance volume between the piston and the cylinder head would be filled with exhaust gas; however, the intake valves open slightly before top dead center and the remaining exhaust gases are swept out of the cylinder by fresh air. The exhaust valves close slightly after top dead center, and the continued outward movement of the piston draws in a fresh charge of air. The period at top dead center, when both the intake and exhaust valves are open simultaneously, is known as the
atmosphere by means of a water-to-air heat exchanger such as an automotive-type radiator. I n small sizes, air-cooled engines may be used; in fact, they can be very attractive in those applications where it is easy to get air to the engine and where the operating locale is such that sea chest clogging is a problem. Engines may be arranged with their crankhafts horizontal or vertical, although the greater number by far are installed horizontally. Cylinders may be arranged in a line or with banks of cylinders in the form of a V, W, or X. In-line and V-type engines are the most commonly used (see Fig. 2). W and X cylinder arrangements permit more compact designs which take less space than the other types; however, access is more difficult, paintenance work is harder to accomplish, and a casualty is more likely to result in extensive damage. Opposed piston engines are two-stroke-cycle engines with two pistons working in a common cylinder. Cornpression takes place between the pistons at their inner dead center position, air intake is through ports at one end i f the cylinder, and exhaust gases flow out through ports at the other end. These engines are usually provided with two crankshafts, one at each end of the cylinder, although one crankshaft is sometimes used together with a rocker arm at one or both ends of the cylinder. Opposed piston engines are commonly of the in-line type; however, they may be arranged in other forms. One of these engines is built with three crank-
The "torque" of the output shaft in units of lb-ft can be computed from the expression
Medium and .High-Speed Diesel
Enkines
T
bhp = 5252 -
marine engineer with information relative to the application of medium and high-speed diesels to ships; a similar coverrlge of low-speed diesels is presented in Chapter VIII. Only those design details which affect the selection, installation, operation and maintenance of a diesel in a ship will.be discussed. Information will be preaented which will enable a ship designer to select the to plan the installation proper engine and its with due consideiration for operation and maintenance, and to prepare specifications to adequately describe the equipment desired. The largebore slow-speed diesel engine employed in many merchant ships is given a comprehensivetreatment in Chapter VIII and, therefore, is not discussed here. 1.2 Descriptioh of the Diesel Engine. The term diesel engine is used to designate any engine in which air is in a cylinder sufficiently to produce spontaneous ignition of the fuel, followed by injection and burning of a measured amount of fuel, the fuel in common use being oil. Although more properly d e s k nated as comptession ignition engines, they are manufactured and sold as diesel engines. The diesel engine is generally a reciprocating engine in which the gas pressure in a cylinder acts on a piston to drive a crankshaft through connecting rods. The power is taken from the crankshaft. The pistons move in the cylinder between the top (or inner) dead center and bottom (or outer) dead center positions. The distance between these dead center positions is known as the "stroke"f the engine, and is numerically equal to twice the radius of the crankthrow of the crankshaft. The diameter of the cylinder is known as the "bore." The bore and stroke are usually expressed in inches. Air is introduced into the engine cylinder through intake valves or ports and then compressed, raising the pressure and temperature of the air. . The "compression ratio" of an engine is the ratio of the volume of the cyliader when the piston is at bottom dead center to that when at top dead center. The term "compression ratio" can be misleading in that it is a volume ratio and not a pressure ratio. The compression ratio of an engine must be sufficiently high so that the air temperature at the end
(3)
-
Section 1 lntrod~~ction Scope. This chapter is intended to provide the
N -.
For a given engine, torque and BMEP are directly proportional; i.e.,
T=-
1.1
247
of compression will ignite the fuel when it is sprayed into the cylinder. Injection of fuel into the cylinder starts somewhat before top dead center and continues for a period of time, which varies with the engine power output. Combustion in the cylinder lags the start of fuel injection by a ~ e r i o dknown as the "ignition delay." Combustion raises the temperature and pressure of the gas in the cylinder, which then forces the ist ton to the bottom dead center position, doing the useful work of the cycle. The burned gases are then expelled from the cylinder through e~haustvalves or ports aad a fresh. y the charge of air is admitted to ~ 0 m ~ l e t e lscavenge cylinder of spent gases ~ r i o to r the start of a new cycle. Some additional terms which are frequently used in dexribing diesel engines are defined as follows: The "displacement" of an engine is the swept volume . is expressed in cubic of all the engine ~ ~ l i n d e r sIt inches as:
where n = number of cylinders in engine B = bore, in. s = stroke, in. The physical size of an engine is approximately proportional to its displbcement. The "brake mean effective ~ressure" (abbreviated BMEP) stems from the days when it was common to take indicator cards of the presrures in an endne cylinder, and to relate the severity of engine loading to the average or mean pressure in the cylinder during one cycle. The BMEP is still used as an indicator of engine loading and is expressed in psi as:
where
C
= number of strokes per cycle (two for %stroke;
four for 4-stroke)
I
4
37.7 C
BMEP
(4)
-
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE ENGINEERING FUEL
VE
EXHAUST VALVES
CONNECTING
START OF COMPRESSION STROKE
!SCAVENGING PROCESS
POWER STROKE COMPLETED EXHAUST IN PROCESS
IN-LINE ENGINE
VEE ENGINE
Fig. 2
TWO STROKE CYCLE
Common engine cylinder arrangements
,. ::'
I
VALVES: INTAKE EXHAUST\
OPPOSED-PISTON .ENGINE-
FUEL
E l I
E
kNOZZLE E
I
E
I I
i
I
INTAKE STROKE
COMPRESSION POWER STROKE (OR EXPANSION) STROKE
EXHAUST\CRANKSHAFT STROKE
FOUR STROKE CYCLE Fig. 1
Two- and four-rtrdte-cycle diesels, cycle evenh
The air for combustion is supplied at the relatively low exhaust conditions is insured. Two-stroke cycle engines pressure of from two to five psig by a scavenging blower benefit from charge air cooling in the same manner as of the positivedisplacement or centrifugal type. The four-stroke cycle engines. A marine engine lends itself scavenging air pressure required is a function of the particularly well to the use of air cooling becciuse of the arrangement and size of the air and exhaust ports (or availability of an ample supply of cooling water. valves) and passages and the speed of the engine. Diesel engines can also be classified according to the With supercharged engines, the combustion air is manner in which they are started. To start a diesel supplied by a compressor of the positive-displacement or engine, it is necessary to rotate it using an external centrifugal type driven from the crankshaft by gears or source of energy so as to bring the engine speed up to a driven directly by an exhaust gas turbine connected to point where the compression temperature of the air in the compressor shaft. The latter arrangement is called the cylinders is high enough to ignite the fuel when it is a turbo-supercharger or, more commonly, a turbo- injected into the cylinder. Once the engine is started, charger. Four-stroke-cycle engines which @re super- the external energy source can be secured and the engine charged by a gear-driven compressor are rare; practically will continue to run. Starting systems are classified by all supercharged four-stroke engines employ turbo- the energy source and the method of applying it. chargers. Some engines use two-stage superchargers Energy sources are high-pressure air, high-pressure with a charge air cooler between the two stages. T&ir hydraulic fluids, or electric power. Methods of applic* cooler ---.-serves .- two functions: it- increases the densitxof tion are by starting motors or using the engine cylinders ,."-the air charge in t6e qrliyder, ,e@iG&ng- thg .engine tta themselves. Air-started engines can use either a b u m z o r e fuel; and it lowers the temperature of the air motor geared to the engine crankshaft --- 3ir rotary-type in t h e 2 n a e r at the beginning-"of< c o s m ~ r en~ ds ~ ~through a disengaging type of drive similar to the throuiout-%i;eerEimainder of the gycle. It is more common automobile engine starter drive, or by admitting -common to use a single-stage supercharger with an air the air dirytly into some or all of the engine cylinders cooler (called an aftercooler) between the compressor through a specially provided air starting valve. These discharge and the engine intake manifold. The two- valves are controlled by a distributing valve which times stroke-cycle engine has a lower exhaust gas temperature their opening to occur just after top dead center on the and less energy in the exhaust gas entering the super- power stroke, and they are designed to close automaticharger turbine. At part load, there may not be cally when the engine fires. Direct cylinder starting is enough energy in the gases to drive the supercharger at rarely applied to engines with a bore less than six inches, the speed required to furnish the engine with sufficient and many engines up to nine-inch bore are started by the air for proper combustion of the fuel. In such a case, it use of starting motors. Hydraulically started engines is necessary to either provide a first-stage blower all use starting motors. Electrically started engines use geared to the engine [5] or to gear-drive the turbo- starting motors, although starting windings may be charger from the engine crankshaft through an over- incorporated in the directly driven power generators running clutch [6]. These arrangements are also for this purpose. beneficial in that rapid acceleration under smokeless Engines may be either unidirectional in rotation or r
__L vmI-
r / -
-* I--I.
shafts and has cylinders arranged along the sides of an equilateral triangle. It should be recognized that the diesel engine need not be of the reciprocating type. Various attempts have been made, and are continuing, to develop a rotary type of diesel engine; however, none have reached the stage of commercial production. I n these machines, a rotor is substituted for the piston, and the cylinder becomes a chamber of other than cylindrical shape. One such type of engine is described in reference [I.].' The description and definitions given in the foregoing must of necessity be modified to make them applicable to engines Numbers in brackets deaignste Referehces at end of chapter.
of the rotary type. Many examples of the various types of engine forms may be found in the literature [2, 3, 41. Another feature of diesel engine design which serves to differentiate between engines is the means of supplying combustion air to the cylinders. The alternatives are naturally aspirated, scavenged,and supercharged engines. With naturally aspirated engines, air is drawn into the cylinder as the piston moves from top to bottom dead center. The pressure in the cylinder at the start of the compression stroke is below atmospheric, due to the pressure drop through the intake passages and valves. A scavenged engine is the two-stroke counterpart of the naturally aspirated four-stroke-cycle engine although the cylinder pressure at the start of compression may be somewhat greater in this type of two-stroke engine.
A
--__I/
250
s
direct reversing. A unidirectional engine, as its title implies, can turn in only one direction; and if it drives an output shaft which must be capable of rotation in either direction, the engine must be connected to the shaft through a reversing device. Direct reversing engines can rui in either direction and most of them can deliver full power in either diction. If it is necessary that full power be available in either direction of rotation, pro&uement specifications must so state to insure that the engine selected does have this capability. To change the direction of rotation, it is necessary to bring the engine to a complete stop and then &art it in the opposite direction. The details of the reversing process will vary from one engine to another; however, the process is automatic and basically consists of cutting off the fuel to the cylinders by moving the throttle lever to the stop position, changing the timing of the fuel injection pumps if necessary, changing the timing of the exhaust and intake valves if used, repositioning the blower reversing - valve on two-stroke-cycle engines having geared, ' positive-displacement, rotary-type scavenging blowers, and reversing the rotation of the starting device. When the starter is energized, the engine should then start and run in the opposite direction. Engines are referred to as being high, medium, or low apeed. There is no clear line of demarkation between the classifications, but in general, "theycan be categorized aa shown in Table 1. There is no unanimity among Table 1
t
.
Low apeed.. ....... . Medium speed. . . . .. . High speed.. . . ..... .
Engine Speed Classifications Piston apeed, fpm
Shaft speed, rprn
1000-1M)O 1200-1800 18003000
100414 700-1200 1800-4000
engine people as to the significance of engine speed. A welldesigned high-speed engine which is not overloaded can give equally good service as a slow-speed engine. Slow-speed engines are of larger size than high-speed engines, but wear rates are comparable; hence it takes longer for a slow-speed engine to wear parts to the same percentage of their original dimension. A balance must be struck between the use of a smaller, lighter, and generally leas expensive high-speed engine and a larger, heavier, slow-speed engine which usually costs more initially but has lower fuel, operating, and maintenance costs. 1.4
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE EN(
Special Requirements of Marine Diesel Engines.
The number of medium and high-speed diesel engines used in marine applications is relatively small compared to the total number of such engines produced, and for this reason it is economically unattractive to produce these engines for the marine industry alone. The medium and high-speed marine engine of today is, therefore, almost universally an adaptation of engines which are built in quantities for service in automotive applications such as trucks, buses, off-highway earthmoving equipment and locomotives, and stationary
applications such as municipal power plants, mobile emergency power sources, gas line compressor stations, and pumping units. The truck and bus field contributes the high-speed engines in the range of 300-400 hp. For intermittent use, maximum speed of these engines will approximate 3000 rpm; for continuous service, speeds of 1800' rprn are common. The off-highway equipment engines are in the power range from 500 to upwards of 1200 hp.and speeds in the 1200-1800 rprn range. Diesel locomotive engines are available in units from 6 to 20 cylinders and ratings up to 4000 hp at speeds from 850 to 1100 rpm. Another group of engines coming from the stationary field includes units in the speed range from 300 to 514 rprn and powers up to 7500 hp. A comprehensive description of a 400-rpm engine rated at 1000 hp per cylinder with a range of cylinders from 6 to 18 is given in reference [7]. Some of the engines just mentioned were designed with marine applications in mind, others require some degree of modification for installation aboard ship. These rnodications are usually in the external hardware of the engine and do not involve changes to the internal working parts, which have undergone extensive development. The changes, where necessary, are those needed to suit the engine to the marine environment, meaning salt-laden air, high humidity, use of corrosive seawater for cooling, and operating from a nonhorizontal platform which is in constant motion (i.e., pitching or rolling at all times). I n many cases, it also means an installation made in confined spaces. I n order to adapt to this environment, the prime requisite of the marine diesel engine is the ability to resist corrosion. Nonferrous alloys are used in many places in a marine engine for corrosion protection where ferrous metals are used in nonmarine installations. To this end, aluminum parts which are exposed to the atmosphere and are not normally coated with lubricating oil should be given an anodic treatment and then painted. All exposed ferrous metal parts should be painted. Care must be taken to insure that only compatible metals are used in the water system. If two metals which are far apart in the electromotive series, such as aluminum and steel, must be used contiguously, they must be insulated from each other. The velocity of seawater through the piping system must be lower than that used in freshwater systems to prevent excessive erosion. Marine engines may be installed with their crankshafts at an angle to the horizontal. For this reason, and because they are subjected to more motion than in many other applications, changes are necessary in the lubricating oil system. Where there is room under the engine, the simplest solution is to use a deep oil pan. This has two beneficial effects: first the oil sump capacity can be increased, permitting longer oil change periods; and secondly, the oil level can be maintained low enough such that the connecting rods will not dip into the oil, thus preventing oil from leaking past the crankshaft end seals. Where there is insufficient space for a deep oil pan, it is necessary to use a shallow pan and a dry sump
system. Details of lubricating oil systems are covered in a later section. While it might seem that the air intake to a marine engine would be dustfree and dirtfree, this may not be the case when operating in harbors, inland waters, or close offshore. In these cases, it can be just as important to provide a good air cleaner as in any automotive or stationary installation.
25 1
The ciose confines of many marine engine rooms gake it especially important to protect personnel from injury due to contact with hot or moving parts. Partic. ular attenti~nshould be paid to adequate shielding for these hazards. It is also extremely important to prevent fires in machinery spaces; to that end, care must be taken to insure that fuel and lubricating oil cannot be sprayed against hot engine surfaces.
Section 2 tharacteris?ics of Diesel ~ n ~ i n e s 2.1 Performance Characteristics. The engine performance characteristics which are of interest when selecting an engine for a particular application are torque, horsepower, fuel consumption, and q eed. The torque output of a particular engine is control ed by the quantity of fuel injected into the cylinder each cycle; in fact, the torque varies almost directly with the quantity of fuel injected per cycle. The maximum torque that an engine can develop at any speed is usually limited by the exhaust smoke condition comidered acceptable, high stress, or high temperature rather than by the engine's ability to pull more load. The maximum power that the engine can develop a t any speed is simply the product of the maximum torque, the speed, and a constant. It is important to know the engine's characteristics and how they are related to the conditions under which the engine is to be used in order to insure a successful application. The diegel is generally referred to as a constant-torque machine, ltnd it certainly is when compared to steam or gas turbines which have stall torque ratios in the order of 3: 1. The stall torque ratio is the ratio of the torque at stall speed (i.k., zero for steam turbines) to that a t rated speed. The torque of a diesel running a t a constant throttle setting will normally rise to about 110 percent of full-load torque in the range of 55 to 70 percent of full-speed rprn and then drop aa the speed is further reduced, as shown by Fig. 3. If the throttle setting is reduced such that the'quantity of fuel injected per cycle is reduced, the engine torque is correspondingly reduced. Typical curves illustrating this trend are also shown on Fig. 3. It is possible to modify the torque characteristics of a diesel by changitlg the fuel injection versus speed chaxactei.istics so as to increase the quantity of fuel injected per cycle as the speed is reduced, whiie at the same time using a turbochaxger which has been matched to the engine for optimum efficiency at the speed for which the high torque is desired [8,9]. I n this manner, a peak torque as high as 140 percent of full-load torque of the normal engine can be obtained at speeds as low as 60 percent of full speed as indicated by Fig. 3. For normal marine drives such as propulsion, generator sets, or centrifugaJ pumps, there is no need for an engine with
P
i
* I
i
j
specid torque characteristics. However, when a diesel is used to drive $winch or a positive-displacement pump, there may be an advantage in having an engine which is designed to develop high torques at low speeds. Such an engine is referred to as one with "lugging capacity." At a constant throttle (or fuel rack) setting, the engine power is, for all practical purposes, a linear function of engine spqed. As indicated previously, the power at any speed is usually limited by factors other than the amount of fuel which can be injected or burned in the cylinder. The manner in which the engine power is limited by the variods parameters is shown in Fig. 4.
NORMAL ENGINE 100%FULL THROTTLE NORMAL ENGINE 80%FULL THROTTLE HIGH TORQUE ENGINE IOOXFULL THROTTLE
0
20
40 60 RPM, % RATED'
Flg. 3 bigin. twque ckarackrMio
80
---- -
100
252
,
MEDIUM AND HIGH.-SPEED DIESEL ENGINES
MARINE ENGINEERING
TURBOCHARGER I SURGE OR SMOKE LIMITED f Lli"ED ~
Fig. 4
1
0 20
II
40
I
60 ENGINE
80 RPM,% RATED
The curves in each case are drawn against a background of constant torque (or BMEP) lines. Data for each curve are developed by regulating the amount of fuel injected per cycle to maintain the parameter constant. Not all engines are designed to withstand the same cylinder pressures, exhaust temperatures, or maximum speeds, and not a11 are smokefree to the same degree. The maximum horsepower is shown by the solid line, and each curve segment is labelled to show the factor which limits the power. The shape of the maximum horsepower curve is generally applicable to all engines; however, the limiting factor in each portion of the speed range may vary from one engine to another [lo-131. Additional factors which could limit the power output over parts of the speed range include the temperature of parts (i.e., pistons, cylinders, heads, or valves), bearing loads, deterioration of lubricating oil, or turbocharger rpm. The fuel consumption of a diesel engine may be illustrated in several ways; the most useful depends upon the intended application and the preference of the user. At a constant speed, the total fuel consumption
I 100
Engine power limitations
I 120
varies almost linearlv with e h ~ h load. e but this relationship becomes nonlinear a t hig%er lo&' as shown by Fig. 5. The variation of specific fuel consumption with speed and load is more commonljr presented in the forms of Figs. 6 cr 7. The curves shown in Fig. 6 are commonly called fishhook curves and show the variation of specific fuel consumption with horsepower for various constant engine speeds. This type of ,curve is of interest for constant-speed applications, such as driving ships service generator sets. The presentation in Fig. 5 is of value in estimating fuel consumption of an engine over a range of conditions when actual data are available for a few points. Figure 5 is also useful during trials for estimating engine horsepower from measured values of fuel consumption and engine rpm. The method of presentation shown in Fig. 7 is particularly useful in the analysis of engines for propulsioh use, as it is possible to superimpose the required power versus speed curves on these fuel consumption curves to determine if the engine size and characteristics are properly matched to the load. The curves of Fig. 7 are typical of normal engine design; however, it is possible for the performanae
map to look as different as Fig. 8 when variable fuel and valve timing features are incorporated in the engine design [14]. Engine manufacturers will generally publish some of these curves in their sales literature. Specific information must be obtained for each specific applic* tion; however, Fig. 9 may be used for estimating the part-load specific fuel consumption corresponding to a propeller load curve if specific engine data are lacking. The specific fuel consumption a t full power will vary from 0.34 lb/hp-hr for we/l-designed medium-speed engines to 0.42 lb/hp-hr for high-speed engines operated near their maximum ratings. Knowing the intended application and the type of engine to be used, an approximate full-power fuel consumption can be selected from this range. A diesel engine has a definite limitation regarding the lowest speed at which it can be operated. This limit* tion can be influenced to some degree by an appropriate design of the engine or its installation. In general, the idling speed of a diesel engine is about 30 percent of rated speed. High-speed, low-horsepower engines may idle a t speeds up to 50 percent, while larger, heavier engines may idle below 25 percent of rated speed. Limitations on the idling speed are associated with the fuel injection equipment, combustion, and the inertia characteristics of the engine and driven machinery. Low-power, highspeed engines require small quantities of fuel to be injected each cycle a t full load; therefore, a t part load it becomes extremely difficult to accurately meter the smaller quantities of fuel required. I n addition, the temperature of the compressed air in the cylinder is less at low speed than it is a t high speed; consequently, combustion can become erratic. I n larger engines, the problem of injecting small quantities of fuel is not as severe, and it can be further reducd by the use of multiple pumps or multiple injection nozzles for each cylinder, or fuel pumps incorporating two plungers of different sizes. Excessively large variations in idling speed may be eliminated by increasing the she of the flywheel. Most medium and high-speed diesels are provided with attached lubricating oil and cooling water pumps driven from the engine crankshaft by gears or belts. These pumps are normally sized to provide the quantities and pressures of the work* fluid to meet full-speed and load requirements as well' as those a t low speed. If unusually low operating speeds are necessary for a given application, the normal pumps may not provide adequate lubrication or cooling. In these cases, the engine manufacturer must be alerted so that either larger pumps, special gear ratios in the pump drive, booster pumps, or separately driven pumps can be provided. I n addition to the problem of idling speed, some consideration must be given to the question of prolonged operation a t light load. As shown in Fig. 4, there is a minimum load below which combustion becomes unsatisfactory. I n this region, unburned or partially burned fuel will remain in the cylinder and wash lubricrating oil from the cylinder wall and find its way into the crankcw, diluting the lubricating oil. Both of these
'"
253
BRAKE HORSEPOWER. % RATED Fig. 5
Fuel consumption versus horsepower
BRAKE HORSEPQWER. % RATED Fig. 6
"Fishhook" curves, specific fuel consumption versus broke horsepower
e
0
I
I
PROPELLER LOADCURVE I
20
40 60 80 ENGINE RPM. % RATED
I
120
100
60.7 'Typlcol apeciflc fuel conrumption mop
"
254
MARINE ENGINEERING
MEDIUM AND HIGH-SPEED DIESEL ENGINES 0
E
110
2 I00
1.
I
I
I
I
MAXIMUM
I
Fig. 8 Spedal rpecitlc fud awaunptim map ENGlNE RPM.X RATED Fig. 9 Eltimafa ot part-load fuel ansumption on propeller load cune
30
I
I
40
50
I
I
I
60 70 80 ENGINE RPM,%RATED
actions tend to increase the wear of engine parts. The partially burned fuel results in increased quantities of carbon and lubricating oil in the engine exhaust passages, particularly so in two-stroke-cycle engines. Operation at elevated loads e l l remove the exhaust system deposits by burning and cause visible smoke in the exhaust for a short period of time after the load is increased. Here again, the engine 'manufacturer is in a position to minimize the adverse effects of light-load operation if he is forewarned that the condition may exist. A diesel engine is normally extremely easy to start. Engine controls are arranged to properly sequence all operations so that starting is simple, reliable, rapid, and automatic. Full load can be accepted immediately; however, if there is no urgency, it is preferable to allow the engine to reach operating temperatures before subjecting it to full load. At ambient temperatures normally encountered in enclosed engine rooms, engines can usually be started without startiqg aids. However, if it is necessary to start the engine in an unheated engine room in the winter, provision should be made for the installation of a starting aid such as the injection of ether into the intake air. Small high-speed engines of the automotive type are usually built with higher compres-
I 90
I 100
I 110
sion ratios than other engines and, in spite of the greater loss of heat during the compression stroke, they will start, at lower temperatures than their larger slowerspeed counterparts, without the use of starting aids. The ability of engines to start when cold varies widely; some are able to start a t temperatures as low as zero F, while otbers may have difficulty at 50 P. If cold starting may be required in service, procurement specifications should make this fact known to the manufacturer. The acceleration characteristics of a diesel are determined by the difference between the torque available and the torque load on the engine at any speed. If the engine is operating near its maximum torque at a given speed, it will accelerate very slowly should an increase in speed be demanded. Or, in the case of a generator set, if the operating load is n e p the maximum, an increase in load may cause a drop in speed and a sluggish return to the desired speed notwithstanding attempts of the governor to effect an early correction. Turbocharged engines may be sluggish in response to load changes if the turbocharger rotor inertia is excessive. If fast acceleration is a requirement of the application, again the engine manufacturer should be so advised. He may be able to supply engines with multiple low-inertia-turbo-
chargers or special controls to provide the desired characteristics. 2.2 Engine Ratings. The rated horsepower of an engine is the power output capability of the engine a t rated speed under specified ambient conditions, duty cycle, and life expectancy as proven by performance, endurance, and environmental testing. To facilitate in the selection of the correct engine for a given application, engine manufacturers publish rating curves for each of their engine models. For most high-speed engines, these curves are of the form illustrated by Fig. 10 and show the recommended rated horsepower for three different operating conditions; namely, maximum, intermittent, and continuous [15]. The limiting horsepower lines from Fig. 4 are included in Fig. 10 for comparative purposes. The maximum horsepower is useable only for special applications where high power is required far short durations. This maximum horsepower serves as a baseline for selecting a rating suitable for a particular application and is determined by tests on a dynamometer in the manufacturer's plant under rather ideal conditions. I n actual service, less ideal conditions prevail; operators may be unskilled, loads may be unexpectedly high or suddenly applied, or extreme operating temperatures may be encountered, any of which can shorten engine overhaul periods, increase wear and cause unexpected failures. To insure satisfactory service performance, the engine is usually rated a t a performance level less than the maximum. For intermittent duty, such as may be expected in a pleasure boat gr for stand-by service, the engine is usually rated at approximately 85 to 90 percent of the maximum horsepower with the speed rated at the maximum value. For continuous duty where the engine will operate for long periods with little downtime or where the load on the engine is a high percentage of the rated load at all times (such as for ships' service generators or workboat propulsion), more conservatism is exercised in
ENGINE RPMSX RATED Rg. 10 Typlcal diesel mcmvfadurwk rating curves
rating the engine. The continuous-duty rating curve is usually 70 to 75 percent of the maximum horsepower, and the rated speed is limited to approximately 90 percent of the maximum. The precise reduction from the maximum rating varies from one application and manufacturer to another; however, the figures stated are representative. The manufacturer's rating curves present the engine performance at standard conditions of atmospheric presaure and temperature, with a simple exhaust system, and with a minimum of accessories. Manufacturers of medium and slow-speed engines generally do not furnish three curves as in Fig. 10,and it is necessary to determine the operating conditions applicable to the quoted rating or rating curve. I n applying these curves or ratinge in the process of selection of an engine, it is necessary to make corrections for atmospheric conditions if the intended installation will impose conditions on the engine which differ from the manufacturer's standard. It is also necessary to reduce the rating by the power required to drive contemplated accessories which were not included during the standard dynamometer tests. These
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE ENGINEERING Table 2
Standard Conditions and Correction Factors for Engine Tests Specification for arine Diesels 100 29.0 1.0 bhpc = C(bhp.)
Mi1it2 Ambient temperature de F Barometric p r e s m (dvf in. Hg Ekhaut back pressure, m. Hg Correction formula Correction factor, C
SAE Test Code J816 [la] 85 29.0. not specified bhp, = C(bhp0 fhp) 29 To 480
+
+
Secl:ion 3 Marine Uses for Diesel Engines
- fhp
E(T)
Po
where bhp, = corrected brake horsepower bh = observed brake horsepower = observed barometric pressure (dry), in. Hg To = observed ambient temperature, deg F fhp = friction horsepower e
a
accessories may include such items aa reverse and reduction gears, battery charging generators, air cornpressors, hydraulic system pumps, or bilge Pumps. Performance redu~tionmust also be made if a complex exhaust system imposes an unusually high back Pressure, or if intake air silencers or cleaners restrict the flow of air to the engine. This is particularly i ~ p o r t a nin t the case of turbocharged engines. Expected service loads and installation conditions must be carefully investigated and specified during the early design stages. It is also necessary to consider the rating and the service history of the selected engine in other applications. A diesel engine is usually capable of operating at a load in excess of its rating, and overloading may not be immediately apparent; however, it eventually can become evident in the form of shorter overhaul intervals, unexpected failures, and higher maintenance costs. There is no standard method of rating a diesel engine; even the baseline conditions of atmospheric temperature and barometric pressure are not standardized, although there are standard conditions set by military specifications and the Power Test Codes of the engineering societies such as the Society of Automotive Engineers (SAE) or the American Society of Mechanical Engineers (ASME). When the anticipated operating environmental conditions differ from the manufacturer's test conditiom, it is necessary to correct engine performance from the operating conditions to standard conditions in order to properly select an engine. If acceptance tests are conducted with conditions other than standard, the test results must be corrected to insure that contractual requirements are met. Unfortunately, no generally accepted correction factors exist. The SAE test code has a correction factor for naturally aspirated engines, but not for turbocharged engines. The military specification for marine engines has a different correction factor which is used for all diesel engines. To avoid misunderstandings, it is necessary that ship specifications and engine procurement documents specify both the standard conditions under which the engine is to be rated and the method by which test data are to be
corrected to verify performance. Two examples of standard conditions and correction factors are shown in Table 2. Corrections to fuel consumption to compensate for differing atmospheric conditions are more complex t h a ~ for horsepower and are not generally used for thm reason. The increasing use of computers to analyze automatically recorded test data may produce relationships which will lead to the publication of generally accepted correction factors for the performance of all types of engines. 2.3 Engine Physical Characteristics. For estimating purposes, an engine weight of 4 lb per cu in. of total displacement is a reasonable approximation. If there are special reasons to require an engine of lighter weight, there are engines in production weighing about 3 lb per cu in. in most sizes; however, the number of available suppliers would be reduced considerably. To estimate the total engine weight of an engine of given horsepower, calculate its approximate displacement, assuming the type of cycle, and values of BMEP and rotative speed appropriate to the intended service, and substitute these in the equation (see Section 1.2 for a definition of
3.1 Types of Ships Employing Diesels. Diesel engines have been utilized in all types of ships, both in the merchant marine and in the navies of the world. The power range in which diesel engines have been used in American-built ships has increased directly with the availability of higher-power engines. The line of demarkation in horsepower between what is normally assigned to diesel and to steam has continually moved upward; however, so has the power installed in ships of a given type. For example, Navy oceangoing tugboats of 25 years ago were powered by four diesel locomotive type engines, each of 900 bhp, for a total of 3600 bhp. But after 25 of improvements, four engines of the same basic type now power a commercial tug with a total of 9600 bhp. This same type of engine and others like i t are available up to 4000 bhp and, no doubt, will be used in tugs of tomorrow if 16,000 bhp can be usefully
ears
as a baseline. The lower-power engines were sized by scaling from the 5000-bhp size, keeping the BMEP and piston speed constant; they are hypothetical engines, but engines are available of approximately the ratings and sizes shown. When more than one engille is geared to the propeller shaft, it can be seen that the gear serves as both a speed reducer and combining gear. The same series of engines could be used in an electric drive propulsion system, with even greater flexibility. Each engine drives its own generator and may be located independently of other engines and the propeller shaft. The enerators provide the power to drive a propulsion motor or motors as the case may be. A single motor may be used directly connected to the propeller shaft, or it may be geared to the shaft. On the other hand, it may be preferable for several smaller motors to be geared to a single propeller shaft. It was previously noted that diesel engines could be
terms>
The displacement is then multiplied by the appropriate ratio of engine weight to displacement (generally 4) to determine the engine weight. The space requirements of a diesel power plant are rather flexible in that it is possible to assemble a plant from one or more units and to select the type of unit to be used. If head room is a ~roblem,small high-speed engines can be used. If width is a ~roblem,in-line engines can be used. If it is necessary to minimize the length, vee-type engi~lesare available. ~ n g i n e scan be furnished completely assembled with all the necessary accessories mounted on the engine and its subbase, or with these accessories loose for mounting where space is available. It is extremely important that adequate t space be ~rovidedaround each engine to ~ e r m i access for maintenance. Fortunately, the space required for maintenance usually coincides with the envelope of the engine. Parts of high-speed engines are relatively small and light in weight; this facilitates handling and minimizes the need for extensive rigging for art removal.
3.2 Shipboard Applications of Diesels. Diesel engines are used either singly or in multiple to drive propeller shafts. For all but high-speed boats, the modern American diesel turns too fast to drive the propeller directly with good efficiency and some means of speed reduction, either mechanical or electrical, is necessary. If a single engine of the power required for a given application is available, then a decision must be made as to whether it or several smaller engines should be used. This decision may be dictated by the available space. Using a mechanical transmission system as an example, Fig. 11 illustrates the flexibility of the diesel power plant in adapting to specific space requirements. I n this figure, an engine with a rating of 5000 bhp irs used
\
In applications where it is necessary to provide rapid maneuvering characteristics with reverse gears or direct reversing bngines, brakes may be installed either on the propeller shaft or on the high-speed pinion shafts of the reduction gear to stop the propeller shaft in minimum time. Many direct reversing engines can be specially adapted to use starting air in the cylinders for braking purposes, and this possibility should be weighed against other means of shaft stopping. Diesel engines are used to drive shipsJ and emergency generators. Emergency generator sets are arranged to start automatically upon failure of the normal power supply, and after a builtrin time delay, assume the electrical load on the emergency bus. For many years, Navy specifications have required that
,I
I 1
I
I
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE ENGINEERING REDUCTION GEAR OUTPUT 170 RPM
I I V-16 ENGINE ( 5 0 0 0 BHP AT 514 RPM) HEIGHT 10'
REFERRED CURVE
2 V - 8 ENGINES (EACH 2 5 0 0 BHP AT 514 RPM) HEIGHT 9.5'
IT (0) .2V-16 ENGINES (EACH 2500,BHP HEIGHT 7.7
20
30
40
50 60 70 80 ENGINE OR PROPELLER RPM.% RATED
90
100
110
Fig. 12 Matching engine to l i p characteristicswing power cunes
I UI
I
2 V-16 ENGINES (EACH 2 5 0 0 B H P AT 7 3 0 RPM) H E I G H T 7.7'
38.5'
m 4 V-16 ENGINES (EACH 1250 BHPAT 1 0 3 0 RPM) HEIGHT 5'
-1
2564 .-'
Fig. 1 1
Compar'wn of various engine arrangements for 5000-bhp plant
emergency generator sets be capable of starting and assuming full load in no more than ten seconds, and it has been demonstrated that this is a reasonable require ment. It is possible to parallel a diesel generator set electrically with generators driven by other diesels or other prime movers such as steam or gas turbines; however, the equipment supplier should be made aware of this requirement if it is needed. Diesels are used as prime movers to power many different types of auxiliaries such as fire pumps, dewatering pumps, cargo oil pumps, compressors, and winches. For engines installed high in the ship, conideration should be given to the use of radiator-cooled or air-cooled engines to avoid cooling water pumps which would be subjected to high suction l i t requirements. 3.3 Selection of Engines. The selection of engines for shipboard use cannot be b-d on any single factor.
There are many possible engine deaigns which are capable of meeting most performance requirements, and numerous factors must be considered such as weight, fuel consumption, cost, availability of competitive engines, manning requirements in terms of skill level and number, and maintenance considerations such as availability of repair parts, necessity for special tools, and the number, type, and frequency of the maintenance required. References [17-221 discuss this subject in detail. The first and possibly the most important consideration leading to the selection of a diesel engine is the definition of what it must do. I n the case of a propulsion engine, this entails obtaining the speed-power curves for d l important modes of operation such as fully and lightly loaded, clean and fouled bottom, towing and running free, and with and without power takeoff loads. Additional information should be aster-
tained regarding the time duration of operations a t each condition. An assessment should also be made of anticipated special operating requirements. For exill it be necessary to spend long periods of time ample: W with engines idling? Will long periods of slow-speed maneuvering be required? Will the operation be primarily point to point with the engines a t full load and speed most of the time? Each of these questions and many more can influence the design of the diesel power plant. When the speed-power curve has been established, an engine can be selected which will develop the required horsepower a t its appropriate rating. Assuming that the ship under considerati~nis one which is expected to operate the majority of its time a t less than full load, the intermittent duty rating would be the appropriate one. A particular engine, or engines, is then selected whose intermittent rating is consistent with the full-power requirements for the ship. The intermittent horse power curve for the engine, similar to Fig. 10, and the light-load lines from Fig. 4 are then superimposed on the speed-power curve. Preferred, acceptable, and lightload operating regions are then added and the resultant plot is illustrated by Fig. 12. Operation in the light-load region should be avoided. The propeller load, curve A in Fig. 12, has been drawn with the power varying as the cube of the s p e d . It can b e seen that operation down to about 70 percent speed is within the preferred zone, and from 70 to 55 percent speed is in the acceptable zone. If ap appreciable amount of time is to be spent in operation below 55 percent speed, where the engine load as dictated by the speed-power curve faIls into the undesired rarige, consideration should be given to the use of two or more engines instead of one. Curve A represents the power to drive the ship with a clean bottom whether that power is produced by one engine or multiple engines. If
the performance of one of two installed engines operating alone is to be evaluated, it is necessary to redraw either the engine performance curves or the speed-power curve. Either the engine performance curves would have to be drawn with ordinates one half their original magnitude or the speed-power curve would have to be drawn with ordinates twice its original magnitude. It is simpler to redraw the speed-power curve, and this is shown as curve B. Now it can be seen t h ~ one t engine can be declutched from the propeller shaft whenever the ship speed is reduced to 62 percent of full speed (the intersection of curve B and the continuous-duty line). Under these conditions, the single engine would operate in the recommended zone, whereas two-engine operation would be in the acceptable zone. At speeds down to about 38 percent, the single engine would be acceptable, whereas two engines would be too lightly loaded below 55Ifpercent low-speed speed.operation is required for substantial periods of time, consideration should be given to a larger number of engines. Using the same procedure as previously, curve C has been drawn to represent the speed-power curve when operating on one fourth of the installed engines. I n this case, one engine could be used for operations up to about 40 percent speed, two engines from about 40 to 62 percent speed, three engines from 62 to 75 percent speed and all four engines above that. I n addition to the improved loading condition of the engines during part-load operations, benefits are derived from the fact that only some of the engines accumulate operating hours, and the total fuel consump tion is less. I n c w s where the speed-power curve can vary with conditions of operation (e.g., different displacements, water depth, hu\l fouling, towing), the extremes of loading should be considered when selecting the engine-
260
MARINE-ENGINEERING
MEDIUM AND HIGH-SPEED DIESEL ENGINES
have a power rating less than 10 percent of that of the main engine and be disengaged when the main engine is used. The quick starting capability of the diesel obviates the necessity of keeping engines running at idle just so that they will be ready when needed. The characteristics of diesel engines and the principles
26 1
governing their proper selection and application have deliberately been expressed in general terms. By following the methods described, unusual applications such as the engine requirements of planing hull boats or hydrofoils can be handled as well as the more conventional ships and boats.
Section 4 Design Considmtions
0
20
40
60
80
100
ENGINE RPM.% RATED Fig. 13 Matching engine to ship characterlrtiu uaing torque CUNW
propeller-reduction gear combination. Curves A' and A" which represent these extremes have been added to Fig. 12 to show the effect on performance. If the ship were designed to absorb full power under the conditions of curve A and then were required to tow a load such that the total resistance corresponded to curve A', the maximum speed permissible would be 85 percent of rated (the intersection of curve A' and the intermittent rating curve); the limiting factor would be engine torque. If, on the other hand, the resistance were reduced to that shown by curve A", no speed increase would be possible without overspeeding the engine, and full engine power could not be utilized. Under these conditions, the choice is dictated by the condition under which it is most important that full power be developed. If full power is required under both conditions, a controllablepitch propeller or a two-speed reduction gear must be used. Figure 13 shows the same conditions plotted with torque and rpm as coordinates to illustrate an alternative method which could be advantageous when most data are available in that form. The engine torque curves
shown-in Fig. 13 are not consistent with the horsepower curves in Fig. 12, which were drawn as straight lines for simplicity. In addition, specific fuel consumption curves have been added. The reduction in fuel consumption at low speeds which is obtained by operation with reduced numbers of engines may be verified from this plot. Figure 13 can also be used to verify that the minimum specific fuel consumption of the selected engine occurs at the ship speed and load most frequently expected. It is not necessary that all engines in a multi-engine drive be identical, although logistics problems are simplified if they are. There are cases where a considerable amount of low-speed maneuvering is required, and, if the required speed is below that corresponding to engine idling speed, the low speed can be obtained by the use of CRP propellers, two-speed transmissions, slipping clutches, or the use of a small engine which is geared to the propeller shaft such that it develops full power at a ship speed slightly above that corresponding to the idling speed of the main engine. The small engine may
4.1 Types of Fuel Used. One of the prime objectives in the development of the diesel engine has been to provide a prime mover which would be capable of burning a wide variety of fuels. It has, however, been necessary to compromise on this goal in order to achieve others such as reduced weight and space, increased reliability, lower wear of parts, good cold starting ability, and increased safety in fuel handling and storage. Over the years, a number of specifications for fuel oil have been developed to insure that the customer would be able to buy fuels meeting the requirements of various engine designs and t o give new engine designers a range of standard fuels from which to select. Operators who maintain a fleet of ships are particularly desirous of supplying one grade of fuel for all of their engines. The most significant characteristics of diesel engine fuels are listed in Table 3. The generally accepted uses for these fuels are: ASTM ID. A volatile distillate fuel oil for engines in service requiring frequent speed and load changes. The flash point of this fuel should be specified as a minimum of 140 F for marine applications. ASTM 2D. A distillate fuel oil of lower volatility for engines in industrial and heavy mobile service. Again a minimum flash point of 140 F is recommended for marine service. ASTM 4D. A fuel oil for low and medium-speed engines; however, it should not be assumed that all low and medium-speed engines will run successfully on this grade of fuel. The advice of the engine manufacturer ,should be solicited before using grade ASTM 4D fuel to insure that the particular engine model can tolerate the wider range of fuel properties permitted by this specification. MIL-F-16884, Marine Diesel Fuel. This Navy specification fuel is generally similar to ASTM 2D fuel except that a higher cetane number and flash point are specified and particular attention is paid to insure that fuels from different sources and lots will be miscible and that good st0rake stability is provided. MIL-T-5624, Turbine Fuel, Aviation Grade JP-5. This fuel is similar to ASTM 1D fuel except for its lower end point and high flash point. It has many require ments which are not tabulated in Table 3 inasmuch as they are needed primarily to meet aviation engine
reguirements, and they are not relevant for marine ap$ications. JP-5 fuel must be provided for turbinepowered aircraft o'perated from ships at sea; therefore, this fuel is used by the Navy in all diesels which are refueled at sea in order to simplify logistics problems. JP-5 can be used successfully in diesels while MIL-F16884 fuel cannot be used in turbines for aircraft use. The increased cost of the JP-5 fuel is offset by the advantage of having to carry only one grade of distillate fuel in tankers. The relationship between engine performance and some of the fuel characteristics specified in Table 3 is as follows: Cetane Number. Cetane number is a measure of the ignition quality of the fuel. Engine performance factors influenced in part by ignition quality are: (a) cold starting, (b) warmup, (c) combustion roughness, (d) deposits under idle and light-load operation, and (e) exhaust smoke density. Each of these performance factors is also affected by other fuel characteristics and engine design parameters. The cetane number requirements of an engine depend on design, size, mechanical condition, operating conditions, atmospheric temperature, and altitude. An increase in cetane number o?er values actually required does not materially improve engine performance. Heating Value. This important property of a diesel fuel is a measure of the energy available from it. The heating value of fuels may be expressed in either of two ways: high or gross heating value and low or net heating value, the difference being the latent heat of the water in the exhaust gas. Heating values may be expressed in terms of Btu/lb or Btu/gal. Since diesel engine fuel consumption is normally quoted in terms of lb/hp-hr and fuel is purchased on the basis of cost per gallon, Btu values on both a weight and volume basis are of interest. It is now customary to use the lower heating value for calculating thermal efficiency of diesels, although in the past the higher heating value was used. In either case, thermal efficiency is of academic interest only. For a comparison of the performance of different engines on different fuels, fuel consumption in terms of Btu/hp-hr is most useful, although care must be taken to insure that the heating values of the fuels are reported on the same basis. The heating value is specified in only one of the specifications listed in Table 3; this is because distillate
.
262
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE ENGINEERING Table 3
Ignition quality-*tam no.. ............... Appearance. .............................
Diesel Fuels ASTMD 976-64T 1D 2D 4D -40(d 40(a) 30'"
Marine Diesel MIGF 16884
JP 5 MILT 56246:
45
SEPARATOR
.. 1 1 1
.
DIESEL OIL SERVICE TANK I $DRAIN L. VLV.
IF VALVE IS USED ALSO INSTALL PRESSURE REL.VLV.
A
STRAINER
FILTER
r
Pour oint, F (mu). ...................... l (e )P.. ~ i ~ ~ ~F ~ ~ i..................... n t , 1.4 Viecoslty @ 100 F: centistokesmin. ......... m u . ........ 2.5 SSU min ............... gax .............. 34.4 0.15 Carbon residue on 10% bctttom, % m a . . ... 0.50 Sulpbur, % ( m u by we1 ht.. ............. Corrodon (mu) )at 212 Fy .;............... No. 3 Color (ma). ............................. Ash, .% (m)by weight. ............. : ... 0.01 Gmwty, API mm/max. ................... Acid number (max). ...................... Nel,ltrrtlity.. .............................. M m e pomt F. ........ !................. Accelerated stability%otal.................. insolubles, mg/100 ml (max). ............ Water and sediment b volume %. ......... TRACE Lower h e ~ t e "sale, g itu/lb (min). ......... Aniliue gramty product (min) ..............
125 or legal 2.0 6.8 32.6 45 0.35 1.0 No. 3
5.8 26.4
45
125
2.0
D.O. MANIFOLD
2.1 6.0 33
D.O. INJECTION PUMPS
45 0.2 1.0 No. 1 0.005 Record 0.50 NEUT Record 2.5
FINAL
k UNIT INJECTORS
0.4 No. 1
6
0.10
-
u
FUEL SUPPLY PUMP
(b)
Fig. 14 Dlagram of a typical fuel system
36/48
NOGS: (a) Lower tem rature or high-altitude operation may require higher cetane number. B below the minimum expected ambient. b S ecify at R r comparison not a apeyification d u e . d) For test methods, see specification referenced.
Go
diesel fuel properties such as volatility, viscosity, gravity, ignition quality, and heating value exhibit interrelationships. It has been established that certain characteristics of fuel can be estimated with reasonable accuracy from two or more measured characteristics such as volatility and API gravity. Charts ahowing these relationships may be found in the SAE Handbook [16] in the section on diesel fuels. For estimation purposes in ship design, a fuel with a representative higher heating value of 19,350 Btu/lb can be used. The corresponding lower heating value is 18,190 Btu/lb. Engine performance on the test stand is corrected to reflect the diierence in the heating value of the actual fuel and the standard value used in design. Viscosity. For some engines, it is advantageous to specify a minimum fuel viscosity because of the power loss due to injection pump and injector leakage. Maximum viscosity, on the other hand, is limited by considerations involving the engine design and size and the characteristics of the injection system. Sulphur. The effect of sulphur content on engine wear and deposits appears to vary considerably in importance and depends largely on operating conditions. It is important to maintain an engine jacket water temperature of at least 140 F to minimize the effects of sulphur in the fuel.
. WATER
PUMP
clear and brinht
Diatiition, 10% point F .................. Distillation, 90% p o w (mu). ............ 550 W i t i p n ; end point F (max) ............. Flash pomt, F (mip). ...................... 100 or
P
I
TRANSFER
Flash Point. The flash point as specified is not directly related to engine performance. It is, however, of importance in connection with legal requirements and safety precautions involved in fuel handling and storage and is normally specified to meet insurance and fire regulations. For marine use, a minimum flash point of 140 F is recommended. Pour Point. Pour point is important in connection with the lowest temperature which the fuel may reach and still be sufficiently fluid to be pumped or transferred. The pour point is generally interrelated with cetane number and volatility. Frequently, low pour pointa may be obtained only at the expense of lowering the cetane rating or increasing volatility. The pour point should not be specifled lower than required. For a more comprehensive dkussion of petroleum fuels, see Chapter 23. I n the design of a new ship, the selection of the fuel ts be used has an important bearing on the selection of engines and the detail design of the fuel handling and storage system. The fuel selection may be specified by the owner or left to the ship designer to provide greater flexibility in optimizing the total design. The selection of fuel for a given engine requires consideration of the following factors: (a) fuel price and availability, (b) maintenance considerations, (c) engine sire and design,
(d) speed and load ranges, (e) frequency of speed and load changes, and Cf) ambient conditions. 4.2 Fuel Oil System Design. The fuel injection system of a diesel engine is, in many respects, the heart of the engine. It must meter extremely small quantities of fuel, deliver the metered fuel at high pressure to the engine cylinder at exactly the correct time, in a precise spray pattern, and a t a specified time stop delivery abruptly and completely. The instantaneous pressure in the fuel nozzles can be as high as 40,000 psi a t full load in unit injectors (the fuel pump and nozzle are combined into one unit with no lengthy fuel line between). In the conventional system, the fuel pressure at full load may be as high as 15,000 psi for some engines. The duration of injection in a high-speed engine can be as short as 0.001 sec. With the high pressures involved and the precise timing requirements, it is necessary to build the injection equipment with close clearances and small tolerances. Nozzle hole sises vary upward from 0.005 in. dia, while the plunger-to-barrel diametral clearances, may be as small as 1.5 microns (0.00006 in.). I n view of these small clearances and high pressures, the most important consideration in the design and layout of the fuel oil handling and supply system for a diesel engine is to insure that clean, waterfree fuel is delivered to the engine. It is particularly important in 11mrine installations to insure that there is no salt water in the fuel at the time it gets to the injection pumps and nozzles. Saltwater-contaminated fuel has been known to erode the small holes in the fuel nozzles and cause pintle corrosion and sticking in a relatively short time, resulting in loss of power, burned pistons, high fuel consumption, and a smoky exhaust. In the typical fuel system illustrated in Fig. 14, diesel oil is transferred to a diesel oil service tank, sometimes called a day tank, after passing it through a water separating device which may be either a centrifugal purifier or a coalescing-type filter. Fuel flows from the
service tank through a strainer to a fuel supply pump which is normally attached to and driven from the engine. The fuel is discharged from the pump and flows through a filter and sometimes also through a final-stage filter before going to the fuel injection pump. It is customary for the fuel supply pump to have a capacity from three to four times that actually required by the engine. The excess fuel flows through the injection pump housing, cooling the plunger and barrel and insuring that the pump cylinder is completely filled at each stroke. The high-pressure fuel is discharged from the injection pump to the fuel nozzles in each cylinder of the engine through high-pressure tubing. Excess fuel flows through leak-off lines from the injection pump and from each fuel nozzle. The leak-off lines are manifolded to return excess fuel to the service tank, d i e charging above the fuel level and preferably against a horizontal b d e . If unit injectors are used, the pump and nodsle are combined in one assembly and there is only one leak-off point from each unit. It is preferable to have a separate return line from each engine to the service tank or tanks, with no valves in the lines. If it is necessary to install shutoff valves in the return line, a pressure relief valve should be installed to by-pass the valve and discharge to one of the service tanks in case the valve is inadvertently closed while the engine is running. If cocks can be installed to divert the flow to the proper tank with no chance of a line ever being completely blocked, the relief valve can be omitted. It is possible for the pressure to build up in a closed return line to the point of rupturing the pipe, spraying fuel into the engine room, and possibly starting a fire. Care should be taken to insure that leak-off lines have a minimum number of joints and that these joints are located so that leakage will not contaminate the engine lubricating oil. The choice between a centrifugal purifier and a c o s lescer-type water separator must be made for each
MARINE ENGINEERING
installation. The purifier can be of the self-cleaning type where the dirt and water removed from the fuel is discharged to a separate collecting tank, which requires infrequent cleaning. Purifiers are available in a variety of sizes, and one unit may be able to serve the needs of all the installed engines. The initial cost of a purifier is higher than that of a coalescer type; however, maintenance costs are lower and logistic problems are simplified. Care must be taken to follow instructions carefully and select the proper ring dams or discharge rings to suit the specific gravity of each fuel being centrifuged. The centrifuge can be equipped with its own heater and transfer pump to make it capable of handling heavier fuels. A coalescer has the advantage of being a static device with no moving parts to wear out, but it does have cartridges which require replacement. It cannot be used, however, with residual fuels or distillates contaminated with residuals. I n a coalescer, a combination filter and water separator unit is used. The oil with entrained water first passes through a phenol-impregnated paper filter element where solid contaminants are rempved, and the finely dispersed water droplets are induced to conglomerate a t an accelerated rate by intimate contact through the capillary openings in the filter paper. Some of these larger water droplets fall by gravity into the water collection sump along with dirt particles. The filtered and coalesced fluid then passes on to the separator unit. Its vertically pleated element of controlled porosity is impregnated with a hydrophobic material, such as molybdenum sulphide or paraffin for preferential wetting by the oily fluid, so that the oil and not the water globules passes through the capillaries. Water is collected in the sump of the separator unit and clean, waterfree fuel is taken off from a connection &,the top of the unit. If care is not taken to change elements when the pressure drop across them exceeds the recommended limit, there is the danger of rupturing the elements and contaminating the fuel in the service tank. The strainer has a metallic element of woven wire, stacked metallic disks, or sintered metal. Woven wire elements can remove particles down to about 40 microns, and if the joints in the wire are welded they can remove particles 2 microns in size. Stacked disks are capable of removing 40-micron particles and have the advantage that they may be made self-cleaning by rotating alternate disks. Sintered metal elements can remove particles in the range of 3 to 25 microns, depending on their density. Sintered metal elements &re difficult to clean and may disintegrate if subjected to'large pressure surges. The fuel supply pump draws fuel from the diesel oil service tank through the filter, and for that reason it must have the capability of operating with a suction lift of from 4 to 6 ft. If the suction lift is too great due to the elevation of the pump or the length of the supply line or the pressure drop in the filter, a separate motor-driven fuel booster pump may be required. The fuel supply pump is of the positive displacement type with pumping elements using either gears, vanes, plungers, or dia-
phragms. These pumps \\-ill have a discharge pressure of 6 to 20 psi for small engines and 25 to 40 psi for large engines. A pressure relief valve should be provided on the discharge side of the pump, either built illto the pump housing or installed separately in the discharge pipe. The fuel from the relief valve should return to the pump suction or to the service tank. The diesel oil service tank is normally located a t a level above that of the supply pump so that fuel can be supplied to the pump by gravity. In some engines, the fuel system is so designed that the fuel service tank must be located below the supply pump to prevent the flow of fuel by gravity into the cylinders of a shutdown engine. Air leakage into the fuel inlet lines can be very troublesome; therefore a minimum number of fittings should be used and all joints must be completely airtight. This is particularly important when the fuel tank is lower than the supply pump. Diesel oil tanks should not be made of galvanized steel because of the danger of forming corrosive zinc compounds in the fuel. Copper or silicon bronze should not be used for fuel tanks either, as their reaction with the mercaptan sulphur compounds in the fuel can result in the formation of damaging copper deposits in the engine combustion chambers. Aluminum bronze and manganese bronze are satisfactory for fuel tanks, as their use does not lead to these problems. Filtration is accomplished upon discharge from the supply pump in filters containing one or more elements made of either treated paper, felt, or woven yarn. The paper elements can be expected to filter particles in the range of 3 to 5 microns, with an initial pressure drop of from 0.5 to 2.0 psi and a pressure drop of between 15 and 30 psi a t the time of replacement. Woven yarn filter elements have a greater capacity to handle dirt, higher flow capacity, and somewhat coarser particle removal capacity. It should be noted that the characteristics of filter elements of any type can vary considerably depending on the filter design. Considerations with paper filters are the porosity of the paper and the material with which it is impregnated, and in the case of woven yarn filters, the tightness of the weave and the depth of the flow path. The particle removal characteristics of a filter should be expressed in terms of particle size and the probability that that size particle will be removed; for example, 2 microns 92 percent, 5 microns 95 percent, greater than 5 microns 99.5 percent. For most diesel engines, a progressive filtering system is used consisting of filters of increasingly fine filtering ability. First there is a strainer to take out large particles, then a yarn type filter to take out particles in the 25-micron range, and lastly, a final-stage filter of the impregnated paper type to remove the finest particles. The yarn-type filter is sometimes eliminated where a clean fuel supply can be assured. In engines with unit injectors, the first-stage filters are of the paper type, and final-stage filters of a metallic type are installed in the body of each injector, one a t the inlet and one a t the outlet connections. Filters may be of simplex or duplex construction, with the latter being used when it is
MEDIUM AND HIGH-SPEED DIESEL ENGINES
not possible to shut down the engine to change filters. Again, it must be emphasized that the major objectives of the fuel system are to deliver clean fuel, free of air and water, to the injection pumps. To this end, filter cases should be installed in locations where they can be easily serviced and the elements can be replaced without introducing dirt and with a minimum of maintenance effort. Jobs that are difficult to accomplish tend to be accomplished less frequently. Steps should be taken to ensure that there is a minimum possibility of air entering the system through joints in the piping on the suction side of pumps. Adequate and easily accessible drain connections should be provided a t the lowest part of the fuel service tanks for stripping water or foreign matter which may accumulate. Systems suitable for handling heavy distillate or residual fuels are described in Chapter VIII. 4.3 Types of Lubricating Oils Used. The engine manufacturer furnishes information regarding the design as well as installation of the lubricating oil system. The manufacturer will furnish all necessary accessories and components and recommend the kind of lubricating oil to be used. Nevertheless, marine engineers should be knowledgeable of lubricating oil systems so that preliminary designs can be prepared prior to the selection of a particular engine, and to alert the engine manufacturer to unusual conditions in specific applications to insure that optimum solutions are obtained when compromises are necessary. Lubricating oils are classified into two broad categories; first by viscosity and second by the severity of the operating conditions which they can tolerate. The most common viscosity designation is by SAE numbers as shown in Table 4. Table 4 SAE Viscosity no.
5W 1OW 20W 20
30
40 60
Viscosity Values of Crankcase Oils
,
Viscosity Range Saybolt Seconds Universal at 210 F at 0 F min max min max 6,000 6,000a less than
12,000b
12,000 48,000
45 58 70 85
less than 58 70 85
110
a Minimum viscosity at 0 F may be waived provided the viscosity at 210 F is not below ... 40 - - STTS. - - -. Minim& &scosity at 0 F may be waived provided the viscosity at 210 F is not below 45 SUS.
Medium and high-speed diesel engines normally use SAE 30 or 40 lubricating oils. For small boat applications where engines are stored outdoors in cold weather, it will be necessary to use winter grades such as 5W or 10W,oils. I n addition to the viscosity, oils are classified by'the viscosity index (VI), which is representative of the slope of the viscosity-temperature curve for each oil. A high VI oil is one in which its viscosity varies little with
265
the temperature, whereas in lower VI oils the viscosity variation with temperature is greater. For engines operating in heated engine rooms, the VI is of lesser importance than in the case of exposed engines which must operate in winter a t low temperatures and, in addition, are subjected to varying loads and infrequent starts. The lubricant in an engine serves to cool rubbing surfaces and provides a hydrodynamic film to prevent metallic contacts. In addition, it carries away products of combustion from combustion chambers and removes metallic and abrasive products. In order to insure satisfactory performance in a variety of engine designs under widely diierent operating conditions, natural petroleum products are specially compounded with oxidation and corrosion inhibitors, antifoaming agents, detergents, dispersants and other additives to produce the desired lubricating oil properties. Oils are qualified by running laboratory tests, both in and out of operating engines. A good brief discussion of these tests can be found in SAE Information Report J304a [16]. Based on tests such as these, oils have been classified by the American Petroleum Institute as to their suitability for use in engines under operating conditions of differing severity. For gasoline engine use, oils are classified in order of their ability to cope with increasingly severe operations as ML, MM, and MS and for diesels as DG, DM, and DS. I n addition, there are numerous military specifications and commonly used descriptors which cover the same basic oil properties [23]. In general, the severity of engine operating conditions and the design of the particular engine will determine the proper lubricating oil to be used. Sustained operation a t high load is not the only condition which may be called "severe." In fact, other conditions such as high sulphur or carbon content of the fuel, widely fluctuating loads or ambient conditions, frequent starts and stops, or atmospheric contamination may impose more severe oil requirements than high loads alone. Approximate military specification equivalents to commercial lubricating oils DG, DM, and DS oils are MILL-21044, MIL-G2104B, and MIL-L-45199 respectively. MIL-L-9000 is a Navy specification oil with increased resistance to the deteriorating effects of water contamination. It is below MIL-L-45199 in detergency level. The best judge of the proper oil to be used in an engine is the engine itself. Where past experience with a particular engine or with special operating conditions is unavailable, the judgment of the engine manufacturer and oil supplier must be relied upon. 4.4 Lubricating Oil System Design. The components of the lubricating oil system are usually furnished by the engine manufacturer and, in many cases, are completely assembled to the engine for installation in the ship as a unit. An oil sump is usually located under the engine and a positive-displacement pump takes suction from the sump and &scharges t h e oil to the engine through a flter, cooler, and strainer, in that order. The
I
MARINE ENGINEERING
MEDIUM AND HIGH-SPEED DIESEL ENGINES
1
I
EXTERNPLL RELIEF
METAL EDGE STRAINER
Dlesel engine lubricating oil consumption will vary depending on engine speed, size, and design details. Typical values of oil consumption are:
DUPLEX PRESSURE
II
COOLER BY-PASS
Medium-speed engines. . .3000-6000 bhp-hr/gal High-speed engines. . . . . .2000-3000 bhp-hr/gal
I - - - - - LUBE OIL - - - - - --,COOLER
\PRESSURE PUMP
,
ENGINE JACKET WATER (A) LUBRICATING OIL SYSTEM W I T H A FULL-FLOW F I L T E R AND WET SUMP
SCAVENOING
I
PRESSURE PUMP TO FILTER, ETC.
( 0 ) LUBRICATING OIL SYSTEM WlTH A DRY SUMP
M
I ------,
L---
JACKET WATER
FINE FULL FLOW
-
STRAINER LUBE OIL COOLER
---J
--
(C) LUBRICATING OIL SYSTEM WlTH A BY-PASS FILTER
-(D)
TO COOLER AS IN(A) OR(C)
ALTERNATIVE SYSTEM WlTH A SHUNT FILTER Fig. 15
Diagram of various lubricating oil systems
pump is equipped with a pressure relief, or in some instances, a pressure regulating valve. This system is shown diagrarnaticdly in Fig. 15(a). The sump tank should be sdiiciently large so that the oil does not splash up to the level of the crankshaft seals and so that the connecting rods will not dip into the oil under pitching an4 rolling conditions. I n addition, the sump should contain a quantity of oil in gallons equal
to about twice the rated capacity of the presswe oil pump in gallons per minute. If space is not available, the sump capacity may be less but not below a one-half minute pump supply. However, under these circumstances, oil change periods will be shortened appreciably. A much preferred solution to the problem of lack of space under the engine is to use a dry sump installation as shown in Fig. 15(b).
Oil change periods will vary with the severity of engine operation, quality of the lubricating oil, and size of the sump tank. With a dry sump, it is necessary to provide an additional pump to move the oil from the oil pan to the sump tank. This scavenging pump should have a capacity a t least 25 percent greater than the pressure pump to insure that the dry sump will, in fact, be dry. The oil flow requirements of engines will vary considerably, depending on such things as the use of oilcooling for pistons, whether the engine is naturally aspirated, supercharged, or after-cooled, and whether it is a two- or four-stroke cycle. The oil pressure pump capacity can be estimated a t about 0.2 gpm per horsepower for preliminary sizing of the system, though it might be half as much for some engines. Pump discharge pressures up to 100 psi can be expected in some engines. Since marine engines may run a t low speed for pr+ longed periods, engive-driven lubricating oil pumps should have adequate capacity to provide pressure under these conditions. Normal practice is to provide fullspeed pressure a t one-third speed. Many engines designed for constant-speed generator drive are found to be inadequate in this regard. ABS rules [24] require that the lubricating oil piping be entirely separate from other piping systems. For other than automotive-type engines, it is good practice to include a motor-driven lubricating oil pump in the system to be used to prime the engine before starting. The motor-driven pump is sometimes installed so that it can circulate oil from the sump tank through a heater and filter and then back to the sump in order to purify the oil while the engine is not runtiing. If this is done, care must be taken to insure that the normal oil supply to the engine can never be blocked off by negligence in realigning the valvivg prior to an engine start. It is possible to overprime opposed-piston engines, and the manufacturer's recommendations regarding means to prevent damage from this cause should be followed. Normally, the ABS rules require that an independently driven lubricating oil pump be furnished. However, for vessels in river or harbor service or vessels below 300 tons, this requirement is waived. I n those applications where the size and design of the engine is such that lubrication before starting is not necessary and an attached pump is normally used, an independently driven spare pump is not required if a complete duplicate of the attached pump is carried as a spare. Lubricating oil must be kept clean and free of abrasives. The best way to control abrasives is to prevent their entrance into the lubricating oil system. The designer should insure that filler caps are provided and located so
267
that foreign matter cannot get into the system n-hen it is being filled. Provision must be made to prevent dead pockets where deposits can accumulate and subsequelltly break loose in large quantities and cause damaging wear. Clean-out openings must be provided a t all locations where sludge is likely to accumulate. The diesel engine lubricating oil must be kept free of abrasive and corrosive q-mterials if it is to function properly. Additives are used to control corrosion, and filters are used to control abrasives. There are three commonly used filtering arrangements: (a) full flow, (b) by-pass, and (c) shunt. Full-flow filtration has become predominant in recent years, and, as its name implies, all of the oil supplied to the engine goes through the filter. This arrangement is shown in Fig. 15(a). Inasmuch as all of the oil going to the engine passes through the filter, it is necessary to prevent oil starvation of the engine in cases of filter plugging. An external by-pass line around the filter, together with a pressure relief valve, provides this protection. The duplex pressure gage shows the inlet and outlet pressures and gives advance warning of impending filter clogging. Normally, this takes place slowly so as to enable filter element changes to be scheduled during nonoperating periods. With the arrangement shown, the pressure relief valve setting can also be checked by means of the duplex gage. The lubricating oil cooler is installed after the filter because it is more effectiveto filter hot oil, as the pressure drop through the filter is less and filteeng is more complete. The simplex metal edge strainer is installed as close to the engine oil manifold inlet as possible to prevent the entrance of foreign matter into the engine. A by-pass filtering system is arranged as shown in Fig. 15(c). In this case, the oil discharged from the pressure pump is divided into two streams; one goes to the oil cooler and thence to the engine, and the other goes through a flow controlling orifice to the filter and thence to the sump. The quantity of oil by-passed through the filter to the sump must be in excess of engine lubricating requirements. The full pump discharge pressure is available for the pressure drop across the filter and orifice. By-pass filtration flow is approximately 5 percent of the pump A shunt capacity. filtering system is shown in Fig. 15(d). In it, the full flow to the engine is made up of oil which flows through the shunt filter and oil which flows in a by-pass around the filter, the quantity of by-passed oil being controlled by an ~rifice. There are three types of filter elements: those made of fine-mesh wire screen or metal edge (such as stacked disks); absorption types which are made of wool or cotton yarn, cellulose, or impregnated paper; and adsorbent types which, by adhesion, hold molecular layers of the contaminants to the filter element. The adsorbent elements contain fullers or diatomaceous earth, chemically treated papers, charcoal, or active clay. These filters are capable of removing additives from oil and should not be used with detergent lubricating oils except as part of an oil reclaiming system which is run separately
268
MARINE ENGINEERING
from the engine oil system. Additives should be restored to the oil after reclaiming and prior to reuse.' To provide an indication of the size of full-flow oil filters, the dimensions and flow rates of elements covered by specification MILF-20707 are given in Table 5. Table 5
Class 1
Characteristicsof Full-flow Oil Filters Max Dia (in.) 3
Maximum Instatled Length (in.) 4
A prox. Flow h t e (gpm) 2
Filter elements may be contained in individual containers, though it is more common to install several elements in one filter case. The elements may be stacked two or more high and arranged in any desired pattern in order to shape the case to suit available space; however, a cylindrical case is most common. It is essential that relief valves not be installed at the bottom of the filter case, where foreignmatter accumulates only to be washed into the engine whenever the relief valve lifts. Lubricating oil coalers are generally of the shell and tube type. For compact units, tubes may be fabricated in other than cylindrical form and include extended surfaces to increase heat-transfer rates. It is recommended that the pressure drop on the oil side not exceed 10 psi a t operating temperatures and that on the waterside be limited to 5 psi. The lubricating oil should be cooled with fresh water, even though it results in a larger cooler. The benefits in terms of faster oil warmup, reduced waterside fpuling, and better temperature control will more than offset this size increase. 4.5 Cooling Systems. As is true in all heat engines, the diesel engine must reject heat to the environment. Quantitatively, this heat is equal to the difference between the heat released by the injected fuel and the work output. The rejected heat is in the form of heat in the exhaust gas, heat transferred to the cooling system and lubricating oil, and the loss to the atmosphere due to radiation and convection from the engine exterior surface. It was previously stated that diesel engine efficiency is now being calculated and reported on the basis of the lower heating value (LHV) of the fuel in order to be consistent with presentations for other heat engines. However, much of the heat balance data in the literature, when reported on a percentage basis, will be found to be based on the higher heating value (HHV) of the fuel. For many years, the standard rule of thumb for estimating diesel heat losses has been, "One third of the heat in the fuel is converted to work, one third is lost in the exhaust gases and radiation, and one third to the cooling system. " The modern medium and high-speed, highoutput engines are more efficient than older engines and rather than one third of the input heat being converted to work, it can be expected to range'between 35 and 38 percent HHV (38 to 41 percent LHV), while about 28
percent is rejected to the cooling water and lubricating oil. While these percentage figures are of historical and general interest, figures in terms of Btu per horsepower per minute gre more useful in design work for estimating sizes of coolers, ventilation heat loads, and piping sizes. Average values for these heat losses are: To cylinder jackets. ....... .20-30 To oil coolers. . . . . . . . . . . . . . 5-20 To air coolers. ............. 5-10 To exhaust ................ 25-40
Btu/hp-min Btu/hp-min Btu/hp-min Btu/hp-min
These values will vary with engine design, load, speed, temperature of the coolant and oil, and degree of supercharging and aftercooling. For preliminary design purposes, the higher values may be used and about ten percent should be added when sizing coolers. After a specific engine is selected, exact values will be furnished by the engine manufacturer. In order to properly size the cooling system for an engine, the manufacturer must be provided with information relative t o the expected ambient conditions under which the engine will operate. For naval ships which must operate in widely varying locations, as an example, it is specified that coolers should be sized on the basis of an 85 F seawater temperatuye. If it is known that the ship being designed will operate in colder water, the cooler size can be reduced; or, on the other hand, if due to peculiar conditions ambient cooling water temperatures are exceptionally high, larger coolers will be needed. The discharge temperature of the seawater from the coolers should be kept well under 130 F to prevent scaling of the surfaces. Engine manufacturers design their equipment so that the water temperature rise of the fresh water across the engine will be between 10 and 20 deg F. This is done to minimize thermal stress and distortion in the engine. The capacity of freshwater pumps is usually in the range of 0.3 to 0.5 gpm/hp. The capacity of seawater pumps should be the same in order to simplify manufacturing and repair parts stocking, provide a margin to accommodate additional equipment such as aftercoolers on turbocharged engines, and prolong seawater cooler cleaning intervals by minimizing the seawater discharge temperature. A typical cooling water system for a medium-speed marine diesel is shown in Fig. 16. Automotive-type marine diesels usually are supplied with all piping, coolers, thermostatic valves, and expansion tanks assem-' bled to the engine. In this case, the only water connection the shipbuilder is required to make is from the sea to the seawater pump suction. The seawater pumps are likely to be subjected to reduced pressure a t the inlet, so to prevent loss of suction it is recommended that pump seals be of a type which will prevent air leaking into the pump under a suction head of 15 f t of water. The expansion tank should be located a t the highest point in the system and all pockets should be vented to the expansion tank. Water piping should be shed to match the pump suction and discharge flanges, or at least
TO WASTE HEAT RECOVERY SYSTEM
CI
G L t N G
THERMOSTATIC TEMPERATURE REGULATING VALVE WlTH BUILD-IN MANUAL CONTROL
cow. 1. D . OF RETURN TO BE APPROX. 3 x I.D. O f VENT LINE
PILLARY TUBE
CTUATING BULB ENT LINES CONNECTED TO IGHEST POINTS OF ENGINE
TO OVERBOARD OR TO WET TYPE MUFFLERS
THROTTLING VALVE TO BE APPROX. 10 PIPE DIAMETERS
JACKET COOLER TO L.O. SYSTEM VP
NOTES
I TO BE INSTALLED ON ENGINE GAGE BOARD. 2 DRAINS TO BE INSTALLED IN LOWEST POINT IN JACKET WATER AND SEA WATER SYSTEMS. USE GATE VALVES. 3 EXPANSION TANK SHALL BE LOCATED IN THE SAME COMPARTMENT WlTH ENGINE. 4 SEA WATER SUPPLY FOR GENERATOR AIR COOLERS WlTH THROTTLING VALVE TO BE PROVIDED ONLY WHEN REQUIRED. 5 SEA WATER PUMP SUCTION PIPING TO BE OF SUFFICIENT SIZE, AND ARRANGED TO LIMIT VACUUM AT PUMP SUCTION TO 6' HG AT RATED RPM. 6 JACKET WATER BY-PASS ACROSS BOTH COOLERS SHALL BE PROVIDED WHEN REQUIRED TO OBTAIN SPECIFIED OPERATING TEMPERATURES 7 TANK FOR INITIATING AND MAINTAINING JACKET WATER TREATMENT. CAPACITY TO BE . -11/2OALLONS FOR EACH 100 GALLONS IN ENGINE SYSTEM. TO USE.CLOSE V A L-V .. E ~ ~ ~ ............ TO TANK AND OPEN VALVES ' C m 8 ' D ' T 0 DRAIN TANK. CLOSEm~..FlLL TANK WlTH CORRECTAMOUNT OF SOLUTION. CLOSE -C:OPEN~A:CIRCULATION OF JACKET WATER WILL FEED SOLUTION INTO SYSTEM. TEST SAMPLE FOR CORRECT CONCENTRATION. ~
kg.16
Diagram of typical cooling water system
to provide smooth transitions if the piping must be smallei-. It may be possible or desirable to replace the seawaterto-freshwater heat exchanger with a hull cooler in cases where the seawater is contaminated or weed-infested. The hull cooler may consist of pipes with gxtended heattransfer surfaces 'hlounted outside'the hull, or simply tanka inside the hull wherein the heat is transferred directly to the sea through the hull plating. Kort nozzle shells have been used in the same manner. Thermostatic valves should be used to automatically regulate the outlet temperature of the jacket water. The outlet temperature should be kept in the range of 160 to '185 F to minimize the size of coolers and to prevent corrosive cylinder wear [25,26,27]. It is recognized that operating personnel prefer to operate cooler engines, as less time is required for cooling down if repairs are
necessary, surface temperatures are not uncomfortable to the touch, and machinery spaces are cooler. It is important, therefore, that the system be designed in such a t a y that the desired operating temperatures cannot be altered easily by the operating crew. Thermostatic controls should be such that adjustment out bf the proper range is impossible and orifice plates should be installed in piping systems once the proper balance is established. The jacket water of diesel engines must be treated to prevent corrosion and to minimize the effectsof cavitation on cylinder liners and jackets. A number of cooling water treatments, including alkaline chromates, soluble oil, sodium boron nitrate, and sodium nitrate-nitrite, are used. The engine manufacturer should recommend the coolant best suited for his engine. It may be desirable, however, for large fleet operators to standardize the
MEDIUM AND HIGH-!SPEED DIESEL ENGINES
coolant treatment used in their fleets, in which case the engine manufacturer should be informed of the preferred treatment. No water treatmeht will last indefinitely; the water must be tested regularly and chemicals occasionaily added. To insure that this is done, provision should be made for drawing of samples from convenient locations and to provide easy access to chemical addition points. A filling funnel located against the overhead in a hot engine room is almost certain to result in neglect of water treatment. For boat engines or other engines which may be exposed to freezingtemperatures, conventional inhibited ethylene glycol antifreeze solutions should be used. Where engines are installed high above the waterline, or where a source of raw water is either not available or unsuitable, air-cooled engines should be considered. It is important to insure that the cooling air is a t a sufficiently low temperature and that the air supply is not restricted by inadequate grill or duct openings to the *weather. The wind direction and velocity should be investigated to ensure that they will not oppose the cooling fan and impede airflow. Direct air-cooled engines are somewhat noisier than the liquid-cooled engines itlasmuch as they have no water jackets around the cylinders to attenuate vibration and noise. This fact should be considered when locating the engine in the ship. Direct air-cooled engines are delivered complete with cooling fan and the necessary cowling. The ship designer must insure that the air gets to the cooling fan and that the hot air from the engine is discharged from the compartment and is not allowed to recirculate back to the fan suction. About 50 cfm/hp of free air is required for air cooling. Air cooling can also be applied to liquid-cooled engines by the use of radiators to transfer the heat from the jacket water to the air. There is somewhat greater flexibility in installing an indirectly air-cooled engine than there is with one cooled directly with air. It is possible to place the radiator remotely from the engine to optimize installation arrangements. The radiators may be installed horizontally or vertically. Care must be taken to insurge that the engine-attached water pump characteristics match the cooling system requirements and provide an adequate flow of water. As with the directly cooled engine, particular attention must be paid to avoiding restrictions in the airflow path and to prevent recirculation of the cooling air. Thermostatic control can be applied either to the waterside, in which case the thermostatic valve directs the water flow through or around the radiator core, or to the air side, in which case the thermostat may operate a valve to divert air around the core, vary the speed of the fan drive, or change the pitch of the fan blades. The last two are more efficient as fan power is minimized at light load or when the air is cold. 4.6 Waste-Heat Utilization System. It is possible to utilize the waste heat from an engine by schemes which range from the simplest of using radiated heat to keep the engine room warm to complex schemes for generating
steam and power from the steam. The two most common uses for waste heat are: (a) heating water which can be used to heat spaces, heat fuel, cargo, or to distill fresh water; and (b) generating steam for use in absorption refrigerntion plants, space heaters, distillers, heat exchangers and low-pressure steam turbines. Almost 100 percent of the heat rejected to the jacket water and lubricating oil and about 60 percent of the exhaust heat are economically recoverable 1281. The amount of heat recoverable depends on the system used and the extent of the recovery equipment employed. The quantity of heat available depends on the design of the particular engine and operating conditions. Average values for heat losses were previously listed; these values can vary considerably, even for the same engine design. For example, in the case of a Fairbanks Morse 38D 8 diesel [29], the heat rejection rate to oil and water has been found to vary: (a) From a minimum of 35 Btu/hp-min a t 720 rpm to 41 Btu/hp-min at 900 rpm. (b) From 36 Btu/hp-min with an oil outlet temperature of 170 and water outlet temperature of 165 to and water at 230 32 Btu/hpmin with the oil at 185 (c) From 36 Btu/hp-min a t full load to 159 Btu/hp-min at 25 percent load. (d) From 36 Btu/hp;min a t full load without turbocharging to 22 Btu/hp-min with turbocharging. The question of whether to use waste heat and how extensive a waste-heat recovery system to design is largely one of economics. A detailed study must be made to develop load-time cycle data so as to determine how much heat is available. At the same time, the demand for waste heat must be analyzed as well, to make sure that there is sufficient heat available to meet the demand a t the time it is needed and that the heat generated can be used. This section briefly covers the basic systems used to recover waste heat, giving the basic engine input data required by the designer to size the equipment to utilize the heat and to devise special arrangements to suit each ship design. The major heat recovery systems are: (a) Engine radiator to air. The air temperature leaving the radiator is between 100 and 150 P and can be used for preheating boiler combustion air or space heating. (b) Normal-temperature, hot-water systems. These use a normally closed system with a thermostat to control the water outlet temperature and a heat exchanger to transfer unused heat to the seawater coolant loop. Hot water to the waste-heat utilization loads would be taken from the system at point F in Fig. 16 and returned a t point E. Additional heat can be added to the water by an exhaust gas exchanger. I n this system, the maximum jacket water temperature is that which can be obtained without pressurizing the expansion tank and will range from 180 to 220 F. A variation of this system is to use a secondary circuit to transfer the waste heat to the utiliring equipment. A heat exchanger is used to transfer the jacket water heat to the secondary circuit, and the
1
f
1
1 1
I
exhaust gaa heat recovery unit, if used, is installed in the secondary circuit. (c) High-temperature; hot-water systems. This system uses jacket water engine outlet temperatures in the range from 220 to 250 P and functions essentially the same as the normal-temperature, hot-water system except that a higher pressure is required in the circulating systems, especially in the engine coolant circuit. I n this system, a pressure control must be provided in the engine coolant circuit which will assure a pressure a t all points in the system sufficiently high to prevent the formation of steam. The source of this pressure may be a static head imposed by an elevated expansion tank or controlled air pressure in the expansion tank. For 250 F water, a pressure of about 20 psig is required a t the engine. In this system, all circulating pumps must be suitable for the higher pressures and temperatures. Engine and piping system gaskets and seals must also be suitable for the imposed conditions. With this high-temperature cooling system, it will not be possible to cool the lubricating oil with jacket water. The heat from the oil cooler must be disposed of in a separate system if it is not possible to use it for preheat in some part of the wasteheat utilization circuit. It may mean that more heat can be abstracted from a normal-temperature system using the heat from the oil rather than from a hightemperature system which does not use this heat. Thermostatic controls must be provided to prevent exceeding the maximum permissible temperature and pressure controls to prevent boiling. Exhaust gas heat may- be recovered in the high-temperature system as well as in the normal-temperature system. (d) Hot-water and steam sgstem with a $ash boiler. This system is quite similar to the high-temperature, hot-water system with the expansion tank replaced by a flash boiler. The pressure in the boiler is lower than that in the hot-water system expansion tank so that the hot water can flash into steam. This type of system is usually designed to operate with a steam pressure of from 2 to 8 psig. The operating pressure is dependent upon the maximum design engine coolant temperature and is set so that the total pressure a t the engine outlet due to the steam pressure and the static head will prevent boiling in the engine jackets. As in the high-temperature, hot-water system operating a t 250 F outlet temperature, a pressure of 20 psig a t the engine is required. If the jacket water leaves the engine a t 250 F and 24 psig and the static head is reduced to 8 psig, an equilibrium condition will be established in the flash boiler with about 0.985 lb of water a t about 235 F being returned.to the engine and about 0.015 lb of saturated steam going to the waste-heat utilization system for each pound of water entering the flash boiler. The 0.015 lb of condensate returned from the waste-heat system is mixed with the water in the flash boiler prior to recirculating through the engine. Using a water pump capacity of 0.3 gpm/hp, the 0.015 lb of steam per pound of circulating water equates to about 2.25 lb of steam per hour per horsepower. I n this system, it is necessary
271
to provide a water level control in the flash boiler and to supply make-up from the condensate return system. Boiler pressure control must be provided to prevent the pressure falling to the point where boiling will occur in the engine jackets. All piping from the engine to the boiler must pitch upward. (e) Ebullient system. An ebullient system may appear attractive where steam is required a t pressures of 12 to 15 psig for use in absorption refrigeration or airconditioning systems or other applications [30, 311. I n the ebullient system, boiling occurs in the engine water jackets. The engine circulating water pump is removed and the flow is maintained due to the diqerence in density of the steam-water mixture a t the outlet and the solid water at the inlet to the engine. System performance is sensitive to restrictions in the cooling water system and to the slope of the cooling water line. Pitch and roll can disturb the flow of cooling water t o the engine. A temperature difference across the engine of about 2 to 3 deg F will be maintained. The steam-water mixture from the engine flows to a steam separator above the engines. The steam pressure must be regulated a t the separator to insure that the pressure does not become too low, causing excessive boiling in the engine jackets, or too high, resulting in an excessive outlet temperature from the engine water jackets. Exhaust gas boilers can be provided with the ebullient system either built into the steam separator or operating in parallel with it. With an engine outlet temperature of 250 F, steam is generated at the saturation pressure of about 15 psig rather than a t 8 psig as in the previous example using a flash boiler. An estimate of the steam production capability of the ebullient system is given in Table 6 €301. Based on the data given in reference [4], the steam production capability of exhaust gas boilers is approximately as given in Table 7. Table 6 Steam Production Capability of an Ebullient System
Type of Diesel Engine
Fuel Heat ~ ~ Btu/hphr
%cycle turbocharged non-
8200
4gde I naturally aspirated &cycle turbocharged
Lb Steam/bhp-hr at Water Jackets with RBted Load ~ Exhaust ~ Manifolds t , ~ ~ Air-cooled Waterc Recovery cooled Unit 1.65
1.95
1.10
8500
1.90
2.35
1.25
7300
1.10
1.35
1.20
N ~ EThe : above data are based on 0-psig steam and 100 F ambient. 85y0 magnesia System or equal. piping is considered to be insulated with 1 in.-
Table 7 Steam Production Capability of Exhaust Gas Boilers
nT" "'
Diese Engine %cycleengine 4cycle engine
Steam Production Caqabilities, Ib/hg-hr 5 P@ 10 pslg 15 ps~g 0.75 0.78
0.70 0.75
0.68 0.74
h
MARINE ENGINEERING
In all waste heat utilization schemes, provisions must to duct the engine air from the outside directly to each be made to cool the engine when the waste heat cannot be engine and provide a three-way valve to permit the utilized. Where steam is generated, it is necessary to engine to take air from the engine room or the weather. Each engine should be provided with its own exhaust provide condensers and feed pumps and to insure proper system. If space does not permit such an arrangement treatment of the make-up water. Fortunately, the water treatments required for boilers and engine water jackets and it is necessary to combine the exhaust ducts from are compatible [4]. This treatment would include a several engines, it is necessary that valves be provided water softener to give zero hardness and a pH value in the branch from each engine to prevent backflow into between 9 and 11. Exhaust gas boilers may be combined an idle engine. The size of intake and exhaust ducts may be estimated with mufaers and may also be provided with supplementary oil firing to insure a steam supply under all using a figure-d 3.5 cfm/hp for'the intake air and 8.5 engine load conditions. The engine exhaust tempera- cfrn/hp for exhaust gas. $hese values may be high for ture conditions must be acceptable to the boiler supplier. most naturally aspirated engines and some turbocharged 4.7 Intake and Exhaust Systems. The intake and engines; however, it is desirable to provide some margin. exhaust system consists of the piping, filters, and silencers Duct velocities of 100 fps for the inlet and 150 fps for the necessary to conduct the outside air to the engine and to exhaust are suggested for preliminary design purposes. lead the exhaust gas from the engine to the atmosphere. When a particular engine has been selected and its To perform effectively, the depression in inlet air actual air and exhaust requirements are known, duct pressure and the elevation of the pressure a t the exhaust sizes can be calculated to meet the allowable pressure outlet must be minimized. The correction factors of drops. If it is necessary to reduce the duct sizes, higher Table 1 indicate the effect on engine power output as the gas velocities may be used [32]. Contrary to what may be believed, a marine engine is pressure st the engine inlet is reduced and as the temperature a t the inlet is raised. Turbocharged engines not always 'provided with clean air, particularly in river are particularly sensitive to intake air pressure and and harbor operations and sometimes in offshore exhaust back pressure. It is recommended that the operations in the vicinity of a desert. In these cases, it exhaust back pressure a t the engine outlet be limited to is necessary to provide air filters or cleaners to remove about 12 in. of water for turbocharged engines and twice abrasive or oily particles from the air. There are that for other engines. The inlet pressure drop in the basically three types of air filters or cleaners: ducting should not exceed 6 in. of water. Excessive 1 Dry inertial. The air direction is changed in the pressure drops in the intake or exhaust systems or a high filter, causing the heavier foreign particles to be separated inlet temperature can cause a loss of power, poor fuel from the air stream. These filters may be of the cyclone economy, high temperatures of engine parts, jacket or impingement type. water overheating, and excessive engine deposits. 2 Dry paper. The air is passed through porous The inlet to the induction air system should be located treated paper which retains foreign matter. so that it is not possible to draw in engine exhaust gases, 3 Oil bath. The air stream is directed a t the surface hot air from ventilation system exhausts, spray from of lubricating oil in the sump of the cleaner. The air seawater, or flammable vapors from tank vents or other reverses direction at the oil surface, and picks up and sources. Flammable vapors are particularly dangerous carries "washing" oil to the filter media. Foreign matter as they can cause an engine to overspeed, and the normal is captured at the media and washed to the oil sump overspeed trip and fuel governor will be unable to shut where it can be drained. it down. Actual filters usually employ combinations of these Piping should be properly supported and provided with expansion joints to avoid strains on the engine manifolds three methods. I n addition, self-cleaning designs are or turbocharger flanges. The velocity through the cor- available. It is possible to obtain filters of reasonable rugated metallic hose type of expansion joints should be si3e with moderate pressure losses from a variety of specified to insure that the type furnished will be suitable. -sources [2, 33, 341. The installation should provide Condensate traps and drains a t the low points of the gages to measure the pressure drop across the filter to engine manifolds should be provided. Provisions should give warning of impending clogging. The filter must be be made for rain covers to prevent the entry of water into installed in a location where is can be removed easily for cleaning. This would appear to be obvious, but for idle engines. Engine air may be drawn from the engine room or some reason it is frequently overlooked when the details ducted directly from the atmosphere. It is simpler from of an installation are developed. Air intake silencers are necessary to prevent blower an installation standpoint to take the engine air directly from the engine room; however, this arrangement has the noise from creating uncomfortable conditions in the disadvantage that the space may be excessively cooled engine room or spaces adjacent to the air inlet ducts. in winter. I n addition, in summer or in hot climates, Positive-displacement blowers generate a low-frequency the air may be heated by other equipment in the space pulsation, whereas the noise from turboblowers is very and reach the engine inlet a t an even more elevated high in pitch and is more likely to be objectionable. temperature, resulting in a loss in power. It is preferable Engines are normally fu+shed complete with an air
273
MEDIUM AND HIGH-SPEED DIESEL ENGINES
I
I
$1
silencer for use when the engine draws air directly from the machinery space. If the air is to be supplied to the engine via a duct, this should be so specified in order to insure that the intake silencer will be suitable. Exhaust mufaers are provided to reduce the pulsations in exhaust line pressure due to the cyclic release of slugs of exhaust gas into the engine manifold as the exhaust valves in each cylinder open in turn. The m d e r also serves to reduce atmospheric noise a t the outlet of the exhaust system. M d e r s may be of the wet or drv type. wet mufflers are infrequently used except & small boats, as they are limited to horizontal installations where the exhaust is through the hull of the ship above the waterline. Seawater is injected into the m d e r and cools the exhaust gas as the water is vaporized. The steam exhaust gas mixture is discharged overboard. With a wet m d e r , care must be taken to insure that the exhaust does not blow across the deck or against the side of adjacent ships. They should be fabricated of AISI 316L stainless steel for a reasonable life expectancy. Dry-type mufflers may be installed horizontally or vertically in the engine room or in the exhaust stack. These mufflers should be provided with spark-arresting features to prevent hot carbon particles from impinging on topside surfaces. I n general, mufflers should be capable of reducing the overall noise of exhaust gases to a maximum of 92 db a t a radius of 10 f t from the end and 2 ft above the muffler tailpipe with the engine operating a t rated load and speed. The noise level permitted may be more or less than this, depending on the particular installation. Figure 17, which was talcen from reference [35], gives an indication of the weight and size of naval dry mufflers with spark arresters. The muffler inlet flange size is the same as the exhaust pipe size. The pressure drop through these mufflersshould not exceed 18 in. of water a t engine rated speed for nonturbocharged engines and 6 in. of water for turbochafked engines. The pressure undulations in the exhaust from a turbocharged diesel are considerably reduced in flowing through the turbine to the extent that a much smaller muffler is generally permitted. However, spark-arresting features are still required. The muffler is generally installed in the exhaust pipe about one third of the distance from the engine to the end of the pipe. This distance will vary with the type of engine, the type of muffler, and the piping arrangement. The precise location of the m d e r should not be fixed until the engine selection has been made. ' 4.8 Starting Systems. To start an engine, it is necessary to rotate it, such that its speed and, consequently, its compression temperature are sufficiently high to insure ignition of the fuel when it is injected into the engine cylinder. The starting system is the means of supplying the energy for rotating the engine. The starting system can be operated with air, electricity, or hydraulic fluid. Air can be applied directly into the engine cylinders or used to drive an air motor geared to the engine crankshaft through an overrunning clutch.
----
L
<*
3
0
T Y P E A - DRY W I T H SPARK ARRESTOR
NOTE: MUFFLERS TO COMMERCIAL DESIGN MAY BE SOMEWHAT SMALLER AND LIGHTER
Fig. 17 Navy muffler dzer
Electric starting can be applied to a small geared motor arranged in a similar manner to the air motor or in the form of special windings in the driven power generator. Hydraulic starting is by a geared motor only. Engines and starting motors are usually designed so that the customer is free to select the medium best suited to a particular ship. The first decision which must be made with reference to the starting system is whether the engine is to be direct-reversing or unidirectional. If a unidirectional engine is to be used, starting can be by means of direct cylinder injection, starting motors, or generator windings. If a direct-reversing engine is selected, there is no problem or choice remaining, for the starting system must be air with direct cylinder injection. Examples of other considerations are: ( a ) a salvage ship may have requirements fgr large quantities of compressed air, which would then be available for starting and would rule out other means; (b) a ship may have no air requirements and it could be more economical to start electrically; (c) hydraulic systems have been used in Navy applications where their nonmagnetic materials and freedom from stray electric currents are the attractive features. A hydraulic starting system has the capability of being charged with a hand pump to make an initial start, after which the hydraulic accumulator can be charged by an engine driven pump. In some instances, it is desirable to take advantage of this characteristic by using a hydraulically started engine to drive a compr&sor which then furnishes air to air starters on other engines
I
MEDIUM AND HIGH-SPEED DIESEL ENGINES
274
MARINE EN
in the plant. It may be seen that the selection of the starting system is tied very closely to the type of engine selected and the nature and duty cycle of the ship and the owner's preference. Some of the general characteristics and requirements of these starting systems may also have an impact on the type selected. The air pressure required with a direct cylinder injection system is usually about 250 psig with a minimum pressure of about 100 psig. The pressure required for an air motor varies from 75 to 150 psig. ABS requires that the starting air storage be in at least two containers with a capacity capable of starting directreversing engines 12 times and unidirectional engines 6 times corlsecutively without recharging. The air can be stored at pressures above that required by the engine and pressure-reducing valves used to lower the pressure to the proper value at the engine. I n this manner, the size of the air storage vessel can be reduced and the air requirements consolidated 'for the engine and other services aboard ship. ABS requires that the electric storage batteries provided be capable of 6 consecutive starts without recharging. Starting system voltage is usually 12, 24, or 36, with cells arranged in series and parallel to obtain the required voltage and storage capacity. Battery charging generators are usually engine driven for automotive-si~eengines and from a separate power source for larger engines. Battery charging requirements can be supplied by the ships' service system with the necessary conversion equipment. Hydraulic starting systems operate at pressures ranging from 1000 to 3000 psig, with the complete system usually supplied by the starter manufacturer. I n general, more than one starting motor can be installed on an engine so that it is possible to start many medium and high-speed engines by means of motors. There are some applications where engines are required to start in low ambient temperatures. Electric storage batteries lose capacity a t low temperatures and provision should be made either to heat the battery compartment, insure that there are sufficient cells to crank the engine even with the reduced capacity, or to use air or hydraulic starting systems. Water and oil heaters and ether starting aids may also be required at temperatures below 35 F. In any case, the minimum temperature at which engine starting is required should be made a part of the procurement specifications. So many options exist in the selection of starting equipment that it is good practice to develop a list of requirements and enlist the aid of the engine manufacturer in selecting final equipment and arrangements. 4.9 Controls. The control of a diesel engine is effected primarily by regulating the fuel injected; this is accomplished by means of a throttle lever which moves the racks of the fuel injection pumps. This control can be applied manually or through various types of governing devices sensitive to engine speed, load, discharge pressure, or flow rate of driven equipment. Rather than going into details of the equipment required, this section
will indicate the control functions which can be obtained and how they can be applied. There are a number of types of governors which may be considered for a particular application. The major alternatives are: (a) Manual. The operator moves a throttle lever which controls the engine speed remotely or directly. This is similar to the gas pedal on an automobile. (b) Limiting-speed governors. This type of governor controls the engine speed at idle and at rated speed. I n the range between idle and rated speed, control is manual. (c) Variable speed. The input lever acts to compress the speeder spring on the governor flyweights and controls the speed automatically to a fixed value for each position of the lever. (d) Isochronous. A governor which ,holds engine speed constant at a set value regardless of load variation. The desired constant speed is set by positioning an external lever or dial. (e) Speed-droop governor. This governor controls the speed at a given value under full-load conditions, but allows the speed to rise as load is decreased. The amount of droop is expressed in percent as 100 times the difference between no-load and full-load speed divided by full-load speed. The percent droop is adjustable in some governors from a stated value down to zero (isochronous); in others, it is adjustable between two values above zero droop. Droop adjustment is internal on some governors and external on others. (j)Load, torque, or BMEP-limiting governors are those with devices to limit the fuel injected as a function of speed to permit governing the speed of the engine with the load limit just above given values (such as a propeller curve). A variation of this type of governor limits fuel input as a function of engine air manifold pressure. This device is used primarily with turbocharged engines to prevent smoking during severe load transients. (g) Lodspeed governing. This governor controls the speed of the load to the desired value when the load speed and prime mover speed are not the same as when a torque converter or hydraulic coupling is used. (h) Electric governors. There are two basic types of electric governors: one which is used on generator sets receives its power supply and speed signal from the alternator; the other, used for propulsion, is powered by a separate a-c line. The speed signal from the second type comes from the electric impulses generated by a magnetic pickup in the vicinity of a rotating gear. This unit permits control over a speed range as wide as 20: 1. The electric unit for generating services senses and responds to the electric load and rate of change in electric load, as well as to frequency (speed). Significant reductions are made in the transient off-speeds as compared with those usually experienced with mechanical speed-sensing governors. The electric unit also senses and responds to the difference in load between paralleled units and permits each engine to assume equal percentages of load under isochronous conditions.
(23 Ovei-speed governors. These are of either the selfresetting or the trip type. The ovcrspeed governor is usually an independent governor whose sole function is to prevent engine overspeed in case the load is lost as when a propeller comes out of the water or a generator is dropped off the line. The self-resetting type will shut off fuel or sir to the engine until the engine speed drops to a preselected value; then it will permit control in the normal mode. Engine speed will cycle between the lower set point and the overspeed limit until the cause for the overspeed is removed or the engine is shut down. The trip type will shut down the engine if overspeed occurs ahd requires manual reset prior to restarting the engine. The overspeed governors may be connected to something other than the normal fuel linkage to insure control in the event of mechanical difliculty with the fuel system. It can actuate a valve in the fuel supply to the engine or in the air induction system. Governors are built which combine several of the features described in the foregoing in one unit; i.e., speed droop and load control. For a more comprehensive discussion of governors and definitions of governing terms, see references [4, 36, 371. The control of a propulsion engine can be remote, from the bridge or other location, or local at the engine. Remote control can be by means of either direct mechanical linkages from the remote-control lever to the fuel linkage on the engine or by means of hydraulic, pneumatic, or electric systems. The choice of control system depends on the number of engines to be controlled, the number of remote-control stations involved, the distance between the remote station and the engine, and the flexibility desired in the system. Governors for variable-speed propulsion engines are generally capable of controlling engines so that a t all speeds and loads up to rated, the periodic or aperiodic oscillations of speed are no more than f1.0 percent of operating speed. The maximum deviation from normal speed when full or partial load is removed or applied suddenly should not exceed 10 percent of operating speed. The speed should return to f 1percent of operating speed in 15 sec or less following the load change. For single engines driving a fixed-pitch propeller, the simplest and cheapest control, and a satisfactory one for small boats, is a manual system. For higher-powered ships, the limiting-speed or variable-speed governor should be used, together with a resetting type of overspeed governor. I n fact, an overspeed governor should be used with all engines except in small boat applications. The resetting type gives the pilot continued control of the engine even though the engine speed fluctuates, whereas the trip type can shut the engine down at a time when the ship's safety may be endangered. Multiple engines geared to drive a single propeller through solid couplings should be equipped with load limiting and reset overspeed governors. Multiple engines driving through sliptype couplings should use speed-droop governors with load-limiting features for protection again& overloading when leaa than the full
275
number of installed engines are used. A single eugiiie driving multiple loads should be protected against overload if the sum of the loads can exceed the erigiiie rating. Engines driving controllable and reversible-pitch (CRP) propellers can be controlled with a single lever to operate at a preset speed-load relationship by integrating the action of two governors, one controlling engine fuel and the other propeller pitch [38,39]. This type of control can also be used when two different types of propulsion plants are used, such as diesels and gas turbinea [40, 411. A two-governor type of control for CRP propellers is particularly useful when a shipoperates under varying load conditions such as towing and free route, or light and heavy displacement. If loadink is constant or varies slowly with time, a simpler singlelever control may be used whereby engine throttle position and propeller pitch are controlled together by means of cams whose relationship to each other is adjustable over a moderate range by a vernier. Generator sets which will not be operated in parallel can use isochronous governors if the frequency must be controlled accurately. For sets which must operate in parallel, a speed-droop governor should be used for d-c units or for a-c units where the frequency may vary with load. Where a constant frequency is required, isochronous governors with load-limiting features or electric governors should be used. Hydraulic governors are generally capable of controlling diesel generator sets so that the steady-state speed regulation is between 0 and 1 percent, the steady-state governing speed level is within f0.25 percent of rated speed, the momentary underspeed is within 7.0 percent of rated speed, and the recovery time following overspeed or underspeed is less thaa 5.0 sec with a prescribed speed band of f0.5 perkent of rated speed. Electric governors are mandatory when close regulation of frequency is required. All generator set engines should be equipped with an overspeed trip as well as a regulating governor. Governors are available which will regulate the engine power in response to various sensors in order to control the output of the driven equipment. As an example, pumps may be driven to deliver a prescribed flow or pressure or to maintain a constant level in a tank. 4.10 Instrumentation. The following instrumentation should be installed on an instrument panel attached to the dngine, .adjacent to the engine, or located at the engine control stand in direct view of the engine operator (asterisks indicate the only instruments normally furnished with automotive-type engines) : (a) Pressure gages 1 Freshwater pump discharge
2 *3 4 5 6
Seawater pump discharge Lubricating oil pump discharge Lubricating oil pressure at engine inlet Piston cooling oil pump discharge (if used) Fuel oil pump discharge 7 Fuel oil pressure at injection pump inlet
MEDIUM AND HIGH-SPEED DIESEL ENGINES
MARINE ENGINEERING
8 Scavenging air or supercharging air pressure 9 Starting air pressure a t engine 10 Lubricating oil at turbocharger or blower 11 Crankcase pressure or vacuum (b) The~momete~s *1 Lubricating oil from engine *2 Freshwater from engine 3 Fuel temperature to engine when heated heavy fuels are used *(c) Tachometer (d) Ezhaust gas pi~ometerindicator I n addition, pressure gages and thermometers should be installed in the piping a t each heat exchanger, filter, and strainer as required for troubleshooting purposes. Engine manufacturers are not in agreement regarding the value of exhaust gas thermocouples and pyrometers. The exhaust gas temperature a t each cylinder discharge into the exhaust manifold is not the same under normal conditions. This is due to variations and pulsations in the airflow to each cylinder and manufacturing variations which affect factors such as the compression ratio, combustion chamber shape, and the fuel nozzle spray pattern. Pyrometers should not be used to set engine fud racks to obtain equal exhaust temperature from each cylinder in the hope that this will equalize the load between cylinders. The proper way to insure equal load between cylinders is to set the fuel racks such that equal quantities of fuel are injected into each cylinder. However, thermocouples and pyrometers are useful for detecting a change in the exhaust temperature of any one cylinder with time as an indication of abnormal cylinder performance. Pyrometry equipment is available in a variety of forms such as: ( a ) each cylinder's thermocouple connected to a pyrometer through a selector switch which is manually operated; (b) each cylinder's thermocouple connected to its own pyrometer; (c) thermocouples indicating the combined exhaust from all cylinders in a bank; (d) thermocouples a t the inlet and exit from the turbocharger; (e) cylinder thermocouples connected t o a pyrometer through a motorized switch together with circuitry in the pyrometer to average the cylinder exhaust temperatures, compare each cylinder to the mean, alarm wheri any cylinder exhaust exceeds the mean by a set amount, and identify the offending cylinder. Sensors are available to measure vibrations,(in terms of amplitude, velocity, or acceleration), bearing temper* tures, bearing weardown, hour meters, power meters, pressures, pressure differences, rates of pressure rise, and rates of change of pressure differences and the same for temperatures. The extent to which this advanced instrumentation of an engine installation is considered is related to the degree to which the plant is automated, the skill of the operators, and the nature of operations, as well as the desired initial cash outlay. 4.1 1 Automalion. There are two justifiable reasons for automating a diesel plant, and both are related t o
saving money. The first reason is to reduce manning and save labor costs; the second is to reduce casualties by . providing better diagnostic data and by eliminating o*rator errors. The difficulty is deciding, for each installation, how far to go. The diesel engine is inherently self-regulating and easy to automate. Most medium and high-speed engines are equipped with enginedriven pumps and governors. Oil pressure regulating valves and thermostatic control of water and oil tempers tures have been commonplace for years. Starting is a simple operation involving no more than several sequencing devices t o control a minimal number of switches or-valves. Turbochargers are self-regulating. Starting from a cold plant, for example, with all sea valves closed involves a sequence which would open the sea valves using remotely controlled actuators, heat jacket water electrically to a given temperature, start an electric motor-driven lubricating oil priming pump, shut down the pump, and energize the starter. When the engine reaches idling speed, the starting equipment is secured. If i start is not achieved in a preset time, the starter is disengaged and the cycle is repeated. This type of operation can be done electricilly, hydraulically, or pneumatically. It is possible to program propulsion controls for multiple engine drives to add or remove engines a t a predetermined speed. The engines removed from the line can be shut down inasmuch as the starting cycle for a warm engine is extremely short and simple, taking less than 10 sec. Electric generating plants can be completely automated to operate unattended, automatically adding and removing engines as dictated by the load. Synchroniaing equipment is available as part of the engine governing system to parallel alternators automatically. Sequene ing equipment can be set up so that generators are used in a specified order to equalize the number of hours on each unit if desired. Automatic shutdown of an engine due to the loss of oil pressure, high water temperature or any other malfunction, will initiate the starting cycle for a designated stand-by, so that operations will not be interrupted as long as stand-by equipment is available. Distilling plants and other engine room accessories can be automated as well so that it is technically possible to remove all operators from the machinery spaces and to control the ship from the bridge or any other desired station. I n addition to automating the plant operation, there are many devices for automating data logging and analysis. Logging may be done in digital form and recorded on typewritten sheets or in analog form as curves of various parameters versus time. Logging may be required for all data or only for readings which are approaching preset limits. An alarm usually sounds when these type of data are recorded. Graphical presentations are very useful as scales can be compressed to require less paper and clearly show trends. For example, a curve of differential pressure across a filter element versus time can be extrapolated to show when
t I
4
d
the filter will require changing, giving the maintenance crew advanced notice to plan the best time for the change. A plot of crankcase pressure versus time may show a change which would be indicative of increasing blowby and tde necessity to renew piston rings. In this manner, engine overhauls can be accomplished only as needed with resultant savings in expenses as opposed to conducting overhauls a t fixed time periods. Another form of automation is the use of computers to compare recorded data with standards, for the particular load and speed, which are stored in memory. Deviations from the norm are recorded and alarmed. The more sophisticated automated plants are controlled by electronic equipment, much of which is in the form of printed-circuit cards with self fault-finding features to minimize the skill levels required to maintain the automatic equipment. In addition to the electronic brains, muscles are required to open and close valves and move the throttle lever. These elements all add to the complexity of a completely automated plant. For a more comprehensive discussion of automation considerations, see Chapter 21. 4.12 Installation Items. Foundations for engines and driven equipment must be sufficiently rigid to maintain alignment when the ship's hull is working in heavy seas. Automotive and railroad-type engines are frequently designed to be mounted on foundations with three or four points of support. This has the advantage of making the engine relatively insensitive to minor working of foundations. Consideration should be given to the use of flexible couplings in cases when alignment cannot be assured. If the engine is mounted on rubber or other types of flexible mounts, flexible couplings are easenti&l. Diesels should be bolted to their foundations with fitted bolts at the drive end only. Clearance bolts should be used in all other locations to permit the engine to expand away from the driven equipment as it heats. Diesel engines are usually provided with a thrust bearing on the crankshaft to keep the shaft in place during shipment. Propeller thrust is taken in a separate thrust bearing housing or in a bearing built into the reduction gear housing. It is essential that the engine thrust bearing clearance be greater than that in the propeller thrust bearing or that the coupling between the engine and the gear provide for longitudinal movement a t least equal to the thrust bearing clearance, and that the engineto-driven-gear alignment be such that the crankshaft is centered in its clearance to prevent the thrust bearings from bucking each other. The same situation must be guarded against in generator sets where the generator and engine each has its own thrust bearing. Medium and high-speed engines have relatively small parts which are easily handled, provided they are accessible to the mechanic while he keeps both feet on the deck. Ctire should be taken to insure that ducts and wireways do not interfere with access to regular maintenance for filters, cylinder-head covers, air cleaners and so forth. Gages and sight glasses should be placed such that the
277
operator can see them without leaving the control station. For medium-speed engines, chain falls and pad eyes are required to pull cylinder heads, pistons, and liners. High-speed engines generally do not have handhole covers in the cylinder block for access to bearings; in these engines, it is necessary to drop the oil pan to renew bearings. Provision should be made, at the time the ship is designed, for bearing renewal, either by providing space under the engine to drop the oil pan and move it out of the way, or to lift the engine off its foundation and roll it over. There are many excellent texts today on the subject of torsional vibration and balancing of engines [42-451. The problem of balancing engines to minimize unbalanced forces and moments has been solved to a great degree by the engine manufacturer's use of computer programs to select optimal firing orders, crankthrow arrangements, and locations for balance weights. Should problems arise due to resonance of engine unbalance forces and foundations or hull structure, they are usually found after ship trials are run and correction consists of local stiffening of hull structure. The type of vibration encountered depends t o a large degree upon the type of hull; for example, the most common vibration in a towboat is vertical [46,47]. Noonan and Zaloumis [48] point out the importance of endeavoring to select machinery which has vibration characteristics m i s matched to hull natural frequencies. If possible, engines should be located a t the nodes of hull vibrations. Torsional vibration, a t one time, was a major problem with diesel engines. Today it is a well-understood phenomenon and engine manufacturers have computer programs for their engines so that they can readily assess the effects of torsional vibrations in various applications. With the use of pendulum or viscous dampers on the free end of the crankshaft and torsionally flexible couplings or fluid drives between the engine and gear, it should be possible to avoid problems due to torsional vibrations. In any case, the engine manufacturer should conduct torsional vibration analyses to ascertain freedom from dangerous critic& in the operating range of the engine. Each new installation should be torsiographed to insure that it is safe. 4.13 Safely Features. Alarms and automatic shutdowns sbould be used to prevent catastrophic failures of engines. On propulsion engines, it is general practice to not shut t h e engine down but to give the captain or pilot the option, in case of an alarm, to either shut the engine down if in a safe situation or to risk wrecking the engine if the ship would be endangered if left without power. In the case of a generator set, automatic shutdown is the general practice. The items which trigger alarms or shutdowns usually are: (a) low lubricating oil pressure with the sensor installed a t the end of the oil manifold remote from the oil pump; (b) high jacket water temperature; and (c) high rpm, i.e., an rpm exceeding 115 percent of rated rpm for ABS vessels. Crankcase explosions are not common in diesels, but when they occur they can be dangerous. An explosion
278
MARIhE ENGINEERING
can occur when the mixture of air and oil vapor in the crankcase is ignited by a spark or hot spot resulting from a part failure. The first explosion is generally weak, due to the rich fuel-air mixture in the crankcase, and the peak pressure does not exceed 25 psig. Should the explosion blow off a cover or should someone open a cover imrriediately after the first explosion, fresh air will rush in and a second, much more violent explosion will occur. There are two ways to pi-event explosive d a m a-e : 1 Design the crankcase strong enough so that the first explosion cannot lift a cover or otherwise let air in. 2 Provide spring-loaded covers to permit the first explosion to relieve and have the cover close rapidly, air out and preventing a ex~losion. Navy specifications and ABS rules require relief valves; the Navy requires them on engines over 6-in. bore and ABS requires them over 8-in. bore. Navy specifications require 1.5 sq in. of relief valve area per cubic foot of crankcase volume, the ABS requires 0.5 sq in. per cubic foot of crankcase volume. Both require warning notices to be posted on the engine cautioning against opening a hot crankcase after an explosion. Engines burn fuel oil which is volatile and combustible and must be contained a t all times. Highpressure fuel lines should be shielded from hot engine parts by the use of a flexible tube around each pipe or a shield over the manifold or the use of water-cooled manifolds. . Fires have occurred when a high-pressure fuel line cracked and sprayed fuel over a hot exhaust elbow. Crankcase vents should be directed to the engine air inlet or overboard. Fumes should not be ducted to the ventilation system where oil will condense and present a major hazard in case of a fire. Engines should be equipped with an emergency shutdown device operated by a pull cable which will trip the fuel racks or shut 08 the air and stop the engine within 60 sec. The pull cable should be installed so as to be operable from a location adjacent to the engine room access so that in case a fire forces abandonment of the engine room, the engine can be stopped. Shields should be provided to protect personnel from hot parts and moving parts in case they are accidently thrown against the engine. Direct-reversing engines should be provided with an interlock to prevent injection of fuel during a reversal until the engine comes to a complete stop. References
1 C. Jones, "New Rotating Combustion Power Plant Development," Trans. SAE, 1966. 2 B. W. Wadman, Diesel and Gas Turbine Catalog, Diesel and Gas Turbine Progress, Milwaukee, Wisc., vol. 33, (Published annually). 3 British Diesel Engine Catalogue, issued by The British Internal Combustion Engine Manufacturers Assn., London.
MEDIUM AND HIGH-SPEED DIESEL ENGINES .
4 K. W. Stinson, Diesel Engineering Handbook, Diesel Publications, Inc., Stamford, Connecticut, 1963. 5 A. K. Antonsen, "The Development of a Supercharged Medium Speed Two-Cycle Opposed Piston Engine, " Proceedings, ASME, OGP, 1956. 6 James R. Ware, "Development of a Turbocharged Two Stroke Twenty Cylinder Diesel Engine for Marine Application, " SAE paper 670949, 1967. 7 J. M. .Moriarty and C. H. Schowalter, "Application of MediumS~eedDiesels to Marine Pro~ulsion." SNAME Spring Mketing, May 1966. 8 Bob Schulz, ,,Mack,s New Maxidyne Constant Power Vehicle Package, " Diesel and Gas Turbine Progress, December 1966. . Bruce W. wadman, ,,Trends in Tailored Truck, Diesel Torque and Homepower Characteristics, Diesel and Gas Turbine Progress, May 1967. 10 H. L. Wittek, "Development of Two New Allis Chalmers Diesel Engines, " Trans. SAE, 1960. 11 W. J. McCulla, "How a Diesel Engine Rates Itself," Tram. SAE, 1959. 12 P. H. Schweitzer and C. G. A. Rosen, "Whither the European Automotive Diesel?" Trans. SAE, 1964. 13 M. V. Kienlen and G. W. Maybach, "High Speed High Output Diesel Engines-35 Years of Development of Railroad and Marine Applications, " Trans. SAE, 1962. 14 L. Wechsler and H. Holler, "Development of a Lightweight High Output Diesel Engine for Naval Service," ASME, Paper no. 58-OGP-2, May 1958. 15 "Engine Rating/Why and How," Diesel Equipment Superintendent, April 1964. 16 SAE Handbook, published by Society of Automotive Engineers, (New issue each year). 17 W. A. Kilchenmann, "Slow Speed Versus High Speed Diesel Engines for Ship Propulsions," Naval Engineers Journal, June 1964. 18 K. Zinner, "A Comparison of High Powered Single Engine and Multi-Engine Plants for the Propulsion of Merchant Ships," ASME, Paper no. 67-DGP-2, April 1967. 19 J. Neumann and J. Carr, he Use of Medium Speed Geared Diesel Engines For Ocean-Going Merchant Ship Propulsion, " Journal, Institub of Marine Engineers, 1966. 20 R. Fredrikson, "The Medium Speed Diesel-An Engine of Increasing Interest to the Shipowner and Shipbuilder, " SNAME, 1968 Diamond Jubilee International Meeting. 21 W. Hempel, "Why has the Medium Speed Diesel Become Competitive in Marine Propulsion," SAE, Paper no. 670950, November 1967. 22 E. A. Butler, R. Kaufman, and T. V. Pedersen, "Advanced-Design Motor-Ship Machinery Plant-20000 SBP," Marine Technology, vol. 4, no. 4, October 1967. 23 F. A. Christiansen and P. I. Brown, "Military and Manufacturer Specification Oils, Their Evaluation and Significance," Trans. SAE, 1963. 24 Rules for Building and Classing Sbel Vessels, American Bureau of Shipping, New York.
279
25 C. C. Moore and W. L. Kent, "Effect of Nitrogen Governing of Electric Generating Sets," MIL-STD-178, and Sulphur Content of Fuels on Diesel Wear," Trans. Superintendent of Documents, U. S. Government PrintSAE, 1947. ing Office. 38 W. H. Krogstad, "Control System Programs 26 H. V. Nutt, E. W. Landen, and J. A. Edgar, "Effect of Surface Temperature on Wear of Diesel Power for Ships with Controllable Pitch Propellers," Engine, Cylinder and Piston Rings," Trans. SAE, 1955. SAE Journal, December 1962. 27 B. A. Robbins, P. L. Pinotti, and D. R. Jones, 39 F. Schanz, "The Controllable Pitch Propeller as "The Use of Radioactive Tracer Techniques to Deter- an Integral Part of the Ship's Propulsion System, " Trans. mine Effect of Operatirig Variables on Wear," Trans. SNAME, vol. 75, 1967. 40 P. K. Wennberg, "The Design of the Main SAE, 1960. 28 Total Energy Handbook, Caterpillar Tractor Co., Propulsion Machinery Plant in the USCGC Hamilton (WPG-715)," Trans. SNAICIE, vol. 74, 1966. Peoria, Ill:, Form M E 0 26690.01, 1967. 41 R- lihuschildt and C- Miller, "U- S- Navy 29 Manual of Heat Recovery, Fairbanks Morse and PG 84 Class CODOG Propulsion Plant," SAE Paper no. Co., Beloit, Wis., Form P 295, 1960. 196730 Vapor P b e Engineering Manual, Engineering 670952, 42 W. I
JJ
I
#
8 i?
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
CHAPTER V l l l
Table 1
BORIC,
Kuri Illies
Low-Speed Direct-Coupled Diesel Engines
in. (mm)
STROKE,
in. (mm)
Typical Engine Design Data
PISTON SPEED, SPEED, rpm fpm (m/s)
BMEP, Psi (kp/cmS)
NO. OF CYL.
MAX.CONT. OVERALL RATING, WEIGHT, LENQTH, bhp tons ft-in. (mm)
Section 1 Survey of Principal Characteristics 1.1 Scope. The possible alternatives when selecting the type of propulsion plant were briefly discussed in Chapter 1. It was noted that of the more usual prime mover selections, only low-speed diesel engines are directly coupled to the propeller shaft. This is due to the low rpm required for efficient propeller operation and the high rpm inherent with other types of prime movers. Overall technical considerations with regard to diesel engines were covered in Chapter 7, which dealt with medium- and high-speed diesels, and therefore will not be repeated here. Instead, only those characteristics which differentiate low-speed engines from medium- and highspeed engines are discussed. Medium- and high-speed engines generally are not designed specifically for marine applications; therefore, the design details of these engines were not emphasized in Chapter 7. The situation is different with low-speed engines, which are designed and manufactured specifically for marine applicatione. For this reason, more emphasis is placed on the design details of low-speed engines. Additional discussion regarding low-speed diesels may be found in works such as references [I] through [7].' 1.2 Engine Data, The continuous rpm &ing of direct-coupled diesel engines is generally in the range of 100 to 225 rpm. Lower speeds are seldom selected due to the strong sensitivity of engine weight to speed as indicated by Fig. 1. The low engine q m in conjunction with piston speed limitations (about 1300 fpm) -require long piston strokes. This means that the engine must be high, wide, and consequently heavy. Higherpiston speeds impose higher inertia loads, bearing reactions, and the l i e , which can have an overriding impact on the design of an engine. As an indication of the effect of piston speed limitations, for two-stroke crosshead engines the piston speed has increased only 20 percent during the past 20 years. Further increases in piston speeds are expected to be minor because they must be made at the expense of either higher engine speeds or larger engine dimensions, both of which are undesired. Nominal design data for an array of low-speed engines,
Numbers in brackets designate Referencesat end of chapter.
KP/PS
LB/BHP
r
40
EXHAUST BEFORE TUF
110
50,
w
Q
88
s
AFTER CYL.
66
30;
C
6 20
;
I
A Y . PRESSURE
PRESS.
t 1150
44
V)
10
2
22
z
0
w
100 500
1000
ENGINE RPM Fig. 1
Weights of low-speed engines
(a) loop ravenging
representative of those manufactured, are shown in Table 1 ; the weights shown in Table 1 are without water and oil. Low-speed diesel en&es are two-stroke, single-acting, and designed with ~rossheads. Due to weight and dimension considerations, they are used only in merchant ships and not in naval vessels. The ratio of stroke to bore also must be considered it has an impact on s venging efficiency- To a great extent, the method 0 scavenging employed determines the stroke/bore ratio permitted. The various methods of scavenging are illustrated by Fig. 2. With the loop and cross-scavenging methods, the strokebore ratio usually falls in the range of 1.8 to 2; higher values of 2 to 2.1 are used with single pistons employing the uniflow method; and in opposed-piston engines, the combined stroke is about 2.9 times the bore. Large cylinder dimensions and supercharging permit high cylinder horsepower ratings. Engines with a rating of 3700 bhp/cylinder have been installed aboard ship, and the tendency is for this value to increase. Outputs of 5200 bhp/cylinder have been obtained on the teat stand. It should be noted that under normal operating conditions a t sea, the engine is loaded to only about 9095 percent of the maximum continuous rating. Typical design data for cylinders of two different ratings are shown in Table 2. Operational data for a spe-
r
(b) (c) (dl Fig. 2
-
Ov3' 0
cross scavenging uniflow ravenging with exhaust valves uniflow ravenging opposed pistam
0.31
FUEL 3CONSUMPTION
-
4
Scavenging methods of low-sped engines
ENGINE Fig. 3
RPM
Characteristic data for a low-speed engine
[a]
cific engine design are shown in Fig. 3. The fuel consumptions indicated in Table 2 and Fig. 3 do not include lubrication and cooling pumps; additionally, they are Table 2 Typical Cylinder Design Data based on a fuel lower heating value of 18,000 Btu/lb. CYLINDER DESIGN B Disregarding electrically driven auxiliaries, a heat Output/cylinder, bhp. ............ 610 A 4000 balance for low-speed engines is approximately as follows Engine speed, rpm.. .............. 225 103 Piston diameter, in. (mm). ........ 20.5 (520) 41.7 (1060) (based on the lower heating value of the fuel) : Piston stroke, in. (mm). ........... 31.5 (800) 78.7 (2000) Piston speed fpm (m/s). .......... 1024 (5) 1360 (7) Heat to power. .................... - 3 9 4 2 % BMEP, psi (kp/cma).............. 103 (7.3) 143 (10.1) Heat to exhaust gases. . . . . . . . . . . . . . .25-35% 996 (70) Compression pressure, psi (k /cma). 498 (35) Heat to cooling water. . . . . . . . . . . . . . .20-30% Max. ~ l i n d e r ressure, psi (ep/cml) 711 (50) 1280 (90) ~ i ratio-com~ustion.. r ........... 1.6 2.2 Remainder. . . . . . . . . . . . . . . . . . . . . . . . . 3 4 % Air ratio-total. ................. 2.5 5.0 temperature after cyl. It is common practice to reduce the heat losses by Exhaust Loo and cross-scavenging, deg F employing heat recovery schemes. For example, an . . . . . . . . . . . . . . . . . . . . . . 625 (330) 700 (370) Uniflow scavenging, deg F (C). . . 635 (335) 800 (425) exhaust gas auxiliary boiler may be used to produce temperature before turbine steam or hot water for preheating heavy fuel, and the Exhaust Loo and crossscavenging, deg F engine cooling water may be used to produce fresh water . . . . . . . . . . . . . . . . . . . . . . . 665 (350) 755 (400) Uniflow scavenging, deg F (C). . . 665 (350) 860 (460) in a seawater evaporator. Mechanical efficiency, percent.. .... 85 93 1.3 Principal Structure. Sections through a typical Cylinder oil consumption, lb/bhp-hr ( /PSh). ...................... 0.0015 (0.7) 0.0007 (0.3) cross-scavenging engine are shown in Pig. 4, and Fig. 5 is consumption, ib/bhp-hr rt section through a typical loop-scavenging engine. (g/PSh)....................... 0.32 (146) 0.34 (155)
(8).
(8).
~ u 3
7
-
-
w
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
MARINE ENGINEERING
ENTABLATURE
BORE FOR PISTON AND STUFFING BOX
CROSSHEAD GUIDE-RAILS CROSSHEAD
BORES FOR TELESCOPIC PIPES
MAIN BEARING
OIL SUMP Fig. 4
Typical uou-scavenging engine [9]
HEAD
CYLINDER HEAD
CYLINDER BLOCK ENTABLATURE STUFFING BOX
Fig. 5
Typical loop-scavenging engine [El
The engine frame basically consists of the bedplate, columns, and cylinder block; but in some cases a top deck is provided between the columns and cylinder block. The frame of an engine is shown more clearly by Fig. 6, which is a detailed view of the engine shown in Fig. 5. The bedplate consists of longitudinal'girders of high structural rigidity which provide substantial support for the main bearings. It may be seen that the basic engine frame is held together by prestressed tie rods which extend from the top of the cylinder block to the lower part of the bedplate. The firing forces are taken by these tie rods, and the other structural components are subjected to compreseive loads only. Crankcase doors fitted with explosion plates (see Fig. 5) are arranged between the columns. The opposed-piston engine shown in Fig. 7 has some interesting characteristics. The stroke of the upper piston is shortened so as to equalize the inertia of the mass associated with the side and center cranks. The forces from the combustion loads on the upper and lower pistons are carried entirely by the running gear connected to the crankshaft; thus, there are no combustion loads on the main bearings or the engine structure. Reference [13], which discusses the design of low-speed diesels in
Fig. 6
Frame without cylinder block [S]
general, describes some of the characteristics of opposed- and 9 illustrate techniques which have been used with piston engines. engines employing the loop-scavenging method. In each In cases where the engine is not equipped with tie case, the lower sides of the pistons as well as the turborods, the columns and cylinder blocks are welded to- charger act as scavenge air pumps. gether and the columns are bolted to the bedplate by With the schemes shown in Fig. 8 when operating in means of heavy bolts. The firing forces are then trans- the higher load ranges, the lower sides of selected pistons mitted from the oylinder head through the columns to deliver scavenge air through the cooler directly irito the the bedplate. charge air line in parallel with the turbocharger. DurFormerly, extensive use was made of cast iron in ing part-load operation, the air supply provided by the engineTrames, but more recently the bedplate, columns, turbocharger operating in parallel with the lower side and top deck have been made of welded steel fabrications of the pistons is not sufficient. In this case, an injector as a means of reducing weight. For rigidity purposes, system [as shown by Fig. 8(a).] or a compressor-drive however, the main bearing supports are usually heavy system [as shown by Fig. 8(b)] may be used to increase steel castings. the air qhantity delivered by the turbocharger a t low Engine cylinders are always arranged in line. I n order speeds. With the injector system, in the lower load to obtain a uniform torque output, the minimum number range the air supplied by the lower sides of the pistons of cylinders is limited to four. Additionally, design com- flows a t a high velocity through injector nozzles, which plications with regard to the crankshaft and engine bed- are arranged immediately following the turbocharger, plate and engine length considerations limit the maxi- and then into the air receiver. This action induces air to mum number of cylinders to about twelve. Within these flow through the turbocharger and stabilizes its operating limits, there is considerable freedom in selecting the condition. With the compressor-drive system, in the cylinder desim and number of cylinders to produce the lower load range the air supplied by the lower sides of the desired output. pistons is admitted to the compressor wheel of the turbo1'.4 Scavenging Systems. As discussed previously charger so as to produce additional torque to drive the and illustrated in Fig. 2, engines may employ either the turbocharger when relatively little exhaust energy is loop, cross, or uniflow method of scavenging. Figures8 available.
LOW1-SPEED DIRECT-COUPLED DlESEt ENGINES
MARINE ENGINEERING
(a) Infector system cylinder air receiver exhaust receiver 4 exhaust turbine 5 air blower 6 silencer 7 underside pump Fig. 8
1 2 3
(b) Compresmr drive system 8 9
oil damper separator (high load)
9a damper (part load) cooler coder infector pipe compreuor drive Loop-ravenginsI air supply system [8]
10 11 12 13
pression is accomplished by the exhaust gas turbocharger, and the second takes place in the double-acting piston scavenge pump; both steps are aligned in a series arrangement. Each cylinder has its own double-acting reciprocating pump that is directly driven by the crosshead. The intake and exhaust valves of these pumps, as well aa the scavenge air intake valves of the power cylinder, are designed as automatic nonreturn valves. Some engine designs include a small electrically driven blower which is used up t o quarter load and when maneuvering. Such an engine is illustrated by Fig. ll. The electric blower considerably improves engine acceleration and gives a clear exhaust. The blower is switched on and off automatically a t a certain position of the fuel lever and requires only about 0.2 percent of the engine output. Many of the turbocharged engines using the uniflow Fig. 9 Sen'er-parallel air supply system [8] method of scavenging are capable of meeting scavenge air requirements a t all engine speeds with no assistance. Some designs, however, include scavenging pumps, such Another air supply method is the series-parallel sys- as shown in Fig. 10, and some designs use small electrically tem as illustrated by Fig. 9. The lower sides of all driven blowers for assistance a t low power levels and for pistons are designed as scavenge air pumps with this cold starting. arrangement. Some act in parallel with the turboWith the uniflow scavenging arrangement shown by charger; however, others may shift between series and Fig. 12, the scavenge ports around the lower part of the parallel operation. During part-load operating condi- cylinder liner are controlled by the piston; the exhaust tions, the piston lower sides arranged in series pump air valve in the cylinder head is actuated by means of a cam from the turbocharger to the air receiver. At higher on the camshaft, and the movement is transmitted to loads all piston lower sides operate in parallel with the theUniflow valve through scavenging a rocker for an arm.opposed-piston engine is turbocharger. An automatic valve arrangement is provided to ensure a proper alignment. shown by Figs. 7 and 13. Figure 12 displays the uniAn air supply system which has been used with engines flow scavenging system for a single-acting engine. In empl'oying the cross-scavenging method is shown by Fig. these designs the turbochargers are capable of supplying 10. It may be seen that the scavenge air is compressed in scavenging and combustion air under all circumstances, two steps and cooled after each step. The first com- including starting, without a supplementary pump.
MARINE ENGINEERING
LOW-SPEED DIRECT-COUPLED DRSEL ENGINES
r?Tl
EXHAUST
TURBOCHARGER
AIR COOLER
SCAVENGE AIR RECEIVER COOLING OIL. INLET
Fig.
1 2 3
4
air compressor 5 scavenge air receiver air cooler 6 combustion space ravenqe pump 7 exhaust receiver air cooler 8 exhaust turbine Fig. 10 Cms-scavenging air supply system [I
11
The scavenge ports a t the bottom of the cylinder are controlled by the lower piston, and the exhaust ports a t the top of the cylinder are controlled by the upper piston. Both scavenge and exhaust ports extend completely around the cylinder periphery. For additional discussion regarding the scavenging of slow-speed diesels, see references [14-161. 1.5 Cylinder Design, The cylinder head, cylinder liner, and piston form the combustjon chamber. These parts are bighly stressed from both a mechanical and thermal point of view. Higher su~erchargingand larger cylinder dimensions incur correspondingly higher stress levels. Simply increasing the thickness of the parts forming the combustion chamber as a means of reducing mechanical stresses is not a satisfactory ~olutionbecause an increased wall thickness would cause increased temperature gradients and higher thermal stresses. As a result many engines employ a "backing system." That is, the wall thickness of the components forming the combustion chamber is minimized by supporting the back
11
Air supply system with blower assist
[9]
side of the walls with a reinforcing structure. A cooling medium is circulated through the reinforcing structure; therefore, satisfactory solutions are obtained to both the mechanical and thermal stress problems. A typical cylinder "backing system" is illustrated by Fig. 14 for a loop-scavengingengine. The cylinder jacket is made of cast iron, and passages for scavenging air, exhaust gases, and cooling water are cast in place. An opening opposite the ports a t their level permits an inspection of the ports in the liner and a region of the piston without disassembly. The cylinder liner shown consists of two parts; the upper liner, which is made of wear-resistant cast iron, contains the scavenge and exhaust ports, which extend for a large arc of the liner circumference. The lower liner serves the primary function of guiding the piston while in the vicinity of the lower dead-center position. The top of the liner is supported by a steel backing ring which is part of the backing system. The liner and port lands are water cooled. The cylinder design of various engines differs in detail, a major distinction being the method of scavenging employed. For example, engines which operate with a uniflow scavenging system do not require water-cooled air ports, as cooling by air is sufficient. Figure 15 shows the combustion chamber of a uniflow-scavenging engine (the figure is a magnification of the combustion chamber shown in Fig. 12). The liner flange is reinforced by a steel band that is shrunk on; holes are drilled into the liner to facilitate cooling with water.
Fig.
12 Uniflow-scavenging air supply system with exhaust valve [I21
1 2
lower cylinder liner 7 upper cylinder head inspection port 8 inleetion valve 3 upper cylinder liner 9 cooling water 4 cylinder lacket 10 exhaust porn 5 steel backing ring 1 1 air inlet porn 6 lower Fig. cylinder 14 Cylinder head of a loop-scavenging engine [8]
r FUEL
INJECTION VALVE
CYLINDER JACKET
CYLINDER LINER
Fig.
13 Uniflow ;cavenging for an opposed-piston design [I01
Fig.
15 Combustion chamber of a uniflow-scavenging engine [12]
288
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
MARINE ENGINEERING
289
PISTON CROWN CONE - COOLING WATER
GUIDE WITH BRONZE
MAIN PISTON ROD PIS
COOLING WATER O U T L E T
LOWER PISTON PART SCUFFING B A N D S
G A S K E T RING
CROSSHEAD-
OIL FROM MANIFOLD COOLING WATER INLET COOLING WATER O U T L E T
CROSSHEAD
Fig. 17 Piston and piston md of a loop-scavenging engine
[a]
a
N G l N G PIPES
OIL DISCHARGE
Fig. 19 Oil-cooled piston with swinging pipes housing 11 securing ring 12 guide rod steel seat 13 guide bracket valve spindle 14 guard Aame guard 15 lower outer spring spindle lower guide 16 lower inner spring spindle upper guide 17 intermediate spring plate bottom guide bushing 18 bottom spring plate top guide bushing 19 lower guide bushing spring keeper 20 sealing ring split collar Fig. 16 Exhaust valve for a uniflow-scavenging engine [I 21
Cylinder heads (or covers) are reinforced thin-walled (see Fig. 12) and are closed by two sets of concentric castings that are water cooled. With many engine de- helical springs. . signs, the cylinder heads are carried rather far down such Some uniflow exhaust valve arrangements employ a that they enclose the pistons when the pistons are in the number of exhaust valves, as opposed to the single valve top dead-center position. This confines the region sub- shown in Fig. 16, but their principle of operation is the jected to the very hot gases to the cylinder head and pro- same. tects the cylinder liner. The firing pressure is t r a m 1.6 Piston and Piston Rod. The pistons of a loop- or mitted from the cylinder head to the cylinder jacket by cross-scavenging engine must be long because the lower the cylinder head studs. part has to cover the scavenge ports when the piston is The exhaust valves in uniflow-scavenging engines are in the top dead-center position. The piston is constructed highly stressed thermally; therefore, heat-resistant steels in three parts; see Fig. 17. The top part, the crown, is must be used in their construction. The cylinder head a steel casting in which the piston ring grooves are mashown in Fig. 15 is provided with a central large orifice chined. The casting is flame hardened in way of the for the exhaust valve, and bores are provided at the piston ring grooves to minimize wear in service. The sides for the fuel valves, starting valves, safety valves, lower part is a grey casting which is grooved to accomindicator cock, and the cooling water connection to the modate a scuffing band made of leaded bronze. The exhaust valve. A detailed illustration of the exhaust center part is a cast iron guide band which also contains valve which fits into the cylinder head in Fig. 15 is leaded bronze scuffing rings. The piston shown in Fig. shown by Fig. 16. The lower part of the spindle guide 17 is water cooled. Cooling water admission and disis water cooled. The valves are opened by cam action charge passages are within the piston rqd, as can be seen
[I 1]
water and also guards against fouling of the cooling water by splash oil. Oil is used to cool the pistons of some engines. A major distinction between the water-cooled and oilcooled pistons is that minor leakage of the cooling medium is not of consequence with the latter. This addition91 degree of freedom permits an arrangement such as indicated by Fig. 19. As with the water-cooled piston, the coolant is introduced at the crosshead; but in this case a swing-pipe scheme, which is characterized by minor leakage at the connecting joints, can be used since absolute containment of the coolant is not essential. For additional discussion regarding piston cooling and lubrication se3 Sections 2.2 and 2.3. The pistons in engines with a uniflow-scavenging system ar? short. Figure 20 (see also Fig. 12) is a section through an oil-cooled piston for a uniflow-scavenging engine. The molybdenum steel piston head is clamped to the piston rod by a cast iron guide skirt and an annular spring of the Belleville type. Piston cooling oil enters the outer annular cooling space with a high tangential I velocity and from there enters the central cooling space Fig. 18 Teleuopic pipes for piston cooling [8] with a high swirling velocity. 1.7 Diaphragm and Stuffing Box. The diaphragm from Fig. 17. The means provided to deliver the cooling and stuffing box for the piston rod form a barrier between water to and from the piston rod are illustrated by Fig. the combustion chamber and the crankcase. The pur18. Telescopic pipes, with their stuffing boxes, are pose of the barrier is to prevent harmful combustion located outside the crankcase; this arrangement ensures residues from entering the running gear and thus conthat there will be no contamination of the lube oil by taminating the lube oil and creating a corrosion hazard.
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
MARINE ENGINEERING
290
PISTON ROD
CYLINDER LINER
I
\
\
1I
THE ENGINE ROOM ! Fig. 21
( A ) CROSSHEAD
Diaphragm with stuffing box [I 11
* Fig. 2 0
Oidwoled pisbn for a unitlow-rcovenging engine
(A) CONCENTRATED PRESSURE Fig. 2 2
(9) UNIFORM
PRESSURE
D e A d o n of crossheods under load [9]
( 0 ) CONNECTING
Fig. 2 4
ROD AT C R A N K P I N
1 bracket for telescopic pipes 2 lubrication oil pump Crorrhead and connecting md with guides on only one side
cn.i
cn. n
cn.r
[a]
29 1
Additionally, the stuffing box w i ~ e soff oil that adheres to the rod and thereby controls the leakage of oil from the crankcase. A stuffing box that has been designed for an engine with the space beneath the piston open to the engine room is shown in Fig. 21. There is one upper sealing ring and two lower rings. Each ring is in three pieces held together by garter springs. For designs in which the lower sides of the pistons are used to pump scavenging air, the sealing arrangement shown by Fig. 21 is not suitable as an additional sealed barrier is required. In this case, a design similar to that in Fig. 11 may be used. Two sealed barriers are provided. One is a seal between the scavenge air and the ambient engine room air, and the other is a seal between the crankcase and the ambient air (similar to that in Fig. 21). This arrangement precludes the leakage of scavenging air into the crankcase. 1.8 Crosshead and Connecting Rod. To minimize the forces imposed upon the cylinder liner by the piston rod, low-speed diesel engines are designed with crossheads. The crosshead pin bearing is a particularly highly stressed element. The reasons for this are that the motion i s oscillatory and not continuous and that the pressure on the crosshead pin is always in the same direction during both the expansion and compression strokes. Under such adverse conditions, a load-carrying hydrodynamic oil film is hardly able to form. An additional complicating factor is that the crosshead pin is of relatively small diameter and deflects under the load such that there tends to be a concentration of pressure at the inner edge of the bearing near the piston rod, as illustrated by Fig. 22(a). With proper attention to design details, however, the pressure on the cro~sheadpin can be made cn.b
CVL. I
SKllON A-A
cn. 7
c
LOOKING AFT CYL.1 T
Fig. 23
Crouhead and connecting rod with guider on two sides [9]
Fig. 25
Built-up cranluhoft [I 21
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
MARINE ENGINEERING
293
LUBRICATING OIL PIPE MAIN BEARING UPPER SHELL MAIN BEARING LOWER S H E L L
OIL SUMP Fig. 26
Crankshaft bearing [8]
nearly uniform. This can be accomplished, as shown in Fig. 22(b), by supporting the crosshead pin with a flanged element which is flexible near the piston rod, which would otherwise be the region of hardest contact. The crosshead in Fig. 23 is symrnetrical with four guides. The four crosshead guides are anchored to the columns on opposite sides. The lower end of the piston rod penetrates and is secured to the cylindrical crosshead pin. Each end of the crosshead pin has a babbitt-lined slipper in way of the bearing area. Because of the danger of a crankcase explosion, sparks in the crankcase must be avoided. For this reason, all crosshead bearings, crosshead slides, crankpin bearings, and crankshaft bearings are always of the babbitt-lined type. An oil passage is drilled into the connecting rod to lubricate the crankpin bearing (see Fig. 23). Provisions are made for inserting shims between the connecting rod and the bottom-end bearing attachment. This permits adjustments to be made to the compression ratio. An alternative crosshead design, in which crosshead guides are on only one side of the crosshead, is shown in Fig. 24. 'A bracket carrying the telescopic tubes of the piston cooling system is attached to the crosshead (see also Fig. 18).
1.9 Crankshaft and Crankshaft Bearings. Crankshafts are usually of the built-up type, with the cranks made of steel castings and the shaft elements made of forged steel. The cranks are shrunk onto the shaft elements as illustrated by Fig. 25. Engines which have a high number of cylinders frequently have their crankshafts made in two sections, as shown in Fig. 25. To reduce the unbalanced centrifugal forces, metal is removed from the crank webs where possible and added to the opposite side so as to produce a counterweight effect. The crankshaft bearings are cast steel shells lined with babbitt. Oil is admitted at the top of the bearing and runs through oil grooves in the top half of the bearing shell to the horizontal split. Figure 26 shows the oil supply passage to the crankshaft bearings. The manner in which the bearing ~ h e l hare secured to the bedplate is also clearly shown in the figure, as are the tie bolts which hold the engine together and resist the firing forces. In some engine designs lubricating oil is supplied through the crankshaft; however, the bearing lubrication arrangement in Fig. 26 is the one employed with most engines because it does not require oil passages to be bored in the crankshaft.
1 air silencer 4 bearing 7 guide blade 2 blower casing 5 shaft 8 turbine wheel 3 blower wheel 6 turbine casing 9 insulation ' Fig. 27 Turbocharger assembly [8]
CONSTANT PRESSURE TURBINE
BLOWER
EXHAUS
CRANK A N G L E -
a
( A ) CONSTANT PRESSURE TURBOCHARGING W
PULSE TURBINE
BL,OWER
Z
zm
a
(6) PULSE TURBOCHARGING
Section 2 Engine Subsystems 2.1 Supercharging. Until about 1940 high cylinder outputs were obtained with double-acting engine designs. This type of cycle permitted short engine lengths and low weights. However, the double-acting engines
were too complicated and were not suitable for operation with heavy fuels. The cylinder output of a diesel engine can be increased effectively by supercharging. The cylinder mean effec-
Fig. 28
Constant-prarsure and puke turbodarging [9]
tive pressure is directly related to the quantity of fuel bulped which, in turn, is related to the quantity of combustion air (oxygen) in the combustion chamber. With the development of gas-turbine-driven air compressors (turbochargers) which operate with the diesel engine exhaust gases, in conjunction with large cylinder dimen-
sions, high cylinder outputs were obtained also with single-acting engines. Compared with two-stroke diesel engines equipped with scavenging pumps, supercharging has permitted the mean effective pressure of low-speed diesels to be almost doubled. Supercharging also enables the specific fuel consumption and specific weight to be reduced. A single-acting design means a simple engine, good reliability, a long service life, minimum supervision and maintenance, and simplified operation. Besides this, a single-acting engine is capable of burning heavy fuel oils. All of these points contribute to economy; therefore, low-speed diesel engines are usually twostroke, single-acting, and supercharged. Figure 27 is a turbocharger assembly that is typical of those employed with low-speed diesels. In addition to the scavenging air which is compressed in the turbocharger, other sources of scavenging air are often provided. Examples are piston scavenging pumps which are mechanically driven by the engine itself, the lower sides of the main pistons (which may be designed as scavenging pumps), and small electric blowers (see Section 1.3). These additional sources of scavenging air ensure that the engine will be operable even in the event of a turbocharger casualty, and during starting, maneuvering, and low outputs. The exhaust gas turbine can be applied to operate on either the constant-pressure principle or the pulse principle. As can be seen from Fig. 28, in a constant-pressure arrangement the exhaust gases from all cylinders are conducted to a receiver and from there to the gas turbine. Since the pressure in the receiver tends to be the average of the cylinder outputs, the gas turbine is provided with a gas supply of nearly constant pressure. With this ar-
.
MARINE ENGINEERING
PACKING
NON-RETURN SCAVENGE VALVE
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
that it lies in the general vicinity of 100 percent. This corresponds to a mep. of about 150 psi and a mep, of about 75 psi. The pressure of the scavenging air itself is approximately 15 psig. With air pressures this low, it is essential that resistailces to flow in the air passages be minimized in order to charge the cylinder with the largest possible quantity of air. The air pressure drop going into the cylinders is minimized & malung the ports as large as practicable and by extending them around a large arc of the liner circumference as illustrated by Fig. 14. In the case of uniflow-scavenging engines, the largest practicable exhaust valve area (see Fig. 16) is provided m a means of improving scavenging and supercharging efficiency. Turbochargers, mechanically driven air pumps, piston lower sides, and electric fans are used in various combinations and arrangements to provide the air required at full load and a t load. As discussed in Section 1.3 they may be aligned in series or in parallel and their alignment may shift depending upon engine load. The supercharging air is always cooled, with an air temperature leaving the cooler of 100 F being common. As shown by Fig. 10, air is often cooled in two stages to attain high scavenging efficiencies; additionally, thermal stresses in the combustion chamber are reduced bv cooling the superchawing air. "In some loop and croG-scavenging engines the exhaust ports, which are opened by the piston as it goes down during the power stroke, are opened before the scavenging ports (see Fig. 2); therefore, there is no problem with combustion gases tending to enter the scavenging air system. With supercharging, however, it is advantageous for the exhaust ports to be closed somewhat prior to the scavenging ports. For this purpose some engine designs are provided with a special exhaust valve of either the rotating or reciprocating type. This permits the pressure in the cylinder to reach approximately the stagnation pressure before the scavenge ports close. A disadvantage of this arrangement is that during the power stroke the scavenging ports can open before the exhaust ports and the combustion gases tend to backup into the scavenging air supply. However, this disadvantage can be resolved satisfactorily with nonreturn valves in the air supply, as shown in Fig. 29 (a detailed view of Fig. 10). 2.2 Cooling. The cooling of low-speed highly supercharged engines is a consideration of overriding importance. The cooling system must remove 20-30 percent of the fuel heat of combustion, which is a large quantity of heat with a large diesel engine. Engine cooling is important for several reasons. The strength of engine materials decreases with higher temperatures; therefore, cooling is necessary to regulate the material temperatures to a level that will ensure the material strength required. Additionally, the viscosity of lubricating oils is sensitive to temperature, and a t elevated temperatures the lubrication oil will break down with a resulting loss of lubricating properties; this re-
CYLINDER LINER
Fig. 29
Cylinder lacket and liner of a cross-scavenging engine
[I 1 1
rangement a supplementary source of scavenging air is required for starting and part loads; such an arrangement is illustrated by Fig. 9. With a pulse arrangement, the exhaust gases from each cylinder or group of cylinders are admitted directly to the gas turbine through a short exhaust pipe. As a result, the flow of exhaust gases to the turbine pulsates; the turbjne is designed such that it utilizes both the velocity and pressure energy in the exhaust gases. Additional sources of scavenging air are sometimes not required with a pulse design, but it is not unusual for mechanically driven air pumps to be installed as a precautionary measure recognizing the possibility of a blower casualty and, on occasions, to meet part-load air requirements. The degree of supercharging, in percent, can be assessed by the expression
A
=
(FEE mep.
where mep. = mean effective pressure of the engine when supercharged mep. = mean effective pressure of the engine with normal scavenging With the original supercharged engines, the degree of supercharging, A, was appro~irnately25 percent. But the degree of supercharging has subsequently risen such
I I
i i
Fig. 30 Temperature distributions in a cylinder liner and a water-cooled piaton [9]
quires that the lubricating oil temperature be carefully controlled. Not only temperatures but also their gradients are important, as nonuniform temperatures cause thermal distortion and stresses. As discussed in Section 3.1, efficient engine cooling is also important in inhibiting high-temperature corrosion and slagging which can occur during the combustion of heavy fuels. At the other end of the scale, excessively low temperatures can cause the formation of sulfuric acid. When the combustion gases are reduced below their dew point, the oxidized sulfur in the combustion gases can be hydrolyzed t o form sulfuric acid, which can be extremely corrosive
P71.
As noted in Section 2.1, cooling of the scavenging air is necessary for several reasons. One is that air of a tigh density is required for higher degrees of superchargng, . Fig. 31 Thermal expansion of water-cooled piaton rings [a] and another is that relatively cool scavenging air reduces the thermal stresses in the combustion chamber. The cooling medium used to control the temperatures flow, a higher coolant velocity, more heat transfer area, of the mechanical elements of low-speed diesel engines or a combination thereof. An additional factor in favor may be either water or lubricating oil, and arguments of water is that oil can carbonize a t high temperatures. can be made in favor of each in the regions of high temThe cooling of pistons is particularly critical. Due to perature. From a heabtransfer point of view, water is the elevated temperatures at which they operate, there is preferred. At the conditions which exist within diesel a hazard of high-temperature corrosion. The temperaengines, the maximum heabtransfer coefficient with ture distributions in a typical water-cooled cylinder liner water is approximately 500 Btu/hr-ftz-deg F, whereas and piston are shown in Fig. 30. (Due to the inherent that with oil is in the range of 300-370 Btu/hr-ftbdeg F ; heat-transfer advantages of water, water cooling can and'the specific heat of water is about 1.0 Btu/lb-deg F as provide piston crown temperatures that are 200 to 300 compared with oil of 0.45 Btu/lb-deg F. Therefore, to deg F lower than oil-cooled pistons.) High-temperature transfer the same quantity of heat a t the same tempera- catalytic oxidation, primarily caused by vanadium ture differences, an oil system must have more coolant pentoxide and sodium pyrosulfate in the fuel ashes, has
-
296
I
I
I
i t
I
11 1
I 2 3 4 5
MARINE ENGINEERING
air cooler turbocharger vent expansion tank freshwater pump Fig. 3 2
6
seawater-freshwater heat exchanger
7 seawater inlet (sea chest) 8 lubrication oil cooler 9 seawater fllter 10 seawater pump Engine cooling diagram
I lube oil valve 5 6 2 electrically driven pump 3 lube oil fllter 7 4 heat exchanger (see Fig. 32) Fig. 33 Lubricating oil
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
running gear lube oil tank cylinder lube oil distributor cylinder lube oil tank diagram
formation of sulfuric acid. The engine cooling water outlet is maintained at 140-160 F; higher temperatures burned as nluch as a half-inch of metal from piston crowns are unsuitable because of excessive combustion chamber [5]. The high operating temperature of the piston re- (and combustion gas) temperatures and because of the sults in considerable thermal expansion, as may be see11 reduction in lubricating oil viscosity at high temperafrom Fig. 31; this can create adverse conditions for the tures. piston rings. It is necessary that the difference between the cooling Tending to offset the inherently poorer heat-transfer water inlet and outlet temperatures be small and that the properties of oil is the advantage that absolute contain- temperatures be uniform so as to minimize thermal ment of the cooling medium is not essential with oil. stresses. Before starting large engines, they are usually Minor coolant leakage within the engine is not of conse- preheated with the warm cooling water from auxiliary quence with oil, whereas leakage of water within the diesels. Automatic control and monitoring of the coolingengine, or oil into the water system, must be avoided. medium temperature is desirable; see Section 3.3 for Therefore, with oil as the cooling medium, there is con- additional discussion on this subject. siderably more freedom in designing the piston coolant When a heavy fuel is used, the injection valve is circuit; see Section 1.5 for additional discussion along cooled with fresh water or diesel oil and the cooling system is maintained in a separate circuit. A separate these lines. The various engine bearings and the crosshead guides circuit is essential with a freshwater coolant to avoid are cooled by the lubricating oil. contaminating the entire engine freshwater circuit in the At the higher temperature levels, seawater has un- event of a leaking injection valve. 2.3 Lubrication. Different oils are used for lubricating satisfactory corrosive and hardness properties; consequently, it is only usable at the lower temperature levels the crankcase mechanisms than are used for lubricating (e.g., lube oil coolers, air coolers). Fresh water, with a the cylinder linen. Since a barrier is provided which low hardness and with a corrosion-preventive additive, isolates the cylinder from the crankcase (see Section 1.7) is used at the higher temperature levels. Cylinder liners, the use of two different lubricating oils is feasible. cylinder heads, and exhaust valves are always cooled with Lubrication of the crankcase mechanisms is relatively fresh water. The fresh-water coolant is cooled by means simple. Conventional lubricating oils of SAE-30 visof a seawater-freshwater heat exchanger as illustrated by cosity are normally employed. Fig. 32. The figure shows that the scavenging air and An engine lubricating oil diagram is shown in Fig. 33. lubrication oil are cooled directly by seawater (as a re- Electrically driven pumps circulate the oil from the sult of their low temperature levels). crankcase oil sump (Fig. 26) to the separator, where Typically, fresh water enters the engine at 120-130 F; foreign particles, water, and water-soluble acids are lower temperatures are not used due to the increased oil removed. It is then filtered and cooled, as indicated viscosity at lower temperatures and the hazard of reach- in Fig. 32, and from there goes to the bearings. Ading the combustion gas dew point with a consequent mission to the bearings is through the bearing shell (Fig.
26) or through holes in the crankshaft, connecting rods (Fig. 23), and similar elements. The lubricating oil pressure at the engine inlet is about 25 psi. The conditions under which the crosshead bearings operate are particularly adverse from a lubrication standpoint. The motion at crosshead bearings is oscillatory instead of continuous and, unless special provisions are made, crosshead bearings tend to operate without an oil film. In order to operate satisfactorily, some engine designs require a very high oil pressure at the crosshead bearing. To supply the required pressure, these engines employ special high-pressure lubricating pumps attached to the crosshead, such as illustrated in Fig. 24. The arrangement shown in Fig. 24 is designed such that the pdmp forces oil into the grooved contact area of the crosshead bearing at the instant the load is a minimum (at bottom dead center). When the load increases, the bearing is supported hydrostatically, and an oil film is maintained during the remainder of the cycle. The consumption of bearing lubricating oil is usually small, resulting from leakage losses and pumping through stuffing boxes. The quantity of oil circulated lies in the range of 2 4 lb/bhp and is circulated 7-14 times per hour. The lubrication of engine liners is a special problem due to the combustion residues and the deleterious components in heavy fuels. Special lubricating oils are required for this purpose [18]. The oils used generally are of SAE-50 viscosity. Cylinder lubrication oil is usually injected into the liner by special pumps driven by the engine itself; however, the pumps can be separately driven. Figure 34 illustrates an oil distribution scheme for cylinder liners. Good oil distribution is especially important in the upper part of the liner to avoid abrasion of the liner and piston rings. The oil delivered by the pump must be injected
PISTON VELOCITY.
Lubrication of a cylinder liner
when the piston is going up so that the piston rings will sweep the oil up into the highly stressed region of the liner. To prevent the cylinder pressure from forcing the lubricating oil back through the pump, small nonreturn valves, often simple ball check valves, are provided. Consumption of the liner lubricating oil is in the range of 0.0005-0.0017 lb/bhp-hr; therefore, the cost of liner lubricating oil is a significant operating expense. Combustion residues (see Section 3.1), together with the lubricating oil residues, collect on the diaphragm and are piped from there to the mud oil tank. 2.4 Fuel Injection. The fuel is injected directly into
y VELOCITY OF FUEL PUMP PISTON
FPy80
1
Fig. 3 4
PRESS.. PSI STROKE.
lNcHES 10000~
8000
-
VALVE OPEN
-10.
TDC
io
20
CRANK A N G L E . DEGREES Fig. 3 5
Fuel inlection diagram
30
,
MARINE ENGINEERING
/
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
CONTROL RACK
Fig. 37
I Fig. 36
Fuel iniection valve
the combustion chambers of low-speed diesel engines. Precombustion or turbulence chambers are not used; therefore, the fuel must be finely atomized. For the fuel to mix properly with the combustion air, there must also be a turbulent flow of air in the combustion chamber. The fuel injection pressure depends upon the viscosity of the fuel but lies in the range of 5,000-10,000 psi. The velocity of the fuel leaving the injection nozzle may reach 1100 fps. The viscosity of the fuel when injected is approximately 35-125 Redw. sec; fuels with higher viscosities must be preheated, as discussed in Section 3.1. The injection period depends upon the combustion qualities of the fuel and the speed of the engine. With low-speed engines, the injection period extends over 1530 dee of crank ande. The injection process begins aboutlo. deg before tip dead center; the iijection is 8;ch that combustion begins 5-0 deg of crank angle before top dead center. A typical fuel injection pressure d i e gram is shown in Fig. 35. The fuel valves are usually spring-loaded, as in Kg. 36; the needle opens when the fuel pressure is about one third of its maximum value, as indicated by Fig. 35. The nozzle generally has several holes of 0.01-0.04 in. diameter.
I fuel pump cam 5 fuel pump casing 2 cam follower 6 wction pipe 3 fuel leakage 7 pmsure pipe 4 spting 8 plunger Fig. 38 Cam-operated fuel pump [a]
Combustion chamber 191
The configuration of the combustion chamber and the arrangement of the injection valves, as regards the position of the nozzles in the cylinder, must allow the longest possible distance for the injected fuel droplets to travel and also ensure sdIieient air turbulence to obtain a homogeneous fuel-air mixture. Figure 37 shows a fuel nozzle arrangement. The fuel pumps have spring-loaded plungers that are usuallv driven by the camshaft; Fig. 38 illustrates a typical arrangement. In exceptional cases one fuel pump may serve several cylinders by using a receiver, but more often each cylinder has its own fuel pump. Exact metering of the fuel and an equal distribution to all cylinders throughout the operating range is a complex problem. This problem is further aggravated by the use of higher mean effective pressures because the difference between the fuel flows at idle and full load becomes larger. One solution to this problem has been the use of doubleplunger fuel pumps; only one plunger operates at low powers and both operate at higher powers, so that fine regulation is obtained throughout the operating range. The plunger stroke of the fuel pumps generally is maintained constant, and the quantity of fuel injected into the combustion chamber is regulated in either of three ways: (1) By opening the inlet valve, thus changing the beginning of injection. (2) By providing a bypass valve between the outlet and inlet of the fuel pump, thus changing the end of injection. (3) By a helical land on the fuel pump plunger which can be rotated in the barrel, thereby changing the length of the injection period.
I I
I 1
Fig. 39
Since marine diesel engines operate over a wide range of speeds, the time available for fuel combustion is variable; it is desirable to be able to change the time at which injection begins and/or ends. The plunger of a fuel injection pump often has a spiral control groove (or helix) and a control sleeve, as shown by Fig. 39 (a detailed view of the plunger in Fig. 38). By rotating the coxltrol sleeve in Fig. 39, the position of the control groove rel* tive to the fuel inlet port is changed thus altering the time at which the injection process ends, and therefore the quantity of fuel injected. For the type of pump shown, the beginning of the injection process can be altered by rotating the cam on the camshaft. The very high fuel injection pressures unavoidably result in some expansion of the pressurized fuel piping and compression of the fuel oil. This can cause a time delay in the pressure rise at the injection nozzle. To regulate the fuel flow satisfactorily and equalize the flow and timing between cylinders, it is desirable that all highpressure fuel pipe lengths be equal, as short as possible, and of heavy-wall construction. Regardless of the type of fuel used, the injection valve must be cooled. When only diesel fuel is used, the cooling medium is the fuel itself, but for heavy-fuel operation, a separate freshwater or diesel-oil cooling circuit is used. Further discussion along these lines is contained in Sections 2.2 and 3.1. 2.5
Starting, Reversing, and Control Arrangement.
Low-speed engines are started by means of compressed air which flows through the starting air valve to the engine cylinders. The starting speed must be high enough
Fuel iniection regulating mechanism
[a]
to produce a temperature at the end of the compression stroke sufficientlyhigh to ensure ignition of the injected fuel. The starting speed is 30 percent or more of the rated speed. The starting torque must be high enough to overcome the frictional losses in the engine and propeller shafting. Frictional losses are high when the engine is cold due to the increased viscosity of the lubricating oil, but the engine frictional losses can be reduced by preheating the main engine and its lubricating oil with warm water from an external source. Reversing the engine when the ship is ip motion is a particularly stringent design criterion in that sufficient starting power must be provided to overcome the hydrodynamic torque produced by the propeller. The st'arting air is compressed and stored in air bottles at a pressure of 400-600 psi. The starting air pressure delivered to the starting valve is at a pressure of about 400 psi. The air bank capacity is specified by the classification societies in the form of a requirement that a reversible engine be capable of at least twelve consecutive starts without recharging the air bank. The required air capacity consequently is dependent upon the number of cylinders, cylinder diameter, ' piston stroke, mep, service air pressure, and the like. If a hot engine is reversed or started after a brief shutdown, the cold starting air can subject the engine to considerable thermal stresses. A starting mechanism for a low-speed diesel is illus-
Il:i111
MARINE ENGINEERING
300
STARTI.NG AIR
DISTRIBUTING VALVES
-k
- TDC
TDC
CRANK ANGLE SHIFT FOR REVERSING
ASTERN 1 cam 2 reno pbton 3 starting air valve Fig. 4 0
4 5
vent starting air
Fig. 43
Fig. 41
Stating air mechanism
TDC
TDC
TDC
Admission of starting air
Several distributing valves actuated by one cam
TDC
TDC
8 sllding blbck 9 fuel regulating wheel 3 starting slide valve 10 slotted plate 4 fuel pump cam follower (ahd. & ast.) 11 stating air valve 5 stating lever 12 cylinder 6 slotted plate 13 air bottle 7 sen0 piston 14 air compressor Fig. 45 Starting and revening mechanbm scheme 1
reversing lever
2 camshaft
! fuel valve open 2 fuel valve shut 3 hg. 4 4
PISTON
POSITION^
1 starting air only, first revolution starting air only, s m d revolution ignition of fuel and starting air normal wmbulon without stating air 5 opening of the start1n.g air valve Fig. 42 Cylinder pressure when starting
2 3 4
trated in Fig. 40. When the engine piston is just beyond the top dead-center position, the distributor valve cam, driven by the engine itself, opens the distributing valve, thus opening the starting air valve pneumatically by means of a servo piston. The starting air valve can be opened only when the starting air pressure is higher than the pressure in the combustion chamber, otherwise there would be a backflow from the combustion chamber into the starting air pipe. In such an event fuel or lubricating oil residues could result in an explosion. It is not necessary that each cylinder have a starting air valve, but it is necessary that the engine be able to start from any crank angle. In the case of two-stroke engines, this means that at least three cylinders must be equipped with starting air valves. The distributing valves
of the various cylinders are actuated by means of a common cam as shown in Fig. 41. The pressure in the cylinder when starting the engine is shown in Fig. 42. During the first two revolutions shown, the engine rotates by means of compressed air only to the required starting speed. During the third revolution a limited quantity of fuel is injected into the combustion chamber and burned; the fuel quantity limitation is necessary to avoid excessively high pressures in the combustion chamber. During the fourth cycle shown, the starting air is switched off and the quantity of fuel injected is aa required for the running condition desired, and the starting process is completed. To reverse an engine, the distributor cam is turned through such an angle that starting air is admitted when
crank angle shift for reversing Fuel Inleetion cam shift for revening
the piston is on the proper side of the top dead-center (2) The starting air system must be interlocked to position to give the desired direction of shaft rotation. preclude manipulation of the starting mechanism Figure 43 illustrates the air adplission position for ahead when the engine is in the normal operating mode. and astern operation. Likewise, all cams which control (3) The quantity of fuel injected into the combusprocesses in the cylinder that have unsymmetrical angles tion chamber during the starting period must be of opening and closing with respect to the dead-center limited to prevent excessively high combustion positions are turned throukh a aimilar angle. The shift pressures (see Fig. 42); the fuel-regulating mecharequired of the fuel injection valve cam is illustrated by nism should have an override which permits the Fig. 44. quantity of injected fuel to be increased to faciliFor two-stroke engines in which the pistons control the tate starting under unfavorable conditions. scavenging and exhaust air ports (loop and cross Faveng(4) The starting mechanism and turning gear must ing) only the cams of the starting air and fuel pumps be interlocked to preclude starting the engine must be turned. In the case of uniflow two-stroke when the turning gear is engaged. engines, the cams actuating the exhaust valves must There are several practical manners in which the cams also be turned. can be shifted from the ahead to the astern positions. To avoid errors during starting and reversing that One possibility is an arrangement with a two-position m y result in engine damage, the following requirements are imposed upon the starting and reversing mechanism: coupling between the driving gear of the engine and the camshaft; the coupling would have end positions in the (1) The starting air system must be interlocked such ahead and astern directions. Another possibility is to prothat starting air can be admitted only when all vide two cams on the camshaft, one for ahead operation cams are in their end positions for ahead or astern and the other for astern operation; by moving the camoperation; admission of starting air must be pre- shaft axially, either the ahead or astern cams can be cluded when cams are in intermediate positions. placed in contact with the cam followers.
111
302
Fig. 4 6
MARINE ENGINEERING
Axid-movement camshaft with cams for ahead and astern operation [8J
A simple starting and reversing mechanism is shown in Fig. 45. For starting, the reversing lever 1 is placed in either the ahead or astern position, thus moving the camshaft in the axial direction. This places the rollers of the starting slide valves 3 and of the fuel pumps 4 in contact
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
with the proper cams. The starting lever 6, which was previously blocked in the stop position by the plate 6 , now can be moved through the slots in the plate to the starting position; this opens the starting air servo piston 7 by means of rods 8. The engine rotates, and upon reaching the starting speed a limited quantity of fuel is injected by turning the fuel regulating wheel 9; the quantity of fuel injected is limited by the plate 10 in contact with the rods 8. After completing the starting procedure, the starting lever 5 is moved to the operating position, thus blocking the reversing lever 1 and a t the same time freeing the fuel regulating wheel 9. Fuel now can be injected as required to attain the desired engine output. Both hydraulic and pneumatic servomotors have been used for moving the camshaft axially. A mechanism for shifting the camshaft is illustrated by Fig. 46; the camshaft is gear driven from the crankshaft as shown in Fig. 47. Another possibility for reversing the fuel injection process is shown by Fig. 48. The camshaft has different cams for ahead and astern operation. The fuel injection pump has a roller for each cam, and by turning the reversing shaft the desired roller is placed in contact with its corresponding cam. I t is also possible to change the distance between the roller and cam by turning the
VERSING
AH~AD CAM FOLLOWER Fig.
48 Fixed-position camahaft with cams for ahead and astern operation
shaft; this means that the piston stroke of the fuel injection pumps and consequently the quantity of injected fuel can be adjusted. A mechanism in which the same cam is used for both ahead and astern operation is shown in Fig. 49. The camshaft is directly driven by the crankshaft through a two-position hydraulic coupling. The coupling permits an angle of displacement asrequired for ahead and astern operation. The cam shown in Fig. 49 serves two fuel injection pumps. The reversing performance of a 65,000-dwt tanker is shown by Fig. 50. The first step when reversing is to internipt the flow of fuel to the engine. The ship begins a gradual decrease in speed, but the engine rpm abruptly drops until the propeller takes charge of the engine and causes it to continue rotating in the ahead direction at about 10-15 percent of the initial rpm. As the ship speed decreases, the hydrodynamic torque developed by r o ~ e l ldecreases er until the reversing speed the ~ - is reached. r The reversing speed is the speed a t which the torque that can be developed by admitting starting air to the engine in the reverse direction is sufficient to stop the engine (and propeller) and reverse it. Once initiated, the actual reversing process is accomplished rapidly, as can be seen from Fig. 50. When accomplished as illustrated by Fig. 50, the reversing maneuver is conservative. In the event of an emergency, starting air can be admitted to the r
AHEAD -ROLLER Fig. 49
One cam for ahead and ostem operation 1 1 1 ]
-
Fig. 5 0
65,000-dwt tanker engine reversing and ship stopping performance
engine in the reverse direction sooner, before the engine can actually take charge of the propeller, as a meam of braking %he ship and decreasing the time required to reach the reversing speed.
Section 3 Overall Considerations
Fig. 47
Camshaft drive gear
[8]
3.1 Heavy Fuel Operation. Low-speed marine diesel engines, in general, burn heavy petroleum fuels that are not only extremely viscous but also contain large quantities of impurities. The constituents, properties at injection, combustion, and combustion residues formed are
considerably different with heavy fuels as compared with standard diesel fuels. As a result, when heavy fuel is used, special arrangements for re paring the fuel before injection as well as special design features of the engine itself are required. In addition, special lubricating oils
304
MARINE ENGINEERING
are required for the combustion chamber when heavy fuels are used. Major differences between heavy fuels and diesel fuels are as follows: (1) The specsc gravity of heavy fuels is higher, in the range of 0.9404.995 as compared with diesel fuels at 0.824.86. Due to the higher specific gravity, the separation of water from heavy fuels is considerably more difficult. For this reason the filling of fuel tanks with ballast water is often discouraged because upon refilling with fuel, the residual water mixes with the fuel. (2) Due to their higher viscosity (200-4000 Redw. sec at 100 F), heavy fuels must be heated to 210-250 F to reach the low viscosity (35-125. Redw. sec) required at the injection valve. (3) Heating of heavy fuels is also necessary to maintain the fuel at the proper viscosity at the injector. Heavy fuels can be used when maneuvering, but the fuel lines to the injection valve also must be heated. For this reason, and also in recognition of the desirability of baving diesel fuel in the lines upon shutdown, sometimes diesel fuels are used when maneuvering and the use of heavy fuels is confined to continuous operation at sea. (4) Due to the longer time required for the combustion of heavy fuels, the injection times for heavy fuel and diesel fuel are different. Heavy fuels must be injected earlier than diesel fuels. (5) Heavy fuels contain large quantities of tarry substances which must be separated and removed from the fuel because the tarry substances together with the lubricating oil residues form gummy deposits which have a deleterious effect on piston rings, inlet and exhaust ports, and other parts. These materials can be separated from the fuel by centrifuging, but it must be done when the fuel is at a moderate temperature. When the fuel is heated above approximately~210F, the tarry materials dissolve and cannot be separated. Therefore, the heavy fuel is heated to a maximum of 185-210 F before centrifuging. As an example, during a 24-hr period up to 880 lb of tarry materials and matter insoluble in normal pentane, 615 lb of sulfur, and 13 lb of incombustible matter may be removed from the fuel to a cylinder developing 2100 hp [19]. (6) Heavy fuels have an ash content (0.03-0.5 percent mineral ashes) which must be removed insofar as practicable due to its erosive effect in the cylinder. The fuel ash content is removed during the centrifuging process. (7) Heavy fuels may contain up to 5 percent sulfur. Due to the high sulfur content, a significant quantity of acid combustion residues passes down between the piston and liner. These acid residues must be prevented from entering the crankcase, otherwise the lubricating oil would be contaminated and consequently there would be a corrosion hazard. Therefore a diaphragm, with a stuffingbox for the piston rod (see Section 1.6), is required between the combustion chamber and crankcase. (8) Due to the higher sulfur content in the fuel, there
is an increased hazard of corrosion by sulfuric acid in the low-gas-temperature regions. The engine cooling s y 5 tem must be designed such that the temperatures of materials which are contiguous with the combustion gases are maintained above the dewpoint of the combustion gases so that sulfuric acid will not form. (9) Heavy fuels contain a number of impurities which can have a vitiating effect on the life of operating parts. Oxides and sulfates of vanadium, sodium, potassium, and zinc are among the most harmful in this respect, and all have melting points (slagging temperatures) in the range of 1100 F to 1600 F. Unfortunately, several of these combine with each other and with the ingredients of the structural metals to form eutectic mixtures with lower melting points, generally in the range of 1050 F, but in several cases, as low as 930 F. The latter temperature is associated with eutectic mixtures of sodium sulfate and vanadium pentoxide which are particularly to be avoided. Contamination (slagging) and corrosion often accompany each other but are not necessarily present at the same time. Generally, the two mechanisms are interrelated in that accelerated corrosion results from the contaminant combining with and removing the film of oxidation which normally protects the metal. Therefore, in the high-temperature region (piston crown, upper part of the liner, and cylinder head) an effective cooling system is especially important with heavy fuels. (10) Due to the combustion residues formed when using a heavy fuel, special abrasion-resistant materials must be selected for the cylinder liner, piston rings, and other moving parts. (11) The specific fuel consumption when burning a heavy fuel is somewhat greater than when a diesel fuel is used for the following reasons: (a), The lower heating value of heavy fuels is in the
range of 17,100-17,500 Btu/lb, while that for diesel fuel is 18,100-18,400 Btu/lb. For this reason, the specific fuel consumption with heavy fuels is increased 2-5 percent. (b) The weight of fuel lost during the separation process depends, of course, on the quality of the fuel. In the extreme case cited previously, by weight about 7 percent of the fuel would be lost during separation. However, a nominal value of 1 percent is considered more representative of the heavy fuels normally burned in slow-speed diesels. (12) Most heavy fuels contain highly volatile components and tend to form gases when heated; therefore, heaters must be placed on the upstream pressure side of the fuel line to prevent the formation of gas in the fuel pump suction and heaters. Additionally, electric heaters should not be used since cracking invariably takes place due to the high localized temperatures which form gas and coke in the heaters. A process for heavy-fuel preparation is shown in Fig. 51. After the heavy fuel is heated in the main fuel tank
FUEL TANK
m
GAS
FILTER FEED PUMP
PREHEATER SEPARATORS
HEAVY FUEL TANK DIESEL FUEL TANK (FOR MANEUVERING)
FEED PUMP PREHEATER'
0 @=
Fig. 52
Gas Inleetion for a dual-fuel engine 191
FUEL DOUBLE- FILTER VlSCOSlMETER
w
3.2 Gaseous Fuel Operation. In the case of liquefied petroleum gas tankers that are driven by diesels, the cargo boil-off can be burned in the diesel engines in concert with fuel oil. Since the gaseous boil-off would be lost to the atmosphere if not reliquefied or burned, the gaseous fuel burned represents a direct savings. ThereFig. 51 Heavy fuel oil treatment schematic fore, low-speed diesel engines in liquefied petroleum gas tankers are sometimes adapted for "dual-fuel" operation, that is, the combustion of both liquid and gaseous fuels. sufficiently to enable it to be pumped, it is pumped Slow-speed diesels can be adapted to burn gaseous through filters to two settling tanks, connected in fuels by mounting a gas injection valve in the cylinder parallel, each having a capacity for 24-hr operation. In head as illustrated by Fig. 52. The gaseous fuel is under these tanks the fuel is heated again to precipitate some a pressure of approximately 55 psi and is admitted by a residues. After being allowed to settle, the fuel is re- hydraulically actuated gas valve during the later phase heated and pumped to the separators. The reheat tem- of the scavenging process. The gas is directed towards perature must be maintained sufficiently low so that the the rising flow of scavenging air, thus ensbring a good tarry constituents will remain in solid form and be re- mixture between the gas and air. moved from the fuel by the separator. In the first The major difficulty involved in the combustion of separator stage (purifier) a small quantity of fresh water gaseous fuels is the problem of knocking. A mean effecis added to remove the ash content and water-soluble tive pressure of about 115 psi can be developed with a acids. No water is introduced in the second stage gaseous fuel containing not more than 10 percent methane (clarifier). The fuel is often reheated between the two or 5 pwcent propane. A higher concentration of separator stages. methane or propane requires a derating of the engine. Two separators are sometimes installed; this provides If, however, more than 10-percent carbon dioxide or for continuous availability of a clean separator, as one nitrogen is mixed with the fuel, uprating is possible. can be operated while the other is cleaned. Additionally, Cooling of the gaseous fuel charge also can be accomtwo separators are advantageous in that both may be plished to advantage. used simultaneously when one is heavily loaded and not Ignition of the gaseous fuel is accomplished by a spray performing to the desired standard. of pilot fuel which accounts for about 5 percent of the After leaving the separator, the fuel is put in either of total heat input. If larger quantities of pilot fuel are two day tanks that are connected in parallel. The fuel used, higher mean effective pressures are attainable. oil ,service pump delivers the oil from the day tanks An automatic switch-over gear is usually provided through a preheater, filter, and viscosimeter to the fuel which enables the engine to be changed from a gaseous injection pump. The line from the last heater to the to a liquid fuel without interrupting the operation of the injection valve is heated by steam. engine. Any proportion of gaseous to liquid fuels can FUEL INJECTION PUMP
306
LOW-SPEED DIRECT-COUPLED DlFSEL ENGINES
MARINE ENGINEERING
. total fuel consumr>be used. At .the full-load ~ o i n tthe tion is approximately the same with either gaseous or diesel fuels. 3.3 Accessories. A great deal of pressure, temperature, and speed data must be talcell in order to ascertain that the various processes within an engine are being performed properly. Audible and visual warning devices also are used to identify measured data that are not within prescribed acceptable limits. Depending upon the number of cylinders, between 60 and 100 bits of data aM recorded. Data which are typically monitored at the control station are as follows: Direction of engine rotation (ahead or astern) Engine rpm.(ahead or astern) Turbocharger rpm Pressures: Starting air Scavenging pressure to cylinder Lubricating oil to and from cooler Fresh water to and from cooler Fresh water for piston cooling Fresh water for cylinder cooling Fresh water for injection valve cooling Temperatures : ~ x h a u s gases t from each cylinder Exhaust gases to and from turbocharger Exhaust gas from exhaust gas boiler Lubrication oil t o and from cooler Fresh water to and from cooler Seawater Heavy fuel (also the viscosity)
ACCELERATION TIMING
OIL ASTERN
6 PROGRAM
VALVES
Fig. 54
Fig. 53
[a]
preloaders which permit studs to be tightened accurately without imposing torsional loads on the studs; this is a desirable capability in connection with cylinder head studs, frame tie rods, and studs in crossheads, main bearings, and the like. The principle upon which the hydraulic wrench for tightening nuts operates is illustrated in Fig. 53. Oil pressure is applied simultaneously to several of the devices that are fitted over permanent nuts to preload the studs; the permanent nuts then can be tightened by hand by inserting a rod into the holes bored in the bases of the nuts. ~ ~application ~ oft a special h tool ~is shown ~ in ~ i 54. Removal of the lower half of a main bearing shell is facilitated by creating a hydrostatic oil film between the bearing shell and the supporting structure. Upon turning the crankshaft, the higher frictional force between the shell and the crankshaft causes the shell to turn with the shaft to the point that it can be easily removed. Special tools are also advantageous in the removal of cylinder liners and in a number of other applications. 3.4 Remote Control and Automation. Remote contrO1 of a propulsion plant either the bridge Or from a special control room and the automation of propubion plant were first in
Hydraulic tool for prelooding a stud '[a]
tion of fuel if the lubrication oil or cooling water pressures should fall below acceptable limits. Occasionally dampers are provided to minimize the torsional vibration in the shafting system; torsional vibration dampers are often mounted a t the forward end These same parameters and others also can be monitored of the crankshaft. Engines are equipped with a turning gear which endirectly at the engine. Mercury thermometers and ables the crankshaft to be rotated slowly. The turning thermocouples are provided for taking temperature me* part operates through the flywheel. surements, and a recording tachometer is provided which A large complement of onboard repair parts are repermits an exact measurement of the engine revolutions. The fuel consumption is measured by means of a fuel-oil quired for diesel engines. Some of the large repair meter and can be checked by gage glasses on the day parts required are as follows : tanks. Each cylinder is equipped with an indicating inpistons and piston rings strument to determine the indicated p-u diagram for the cylinder liners cylinder. cylinder heads Speed regulation of the more recently built slow-speed stuffing boxes for piston rods diesels is accomplished by a hydraulic governor. Regfuel injection valves ulation of the engine speed is accomplished by altering starting valves the output of the fuel injection pumps. The accuracy safety valves of regulation is within limits of about a 1 percent of the exhaust valves for uniflow engines ordered speed. In addition to a speed regulator, an overspeed trip is provided which interrupts the fuel injection process in the event that the normal speed This list 'of large and heavy parts is only representative; regulator is not capable of maintaining the engine rpm furthermore, there are numerous smaller items. The large dimensions and heavy weights of many of within tlie prescribed limits. Circumstances which the low-speed engine parts require special tools. These could cause the overspeed trip to actuate would be very bad weather during which the engine rpm would tools make it possible for the small number of people surge periodically due to the propeIIer partially coming aboard ship to properly execute with ease many mainteout of the water, the loas of a propeller which would nance tasks that otherwise would be onerous. Special cause the engine to suddenly race, and analogous situ* tools effectively increase the availability of the engine. Special tools typically provided include hydraulic stud tions. Provisions are also made to interrupt the injec-
Hydraulic removal of main bearing lower shell
I
c
1
mation of engine operations are advantageous from several points of view. First, but not necessarily the most important, a reduction in routine supervisory labor is permittted; particularly the night-service labor. Engine operation is supervised at the bridge in a convenient manner the monitoring The engine need be manned a t night in bad weather or close waters for safety reasons only (so that the engine can be switched to manual operation if need be) and when maneuvering (this includes preparing the engine for operation and shutting the engine down). An additional advantage is that engine maloperation is automatically sensed and alarmed. Furthermore, in some cases remedial measures are taken automatically;
.
r - - - -,-,----
DIESEL
~ Fig..
55
ENGINE
Cantrol system rchqmatic diagrom for a direct-revehing d l h l engine
as an example, in the event of failure of an auxiliary such as a lubricating pump, the pump is started and put on the automatically. Thus, automatic provides an additional degree of engine safety in that prolonged engine msloperation is avoided, The remote control and automation principles in the case of low-speed engines are generally the same as those with other typesof propulsion plants. But the automation of low-speed engines * important as these engines are of larger power with corremeans tha% the engine's temperatures and temperature gradients are high. with thermal stressw being a matter of grave concern, it is advantageous to be able to automatically monitor the temperatures throughout the engine. A schemstic diagram of a systemfor a directreversing diesel engine is shown by ~ following i 55. ~ When . seoperating an engine automstically, the quence is followed: Starting the engine "ahead" or " a s h " (1) Check that the fuel admission is set to zero. (2) Move the starting air lever to "ahead" or "astern"; this positions the camshaft so that
LOW-SPEED DIRECT-COUPLED DIESEL ENGINES
MARINE ENGINEERING
Fig. 56
Low-speed dieael installation
The following alarms are usually provided starting air is admitted to the "ahead" or mated control system: "astern" ~ o r t of s the distributor. (3) Check that ingine rotation is"aheadJ' or "astern." Main engine cooling 8 8tC.m (4) Wait for the engine to reach the adjustable firing , Pump suction and dehery pressures speed. Cooler inlet and outlet temperatures Cooling water tank level gage Raise the fuel admission linkage to the starting (5) Reserve cooling water tank level gage position. Outlet temperature of each component cooled (6) After a time delay, cut off the starting air. Main engine h W i n g oil Pump suction and delivery pressures (7) Check to determine that the engine speed is above Cooler inlet and outlet temperatures or below the firing speed. Sump tank level gauge Pressures to and from filters (8) If the rpm is above the firing speed, raise the fuel Pressure at turbochargers admission linkage to correspond with the conTemperature at each main bearing troller setting. Pressure at main bearing inlet mainfold Cylinder lubricating oil pressure (9) If the engine rpm is below the firing speed, repeat Temperature of propeller shaft bearings actions (1)-(7). When the engine has failed to Pressure of oil to reduction g e m fire three times, the starting sequence is disTemperature of reduction gear bearings Temperature of thrust bearing continued and the control system gives an alarm. Main engine exhauat goe Temperature of gas leaving cylinder Changing engine revolutions Temperature of ges to and from blowers (10) Move the fuel admission linkage to correspond Main engine pre8eure charging 8y8tem with the new telegraph setting. Temperature of air at inlet Air pressure leaving filters Reversing the engine Temperature of air leaving blowers (11) Move the fuel admission lever to zero. Temperature of air leaving coolers (12) Wait until the revolutions have fallen to the Turbocharger speed Temperature inside scavenge belt firing speed. Main engine fuel oil Perform normal starting actions (1)-(9). (13) Oil pressure to and from high-pressure fuel If the engine has a barred speed range (i.e., a vibratory 0ii;:f::ure to and from heaters resonant frequency within the operating range), the fuel Main engine starti9 air lever setting is slowly increased until the engine speed is Pressure in each an bottle Pressure in starting manifold on engine above the ba&?l _speedrange.
,
_
- ,
in an autoAhwn Low High Low Low High Low High
LO^
High Low Low High Low High High High High High Low High High Low High High, Low High, Low Low Low
For additional discussion regarding the automation of diesel propulsion plants, see Section 2.4 of Chapter 21. 3.5 Installation Aboard Ship. Low-speed diesel engines are not rigid either in bending or in torsion; for this reason, it is not possible to place vibration isolation material between the engine and the hull foundation. A rigid hull foundation, with a high resistance to vertical, athwartship, and fore-and-aft deflections, is required. The engine is bolted to the hull foundation with fitted bolts used a t one end of the engine (near the thrust bearing) and clearance-fit bolts towards the other end. Chocks are accurately machined and fitted between the engine base and the hull foundation so as to uniformly support the engine and avoid imposing stresses on the .engine frame. The design of hull foundations does not lend itself to an exact analysis; instead, it is influenced ' greatly by previous successful experience. The engine room must be designed such that there is sufficient overhaul space above the engine for the removal of cylinder heads, pistons with piston rods, and cylinder liners. Furthermore, these and other replacement parts which are both large and heavy must be lifted by cranes. Consequently, the engine room must be designed to permit the replacement parts to be lifted from their storage area, transported, and lowered to the engine by crane. An engine room skylight, or similar opening, also should be provided to transport replacement parts to and from the ship. Due to the low-frequency noise generated by lowspeed engines, the operating platform can be located a t the engine itself. But special control rooms are often preferred as the noise level in the control room can be made to be approximately 30 db less than the 100 db in the engine room. The auxiliary equipment is arranged in groups to facilitate their control and surveillance. For example, the lubricating oil equipment, including pumps, filters, coolers, separators, and their fittings, are grouped together. Similarly the cooling water, heavy fuel preparation, and electric plant equipment are arranged in groups. A low-speed diesel engine with a rating of 18,000 shp a t 118 rpm as installed in a container ship is shown in Fig. 56. Electric power is produced by the generator mounted directly on the line shafting. Operation of the entire plant is automatic, and it is remotely controlled from the bridge. The engine room is completely unattended for 16 hours of the day.
309
References
1 K. Illies, Schiffsbetriebstechnik, yieweg-Verlag Braunschweig, 1969. 2 K. Illies, "Neueste Entwicldungen im Schiffsmaschinenbau," Jahrbuch des Schflahrtswesens, 1968. 3 S. Bock and G. Mau, Die Dieselmaschine im Landund Schiffsbetrieb, Friedrich Vieweg u. Sohn, Braunschweig, 1968. 4 W. Henschke; Schijj'bautechnisches Handbuch, VEB Verlag Technik, Berlin, 1958. 5 F. Mayr, Ortsfeste und Schiffsdieselmotoren, Springer-Verlag, Berlin, 1948. 6 F. Sass, Bau und Betrieb von Dieselmaschinen, Springer-Verlag, Berlin, 1948. 7 F. A. F. Schmidt, Verbrennungskraftmaschinen, Verlag R: Oldenbourg, Munchen, 1951. 8 Illustrations through the courtesy of MAN. 9 Illustrations through the courtesy of Sulzer. 10 ~llustrationsthrough the courtesy of Doxford. 11 Illustrations through the courtesy of Fiat. 12 Illustrations through the courtesy of Burmeister and Wain. 13 T. W. D. Abell and J. F. Butler, "The Future of the Large Direct-Coupled Diesel Engine," SNAME Spring Meeting, 1966. 14 J. A. Smit, "The Future of Diesel Propulsion," SNAME Spring Meeting, 1966. 15 H. Andresen, "Slow-Running Marine Diesel Plants," SNAME Spring Meeting, 1966. 16 E. A. van der Molen and Ir. H. van der Wal, "Air Consumption Data and Practical Performance Data of the Stork Uniflow-Scavenged Two-Stroke Marine Engine," Trans. IME, 1966. 17 K. Knaack, "Taupunktverhalten von Verbrennungsgasen mit hoheren Luftuberschusszahlen," Schiff und Hafen, 1968. 18 M. J. van der Zijden and A. A. Kelly, "Combating Cylinder Wear and Fouling in Large Low-Speed Engines," Trans. IME, 1956. 19 Hugo H. Scobel and Jochen Richter, "A New A p proach to Maintenance and Operation of Large-Bore, Two-Stroke Diesel Engines, and Experience in Operation of Periodically Unattended Engine Rooms," SNAME Diamond Jubilee (Spring) Meeting, 1968. 20 D. Gray, Centralized and Automatic Controls in Ships, Pergamon Press, 1966.
r(
CHAPTER I X
semar
I
I Reduction Gears
Section 1 Introduction 1.1 Early History. It is generally acknowledged that Dr. DeLaval was the first to apply a reduction gear for ship propulsion with a 15-hp experimental unit in 1892. This was followed by Sir Charles Parsons with .his 10-hp experimental geared-turbine unit in 1897. However, these were experimental units and functional marine reduction gears did not make their debut until some years later. Just after the turn of the century, the steam turbine was being championed for ship propulsion by Sir Charles Parsons and others. I n a study in 1904 of the prob* bility of the steam turbine becoming a successor to the reciprocating engine, Admiral George W. Melville and Mr. John H. MacAlpine, consulting engineers, reported: "If one could devise a means of reconciling, in a practical manner, the necessary high speed of revolution of the turbine with the comparatively low rate of revolution required by an efficient propeller, the problem would be solved, and the turbine would practically wipe out the reciprocating engine for the propulsion of ships. The solution of this problem would be a stroke of great genius." Parsons in 1909 said, "The solution may be found in reverting to some description of gearing. . . and if a satisfactory solution can be found, then the field of the turbine at sea will be further extended." Parsons carried on further research and experimental work in applying the helical gear to large-scale marine installations, and in 190!9-1910 he equipped the Vespaaian with a geared-turbine plant. The gear was rated a t 1095 hp and reduced the turbine speed of 1450 rpm to a propeller speed of 73 rpm. George Westinghouse, in 1909, demonstrated in a shop test a 6000-hp gear which reduced the speed from a 1500-rpm turbine to a 300-rpm hydraulic dynamometer. This gear was the forerunner of the 6500-hp gears installed in the collier Neptune a short time later. The adoption of high-speed helical reduction gears in connection with marine propulsion was rapidly accepted by engineers all over the world, and this type of equipment had a very rapid development. At the end of 1910 the total power of geared marine turbines was about 15,000 shp, whereas 30 years later marine propulsion of this type in service totaled over 100,000,000 shp. It is interesting to note that the last large ship built with direct-connected turbines was the passenger liner Ile-&-France, which went into service in 1927. The
IZede-France had a propulsion plant of 52,000 shp divided arriong four screws. The turbines were designed by Parsons and were of the reaction type. The main turbines contained a total of more than 800,000 blades and weighed 1065 tons. With the further development of the steam turbine, still higher turbine speeds could be used to advantage and the single-reduction gear no longer met the need. Engineers began development of the double-reduction gear, where practically no limits were imposed on the speed ratio that could be obtained. This permitted both the turbines and propellers to be operated a t speeds suitable for their individual maximum efficiencies. Double-reduction gearing was first used about 1917. For a few years during and after the first world war, many ships were equipped with this type of gearing. Due to many unknown factors entering into the design and use of this new type of reduction gear and also due to the unusual operating conditions during the first years of use, considerable difficulties were experienced and many casualties of reduction gears occurred, which more or less slowed the general adoption of doublereduction gearing. However, by adhering to sound design principles, it was possible to eliminate early mistakes and develop satisfactory double-reduction gears and to greatly increase the application of this type of power transmission. This is not to suggest that the development of double-reduction gears immediately made single-reduction gears obsolete. For higherpowered naval ships with propeller speeds above about 200 rpm, single-reduction gears remained in general use until the early 1930's. Then, the higher rotational speeds of the more modern steam turbines brought about the demise of single-reduction gearing for turbine drives in all categories. Single-reduction drives still remain the standard, however, , for high- and intermediate-speed diesel engine service. The development of propulsion gearing has been one of a continuous improvement and refinement in materials and in manufacturing techniques and equipment to provide greater reliability and longer life. The horsepower ratings of gears have increased to keep pace with the requirements for larger and faster ships. There are only a few step-advances that can be identified, the step from single to double reduction, the introduction of welding to the construction of gear wheels and casings,
I
t
and the introduction of higher hardness pinion and gear materials with the attendant higher gear *toothloadings. The reliability, high efficiency, and long life of the modern reduction gear is well known and its low noise level makes it completely acceptable in the engine room. These factors have been in large part responsible for the continuing popularity of the geared-turbine drive for ships. 1.2 Articulation and Gear Arrangement. The early reduction gear designs incorporated many devices to minimize the effects of bending and torsion of the pinion and of inaccuracies in machining and alignment. However, experience has demonstrated that such devices are unnecessary, and gear elements are so proportioned and machined that uniform tooth pressures are obtained without the use of mechanical devices to compensate for pinion deflections. Figure l(a) represents the simplest arrangement of a marine reduction gear, i.e., one pinion meshing with a gear as used, for instance, for connecting a propeller to a diesel engine or to an electric motor. It is not used for propelling equipment with a turbine drive but, on the other hand, it has found a wide application for turbine driven auxiliary equipment on board ship such as generators and circulating pumps. Figure l(b) is a drive with two pinions as used frequently with diesel engines of comparatively large power. It is not used for direct connection to high-speed turbines, but is often used in the second reduction gear unit of a turbine drive using doublereduction gears. Figure l(c) represents the early type of single-reduction gear for a turbine drive, the principal difference between this reduction gear and the one shown in Fig. l(b) being in the number of pinion bearings. The third bearing located between the two helices is necessary because of the wide tooth face in relation to the diameter of the pinion. While many ships with reduction gears built according to Fig. l(c) are still in successful oper* tion, this design must a t the present time be considered obsolete. It was used for speed ratios up to or slightly above 20 to 1. Figure l(d) is the usual arrangement of a double reduction gear for turbinedriven ships. The two input pinions are driven by the two elements (high-pressure and low-pressure turbines) of a cross-compound turbine. Power is divided between the two input pinions by the turbine characteristics and is normally split approximately equally between the two turbines. Note that the second reduction gear is common to both highpressure! and low-pressure trains but that, although it transmits the power from both turbines to the gear shaft, the tooth portion is designed to transmit the power from one turbine. The terms "tandem" and "articulated" are also applied to this arrangement; tandem because of the disposition of the first and second reductions, and articulated because a flexible coupling is generally provided between the first reduction or primary gear wheel and the second reduction or secondary pinion.
31 1
GEARS
Figure l(e) represents the "nested type" doublereduction gear, which has also been used with crosscompound turbines. The configuration shown has the second reduction helices divided to provide space for the first reduction and is additionally referred to as a "split secondary." The nested type may also be arranged as a "split primary. " Figure 10 illustrates the type of gear referred to as a locked-train double-reduction gear. In it the power of the single input pinion is equally divided between the two intermediate-speed elements. Its advantage is that the gear elements are proportioned for one half of the input horsepower and are therefore smaller than would be the case with a single intermediate element. The overall size and weight are reduced, but offsetting this advantage is the added number of parts, the need to provide torsionally flexible shafts between the first and second reductions, and the need to "time" the assembly to equaliie the power split between the two trains. The term "dual tandem" is also applied to this type of gear. Figure l(g) is a locked-train type of double-reduction gear for a cross-compound turbine. This arrangement has become standard for high-powered naval ships and is coming into use for higher-powered merchant ships because it minimiaes the total weight and the size of the assembly. Figure l(h) is a planetary gear. It has a single input "sun pinion" which drives three or more "planet gears." These planet gears are mounted on a planet carrier which is solidly connected to the output coupling. The outer "ring gear" is held stationary in the gear housing. This type of gear has been applied to turbine-generator drive gears and to main turbine drive first reductions. It has also been considered for the second reduction of main reduction gears. Many other reduction gear arrangements are possible and have been used. These can be very special as in cases where more than one type prime mover is coupled to the propeller. 1.3 Methods of Manufacture. Nearly all gears produced in the U. S. have their teeth cut by the hobbing process. In this process the cutting tool is a hob, a rotary cutter having one or more leads, whose teeth are accurately formed to the "basic rack" tooth form selected. I n the hobbing process the teeth are cut and the true involute form of the tooth flanh is generated by the continuous rotation of the hob and the gear blank. The hob determines the dimensions of the teeth in the plane normal to the teeth. The other factors determining the tooth geometry, number of teeth, and helix angle are obtained by selecting change gear ratios for the hobbing machine; the selection of the change gear ratio provides a choice of these variables without a change of tooling. By adjusting the helix angle (which affects the tooth profile in the plane of rotation), it is possible to use a given hob (which dictates the tooth profile in the plane normal to the teeth) and produce a favorable number of teeth within rather broad limits.
ef
KEDUCTION GEARS
MARINE ENGINEERING
(a) Single reduction, single input
(f) Double reduction, single input, locked train
For this reason, manufacturers standardize with a small number of hobs. The other cutting process which has been used in the U. 5.. and is still used abroad for large gears is shaping. I n this process the shaping cutter is either in the form of a basic rack section or a small gear, stroking in timed relation to the rotation of the blank to generate the tooth form. Post-cutting processes are generally applied to further refine the accuracy and surface finish of the gear teeth. I n the shaving process, which is the most popular in the U. S., a multitooth cutter in the form of a small gear is pressed tight in mesh with the gear being shaved. As the gear is rotated rapidly and the shaving cutter fed slowly across the gear face, a very light cut is taken from the tooth flanks. This results in a finer tooth surface and a more precise involute form than can be produced by hobbing. The shaving process also makes possible the correction of slight mismatch in the helix angle of the gear and pinion by selectively shaving that portion
of the face width which indicates the heaviest tooth contact. In the lapping process, the gear is rotated in mesh with its own pinion or pinions, or with a cast iron lap having the same face width. An abrasive is placed between the mating teeth and lapping is continued until proper surface h i s h , involute, and face contact are obtained. In the grinding process, as applied to gears of large diameter, the flanks of the gear teeth are formed by the tip of a grinding wheel (which passes over the flanks to generate an involute form) and by the action of the @;rindkgmachine to generate the correct tooth form and helix angle. Although not within the scope of this chapter, the inspection, installation, and alignment procedures used in connection with reduction gears can have a major impact on their successful operation. These subjects are comprehensively covered by reference [I],' which was prepared by the Society's Panel M-12, and reference 121, which describes methods used with naval ships.
(dl Double reduction, double input, articulated
Section 2 Tooth Design hctors 2.1 Tooth Contact Pressure. The most important factor in the design of a reduction gear is the tooth contact pressure; that is, the pressure which exists between the mating tooth surfaces when force is transmitted from one to the other. This factor determines the durability of the working surfaces of the teeth. The tangential force transmitted per unit of gear face width is determined from the expression:
(b) Single reduction, double input
(9) Double reduction, double input, locked train
where
W t = total tangential tooth load, lb F , = effective face width (at pitch diameter), in. R P M , = pinion revolutions per minute H P = horsepower transmitted (per mesh) d = pitch diameter of pinion, in. (el
Dwble reduction, double input, nested
neither of these expressions was an accurate measure of the actual load-carrying capacity of a reduction gear because they did not take into account the contact pressure between mating teeth. The contact pressure is the proper design criterion because it is the factor that determines the satisfactory operation and durability of gears. For many years, in the U. S., the allowable gear tooth pressure for turbine drives was related directly to the pinion diameter so that the loading was specified as "pounds per inch of face per inch of pitch diameter." This was logical since the curvature of the pinion tooth as it affects contact pressure, or more precisely the compressive stress a t the contact surface, is directly proportional to the pinion diameter. Then,
.
I
Fe
(allowable) = J . d
The allowable tooth load per unit of face width where J is an experimentally determined constant, increases with the diameter of the pinion because of the pounds per inch of face per inch of diameter. When gear dimensions are known, the J factor can be decreasing curvature of the contacting surfaces. I n early gear designs, particularly in Britain, the allowable calculated as follows: tooth pressure per unit of face width was taken as proporJ='- W 126,050 HP tional to the square root of the pinion diameter; that is F. d - R P M , . dz . Fa W 6 (allowable) = C& Fe (2) The foregoing relationship, although an improvement, is not precise because it ignores the effect of the curvature where of the mating tooth. A further refinement which takes C = experimentally determined,constant this into account is Yet another expression related the allowable pressure to the two-thirds power of the pinion diameter. However, 1 Numbers in brackets designate References at end of chapter.
-
(c)
Single reduction, double Input, three-bearing pinions
(h) Single reduction, planetary
Rg. 1
Gear arrangements
-
W, =
Wt COB 4,
126,050 H P
.cos # = cos 4, . cos # . RPM,
(6)
d
where W, = total load normal to contact lines, lb 4. = pressure angle (plane normal to teeth) # = helix angle The average total length of all the lines of contact is n
n
Lam
=
average
length of
contact
lines^
Z = length of line of action, in.
P, = normal base pitch, in. Dividing equation (6) by equation (7) the loading per unit of contact line length is equal to
where Wnl = tooth load per inch of contact line, lb/in. The radius of curvature bf the pinion tooth at the pitch diameter is d sin 4 PP =
5-&G&
where p, 4
= =
#*
=
radius of curvature, in. pressure angle in plane of rotation = tan 4, tan-' cos # base helix angle (helix angle at base circle diameter) = sin-' sin # cos 4,
and of the gear tooth is
D sin 4
=
Fig. 2 Imoluh geometry
F.
d
(4)
K---= Wt R + l F..d R
JR+l R
-
126,050-HP ( R + l ) RPM, .dp.F. R
where R = gear ratio K = experimentally determined constant This factor K representing the allowable tooth surface stress is the familiar "K-factor" by which gear loadings are now generally specified. Note that the K-factor is simply the loading per inch face per inch diameter, J in l)/R. Where gear equation (3), multiplied by (R design detsils are known the K-factor oan be determined by the failowing relationahips:
+
It can be shown that the K-factor is a good measure of tooth surface stress, i.e., the maximum compressive stress to which the tooth materials in contact are subjected. Referring to Fig. 2 it can be seen that the total tooth loading in a helical involute gear is carried by a series of straight contact lines extending diagonally from the tip to the root of each meshing pair of teeth. The total force normal to the surfaces in contact is
GG&
=
RP,
where D = pitch diameter of gear, in. R = gear ratio The relationships of helical involute geometry involving pressure angles, lineg of contact, lines of action, etc. are described in numerous places in the gear literature, e.g., reference [3]. The tooth elements in contact may be considered to be elements of two tangent cylinders in contact under an applied force. The compressive stress between two cylinders is given by the Hertz equation:
where
8 = maximum compressivestress between surfaces, n r
psi
- = loading per inch of length, Ib/in. L E = modulus of elasticity, psi rl, rs = radii of cylinders, in. substituting equations (5) to (10) in this Hertz equation, the compressive stress between the pinion and gear teeth becomes :
dx)
where
(allowable) = K
315
REDUCI'ION GEARS
MARINE ENGINEERING
(11) & = (4580 Z sin 2$ d~ The first term includes the modulus of elmticiti and geometric factors which are chosen by the gear designer. However, within practical limits, for steel gears with wellproportioned tooth geometry, this term cannot be varied significantly. The second term is the square root of the K-factor and shows that gears of equal K-factor will have nearly equd compressive stress. With the compressive stresspropor tional to the square root of the K-factor, it would follow that if the allowable stress is considered to be directly proportional to the material hardness, then the allowable K-factor should be proportional to the square of the material hardness. Despite the apparent mathematical exactness of these formulas, many effects on tooth durability are not evaluated by them. Some of these effects, such aa the bending and torsion of the pinion, can be analyzed; but others can be evaluated only by service experience. Among the, latter are the prec&ion-with which the tooth surfaces are formed and the tolerance to small misalignments, vibratory forces, and the inevitable foreign particles which find their way into the teeth mesh. Satisfactory values for the K-factor have been established by experience for the materials in common use, and the commonly specified values are discussed in Section 3.8. I t may be noted that the K-factor controls the size of the reduction gear unit. For a given set of horsepower and rpm conditions, the volume and weight of the gear will vary in nearly inverse proportion to the K-factor. I t is alm significant that the K-factor and the pitch, or coarseness, of the teeth are independent. Under the as~umptionthat the tooth pressure is uniformly distributed over the contact lines, that is, uniformly distributed from the tip to the root of each contacting tooth, the contact compressive stress is affected to only a slight degree by a change in pitch. However, the practical requirements for greater tip relief with coaxser teeth make the tooth extremities of coarser teeth less effective in carrying their share of the load. In other words, the assumption of uniformity of tooth pressure from tip to root, which leads to a minimal calculated value of surface stress, is less valid for coarser teeth. Unfortunately, there is no precise procedure for evaluating this effect. 2.2 Tooth Bending Strength. I n addition to providing the surface necessary to sustain the contact loading
.
31 6
MARINE ENGINEERING
imposed upon gear teeth, the teeth must also withstand the bending moments tending to bend or break the teeth at their roots. Since the teeth are cyclicly loaded at a high rate, the bending stresses in the root portion of the tooth must be kept well within the fatigue or endurance limit of the material. To arrive at a formula for bending stress, it is necessary to make the same assumptions of uniform disthbution of tooth pressure over all limes of contact. The loading per inch of contact line developed earlier is
By substituting equations (6), (7), and (8) into equation (14), the bending stress in the root of the helical gear teeth becomes
Equation (15) contains the important variables affecting bending stress. Further refinement, or a more precise assessment of the stress as it determines the bending fatigue strength of the teeth, can be made by including two additional factors. One is the compressive stress across the tooth root cross section due to the rfdial component of the tooth load which acts to reduce the For a spur gear tooth the highest bending stress occurs bending stress on the tension side; the other is the stress when the load is acting a t the extreme tip of the tooth. concentration created by the root radius adjacent to the The tooth form factor Y which relates the tooth loading critical bending cross section. Both of these additional factors are included in the bending strength derivation to bending stress a t the root is in the military specification for reduction gears [4]. .e Referring to equation (15) it can be seen that the bending stress Sa is directly proportional to the tangential tooth load per inch of face and inversely proportional to where the first power of the tooth dimensions. Other variables t = tooth thickness at root, in. are of secondary importance and change very little with h = tooth height dimension, in. well-proportioned teeth in the usual range of helm angle These tooth dimensions are shown in Fig. 2. For a spur and pressure angle. As a good approximation, the bending stress formula can thus be simplified to gear, the bending stress at the root is computed as (I3)
where
This s8me relationship also holds for helical gears but, C = a constant depending on the tooth proporalthough the assumption that the loading is tip-applied tions, helix angle, pressure angle, etc. is good for spur or low helix angle gears, it is invalid for Wt U = unit loading = NDP steeper helii angles where the loading extends diagonally Fa over a portion of the tooth. Therefore, a diagonal NDP = normal diametral pitch of teeth loading factor should be applied. This is particularly The normal diametral pitch is in inverse proportion to true when comparing designs with different helix angles. The stress, as given in the foregoing, should be reduced the linear dimensions of the tooth cross sections and is by a factor, k, which is a function of the helix angle. therefore an accurate reference for tooth size. As a result, the unit loading, which is simply the tooth loading Equation (13) then becomes per inch of face multiplied by the normal diametral pitch, is a convenient measure of bending stress, just as the K-factor is a measure of surface stress. The allowable unit loadings are generally in the range of 6000 to where 8000. However, this range may be safely exceeded with k = diagonal loading factbr DroDer standards of alignment accuracy, metallurgy, etc. High-powered naval vessels employ unit loads well Values for the diagonal loading factor are given in Table above 10,000. 1. Factors for interinediate values of the helix angle From the considerations of bending stress alone, it may be determined by interpolation. would appear quite easy to lower the bending stress simply by increasing the size of the teeth. However, this entails compromises with surface stress, scoring, and noise considerations; consequently the tooth pitch must Table 1 Diagonal Loading Factors be selected to provide the best balance of all factors. HXLIXANGLE k 2.3 Tooth Scoring Factor. The action of two involute tooth surfaces when rotating in unison is such that the contacting surfaces both roll and slide over each other. At the pitch line point of contact, the sliding component is zero and the contacting surfaces are in pure rolling
REDUCTION GEARS
contact. But the sliding component increases with the distance from the pitch line and is a maximum at the tooth extremities, tip and root. This sliding action, if sufFiciently severe, can cause scoring of the tooth surfaces. This scoring or galling is an actual fusing or welding together of particles of the contacting surfaces. Under the continued motion, particles are torn from one surface and either deposited on the other surface or released. Scoring results from tooth pressure in conjunction with a sliding velocity. The tendency to score is usually assessed by means of a scoring or PVT factor which places a numerical value on a combination of the contact pressure and sliding velocity. A definition of the terms and formulas for calculating P,VT can be found in [6]. A number of additional factors. such as lubricant and tip relief, influence the tendency of gears to score such that considerable expertise and practical experience are
3 17
required to seleat the tooth form, niaterials, surface finish, and lubricant to avoid scoring difficulties. Coarser teeth are more prone to score than finer teeth so that with coarser teeth it may be necessary to modify the involute form to relieve the contact pressure at the tooth tips to avoid scoring. Lubricating oils vary in their ability to prevent scoring, and it may be necessary in some gear designs to use oil having a higher E P or "extreme pressure" quality. Scoring, which results in a serious deterioration of the tooth surfaces, is not to be confused with the minor scratching of the tooth surfaces that results from the passage of minute particles between the teeth. Scratching under certain light reflection can appear to be scoring. Scoring, however, will be rough to the touch. For a comprehensive discussion of the various modes of gear tooth failure and some practical experiences in thi$ regard, see reference 151.
tear Design 3.1
Determination of Approximate Size of Gears.
While the detail design of a reduction gear requires a high degree of skill, it is fairly easy to establish approximate dimensions of a reduction gear. As an example, consider a doublereduction gear which is to be designed to meet the following requirements : Shaft horsepower. . . . . . . . . . .25,000 hp at 108rpm H P turbine. . . . . . . . . . . . . . . .12,500 hp at 6100 rpm LP turbine. . . . . . . . . . . . . . . . .12,500 hp at 4100 rpm First reduction K-factor . . . . .K1 = 140 Second reduction K-factor. . . .Ke = 110 A conventional arrangement, as illustrated by Fig. l(d), haa been selected and suitable dimensions for the pitch diameters and face widths are to be computed. The H P and LP turbines develop equal horsepower; however, the H P turbine turns faster than the LP turbine. As a result, the H P side will require a larger gear reduction and will control the size of the second reduction elements; therefore, it will be computed first. The overall reduction ratio of the H P side is 6100 to 108. As a first approximation, the ratio of the second reduction can be taken as the square root of the overall ratio minus 1.0. (For a locked-train gear, 3.0 would be added to the square root of the overall ratio.) The second reduction ratio then becomes
The loading per inch of face per inch of pitch diameter for the first reduction can now be computed.
125.5 lb/in-in. The next step is to equate two expressions for the tangential tooth load as follows, . HP = Jl . Fa.d wt = 126,050 d . RPM solving for dsF, 126,050 HP - (126,050)(12,500) = 2058 d ' ~ ,= J . RPM (125.5) (6100) Generally, the most economical reduction gear is one where the pinion diameter is as small as possible with relation to its working face. However as will be seen later, the face width-to-diameter ratio cannot be too high if excessive deflections are to be avoided. Ratios of 2.0 to 2.25 represent good practice and 2.25 is selected. With this stipulatipn, the computations may proceed: F, = 2.25 d
and the first reduction pinion diameter is dl = @E= 9.71 in.
afid the first reduction ratio is
with an effective face width of
Fe1= (2.25)(9.71)
= 21.8 in.
The first reduction gear is next computed aa
MARINE ENGINEERING
D l = Rid1
=
(8.66)(9.71) = 84.1 in.
Similar cdculations can now be made for the second reduction : J z = Kz- Rz = 'lo 6'52 = 95.4 lb/in-in. 6.52 1 Rz 1
+
+
This LP first reduction is larger than it would be if it were designed to the maximum permissible K-factor, but this may be offset by the economy of using the same part for both first reduction gears. The pitch diameters as determined in the foregoing must now be laid out to determine if centerline positions 23,460 in. and other arrangement considerations are acceptable. The optimum gear arrangement may require adjusting Again selecting F , = 2.25 d, the second reduction pinion the choice of ratios between first and second reduction0 diameter becomes and the choice of face width-to-diameter ratim. With the approximate diimeters and face widths as d za = A 23 460 = 10,430 in.' determined in the foregoing, the designer will next check 2.25 to determine that bending and torsional deflections are acceptable. Formulas for these deflectionsare developed in the following section. He may select a lower L I D with an effective face width of ratio if these deflections are too high and then adjust diameters and face widths accordingly. Fez = (2.25)(21.8) = 49.0 in. Tooth pitch is then selected to provide the best and a second reduction gear diameter of balance between bending stress, scoring factor, and noise. The best compromise in this regard is generally the finest Dz = (21.8)(6.53) = 142.1 in. pitch permitted by the bending stress or unit loading limits. This will result in an acceptable bending stress, The LP first reduction can be proportioned in the same manner, but it is desirable to design the arrange minimum scoring factor, and minimum noise level. Tooth pitch, addendum, dedendum, pressure angle, ment such that the second reduction pinions on both the HP and LP sides are identical. Since the first reduction etc., and tooth proportions, are made to suit the stangear speed on the LP side must be the same as that on dards for which the manufacturer has tooling. These the HP side (704 rpm), the first LP reduction ratio will be: standards are in small enough increments that no significant compromise is involved. Numbers of teeth are chosen to provide "hunting tooth" combinations between mating pinions and gears, and diameters or helix angles are adjusted to the precise values determined Proceeding as before by the numbers of teeth. A hunting tooth combination is one in which the numbers of pinion and gear teeth have no common prime factor. This means that each tooth will with every tooth of the mating element and 126,050 - H P (126,050) (12,500) - (119.5)(4100) = 3216 in. 8 thus mesh d l Z ~=, avoid any wear or tooth spacing pattern that can J1 . R P M give rise to asub-harmonicof the tooth meshing frequency. Selecting F , = 2.25 d As noted previously, the design of gears is based on the tooth pressure being uniformly distributed across the dl a = - - - 1429 in.' entire face width. Many factors adversely aiTect this tqoth pressure distribution and must be taken into dl =9= 11.26in. account. Among these factors are torsional and bending Fd = (2.25)(11.26) = 25.3 in. deflections of the pinion, accuracy of manufacture, deflections due to centrifugal force, strains due to Dl = (11.26)(5.82) = 65.5 in. temperature variations, and casing distortions due to It may be desirable to use the same first reduction temperature differences and hull deflections. Two of gear on the LP side asused on the HP side. In this case, these factors, torsional and bending deflections of the pinion, are important in proportioning gear elements and, D dl='-"- 14.45 in. fortunately, are readily evaluated. Ri 5.82 3.2 Torsional Pinion Deflection. When' subjected Fel = 21.8 in. to a uniform tooth pressure, a pinion will deflect torsionD l = 84.1 in. ally as shown in Fig. 3. The teeth will separate from the mating gear teeth by the distance y. However, since the 126,050 H P (126,0a)(12,500) W t= d l . R P M - (14.45) (4100) = 26,600 lb pinion is always free to shift endwise to balance the load
Fi:.
REDUCTION GEARS
between the two helices, the separation after this axial shift will be yl on the helix ne& to the coupling m d yz on the helix away from the coupling. The torsional deflection in the space between the helices has no effect on the separation. The separations will then be
r
.
F 1
where
yl = tooth separation at driving end, in. y~ = tooth separation opposite from drioing end, in. c = -where 4 = diameter of pinion bore; d4-d,d c = 1.0 for a solid pinion J = tooth loading, lb/in-in. F , = effective face Foidth of pinion, in. d = pitch diameter of pinion, in.
These equations are based on a uniform distribution of tooth pressure, endwise freedom to equalize load between both helices, an effective diameter for torsion equal to the pitch diameter, and a s h a modulus for steel equal to 12.0 X lo6psi. 3.3 Bending Pinion Deflection. In addition to torsional pinion deflections, the tooth loading will cause the pinion to deflect due to bending stress as shown in rig. 3. Fi. 3 PMonddecHon The pinion can be assumed to be uniformly loaded, m d by using the deflection equation for a simply suppotbed, uniformly loaded beam, the tooth mparation due to d i i t i o n and amount of the helix angle corrections are bending is found to be known, the light-torque contact pattern will be a good indication of the eontact pattern under operating conditions. where Such a light-torque contact check will be made a t the factory to confirm the correct machining and assembly f = tooth separation due to deflection, in. of the unit, and the check will be repeated in the ship F = distance between ends of bertrings, in. installation to confirm that the factory alignment has The remaining terms are as defined previously. been duplicated. These contact chmks can be made by This expression is based on a uniform distribution of observing the transfer of a marking compound such as tooth pressure, the tooth pressure acting over the red lead, Prussim Blue, or light layout lacquer, from one distance between the ends of the bearings, the effective dement to the other. Uniform transfer of compound diameter for bending equal to the pinion pitch diameter over the fuli face width will indicate uniform face contact including the space between helices, the pinion simply under light loads. While satisfactory contact checks supported at the inner ends of the besrings, and the c m be made with v q light torques, they can be made modulus of elasticity for steel equal to 30.0 X lo6psi. with greater reliability with higher torquea. When A generally accepted value for the allowable deflection light-loads are not sufficient to bring about uniform due to torsion and bending is 0.001 in. However, other contact, a quantitative measure of face contact can be d e d s can add to these calculated values. The totd made by gaging the opening between meshing teeth with d e c t can be observed by tooth contact patterns under feeler gages graduated in 0.0001-in. steps. full-load operation,or by estimating from experience on Despite the care which m y be taken in factory and similar gears, or by analysis. This sum may exceed installation tests, the find quality of tooth contact must 0.001, but the gearing can be made perfectly satisfactory be judged after full-power operation in the ship. For by machining corrections into the h d i angles so that this observation, the teeth of each pinion or gear may be the tooth contact will be uniform under f d - l d operat- coated in a band extending across each face with copper ing conditions. When this is done, the cold light-torque by the application of a weak acid copper d p h a t e contact pattern will not be uniform. But since the solution, or with a thin mat of layout lacquer.
3N GEARS
MARINE ENGINEERING ck
'
S
i
E
w = WEIGHT GEAR PER OFWINO PINION OR R
.
TOTAL REACTION PER BEIRlNO
m b D REMION ---ASTERN REACTION
Fig. 4 Typical reduction gear bearing reaction diagram
I
3.4 Slow-Speed Gear Misalignment. An important source of misalignment in the second reduction mesh can be due to the differencein the magnitude of the forward and after slow-speed gear bearing reactions [7]. Figure 4 is a typical bearing reaction diagram for a doublereduction gear. It may be seen that the gear bearing reactions consist of one or more components due to the torque loadings and a component due to the static weight of the pinion or gear supported. With the exception of the slow-speed gear bearing reactions, none are affected by external influences. However, such is far from the case with the slow-speed gear bearings. When the static loads imposed on the forward and after slowspeed gear bearings are different in magnitude, as opposed to being equal as shown in Fig. 4, the resultant reactions will not be in the same direction. This will cause the forward and after gear bearing journals to ride in different positions within their bearing clearances. The slow-speed pinions are not subjected to a similar influence; therefore, there results a crossed-axis condition between the slow-speed pinions and gear. The foundations of slow-speed gear bearings and line shaft bearings are completely dissimilar. Slow-speed gear bearings are located very close to the lube oil sump
tank provided beneath the slow-speed gear (see Fig. 6 of Chapter 1) and, therefore, their foundations become very warm when at operating temperature, causing an attendant thermal rise in the position of the slow-speed bearings. On the other hand, little heat is generated in line shaft bearings, and they operate at a temperature little above the ambient. This being the case, it is unavoidable that the l i e shafting have an influence on the slow-speed gear bearing reactions when the plant goes from a cold to the operating condition. When going from a cold to the operating condition, the slowspeed gear bearings will rise about 15 to 30 mils higher than the line shaft bearings. Prior to the late 1950Js, misalignments due to this source were generally disregarded and the slow-speed gear shaft was aligned concentric to the line shafting. It is easily shown that*the forward slow-speed gear bearing on many of the older ships carried no static load when in the operating condition. It is speculated that the disregard of this factor led to a number of their problems. Although many obstacles are often encountered, if pursued sufficiently early in the design stage it is usually possible to design a system that will not experience
difFiculties of this type. When investigating potential problems of this type, the first step is for the gear manufacturer to state the allowable difference between the static reactions of the forward and after slow-speed gear bearings. The allowable differences usually fall in the range of 20 to 30 percent of the static reactions, and must often be assumed for preliminary studies. Beyond this point a technique, similar to that described in Chapter 11, is employed to ensure that there is adequate flexibility in the shafting system to avoid an excessive variation in the slow-speed gear bearing reactions and to allow reasonable alignment tolerances to be specified. 3.5 Other Deflections. There are other deflections that can act to affect the uniformity of tooth contact acrbss the tooth faces. The gear housing structure will deflect under the forces applied to the bearings and may deflect to misalign the teeth; an example would be the case in which the support of one pinion bearing is more flexible than the support of the bearing at the opposite end of the pinion. Gear casings are also subject to thermal strains and these can affect tooth alignment. For instance, the casing support structure for the bearings in the middle of a double-reduction gear housing may be at a higher temperature than structure which supports the end bearings. The rotating elements are also subject to elastic and thermal strains. Gear rims that are attached to their hubs by a series of thin plates or cone members are deformed by the action of centrifugal forces. The design must be such that these deflections do not have a significant effect on the tooth portion. Thermal strains can also be important, particularly with wide face widths. If the pinion whose teeth mesh a t a higher rate is allowed to reach a temperature higher than its mating gear wheel, the uniformity of tooth contact across the faces of both helices will be affected. 3.6 Critical Speeds. Pinion and gears, designed as they are for stiffness to resist tooth forces, have lateral critical speeds that are well above any operating speed. They will run free of vibration with normal procedures for balancing. Balance is a particularly important consideration with the first reduction pinion since it rotates at turbine speed, and it must be given the same high degree of dynamic balance as the turbine. Coupling shafts connecting the turbine to the pinion are an important element in determining the lateral critical speeds of the turbine rotor-coupling-pinion assembly and must be considered in evaluating turbine ' critical speeds. The combination of the propeller, shafting, gears, and turbines forms a system which can vibrate torsionally in response to the impulses from the propeller blades. With the very early gear designs, manufacturing irregularities in the gear teeth occasionally were a source of serious torsional vibration; however, the precision with which modern gears are manufactured has eliminated this as a
32 1
source of torsional vibration. As discussed in Chapter 11, the first three modes of torsional vibration warrant careful analysis. I n the first mode of torsional vibration with a geared-turbine drive, the angular vibratory motion is greatest a t the propeller, but the vibratory torque is a maximum at the reduction gear. This mode generally occurs within the operating range, being well down in the operating range with arrangements having long shafts but relatively high in the operating range and potentially dangerous with very short shafting arrangements. The first mode of torsional vibration must be evaluated to ensure that the vibratory torque in the gear train, when added to the torque transmitted under steady power conditions, will not be deleterious to the reduction gearing. The inertia and elastic factors of the turbines and gears have no significant effect on the first critical speed; it is controlled by the inertia of the propeller and entrained water, the stiffness of the shafting, and the number of propeller blades. The second mode of torsional vibration is one in which the two turbine branches vibrate in opposition and it may occur in the operating range. When this is the case, vibratory torques must be evaluated as for the first critical. However, by employing a so-called "nodal driveJJarrangement, it is possible to render the second mode incapable of excitation. In a nodal drive arrangement, the two turbine branches are tuned by adjusting the dimensions of the quill shafts, such that they have identical frequencies with the slow-speed gear, shafting, and propeller considered nodal points. As a result, all motion in the second mode is in the turbine branches and propeller excitation cannot excite this mode since the propeller is on a node. The third mode of torsional vibration, in which the slow-speed gear is an antinode, may be of concern. It is usually well above the operating range, but the trend toward larger numbers of propeller blades may cause it to be of importance in the future. It is not possible to avoid tooth separation and the attendant banging sound from the gearing during deceleration through a critical speed with little or no power being transmitted by the gearing, or when a critical occurs at very low power. However, this is a transient condition at low torque levels and is not damaging to the gears.' 3.7 Gear Case. The function of the gear case is to furnish adequate support for the bearings as well as to provide an oil-tight enclosure for the reduction gear. Typical gear cases may be seen in Figs. 5 and 6. All journal bearing load reactions are in planes perpendicular to the axis of the revolving shafts. In many instances, and particularly in connection with double-reduction gears, the bearing supports will have to support bearings at different elevations. It is of the utmost importance that these bearing supports including the cap have sufficient structural stiffness to prevent any measurable deflection under varying load conditions. Due to the
MARINE ENGINEERING
REDUCtiON GEARS
AFT -
AFT END F L E W T l q
Fig. 5(a] Miculaied doublarsduction pear
direction of rotation of the different shafts and the location of pinions in reference to gear bearing loadings, Fig. 4, reactions may occur a t any angle to the axis and it is important that the bearing cap construction takes this into account. It must be borne in mind that for satisfactory operatian of the gears and to minimiae wear the revolving shafts must operate continuously parallel to each other. The gear c w construction is the only means provided to maintain the diierent shafts in their correct relation to each other. The construction and stBneaa of the gear case must be
studied and compared with the structure and rigidity of the foundation below the gear case whereby the gear cape is secured by bolting to the ship structure. The gear casing generally is rigidly bolted to the foundation to form ti combined structure to prevent deflections between the gear and pinion axes which may be caused by deflection in the ship structure when operating in a heavy sea. Since the strains in the hull due to the loading of the ship and to the forces imposed by the seaway are imposed on the gear casing, designers are attracted to means for isolating strains in the ship's hull from the
Fig. 5(b) Miculatsd dwbb-reduclion gear
main gear casing. Two, three, or four points of support between the gear case and its foundation can effectively accomplish this isolation. Except for small auxiliary gears, the casing has separate inspection covers for convenience in inspecting the condition of the gear teeth, and is arranged so that bearings, flexible coupliiga, and oil sprays may be inspected and replaced without having to dismantle large sections of the casing. The casing enclosure acts to attenuate the noise that is generated by the meshing teeth. Even thaugh the most effective means of reducing gear noise is the precision
with which the teeth are machined, the gear case offers a means of further quieting by designing t o minimine the transmission of sound. Some of the smaller gear cases are made of steel castings. However, in-the propulsion gear siaes, the casings are of fabricated steel construction. Some castings may be employed in the fabrication for the heavier sections, but the trend is sway from castings in favor of sections burned out of heavy plate or formed from plate material. 3.8 Pinions and Gear Wheels. Pinions am made of a one-piece forging and may be hollow-bored to accommodate a quill shaft. The requirement for strength and
MARINE ENGINEERING
Fig. 6(a)
REDUCTION GEARS
Locked-train reduction gear
rigidity generally precludes making the tooth and journal portion of the pinion in more than one piece. The material is usually a nickel-steel, through hardened to the desired hardness. The tooth portion of the gear wheel is usually a carbon steel forging that is either welded to a center portion made up of steel plate in the larger diameters, or integral with the center portion in the smaller diameters. For many of the older turbine drive gears now in service in the U. S., the most common materials are pinions in the 200-240 Brine11 hardness number range running with gears in the 160-190 Bhn range. With these hardnesses, K-factors of about 90 for the k t reduction and 75 for the second reduction were generally applied. Higher hardness materials with. higher Kfactors have wide application in naval combatant ships and are tending to be accepted for commercial service. In the more recent designs, K-factors of 140/110 (for the first and second reductions respectively) have been applied using through-hardened pinions with a Bhn above 300 and through-hardened gear rims with a Bhn above 220. The higher K-factor gearing has the advantage of increasing the power capability of a given size of gear unit approximately in direct proportion to K-factor. The higher K-factor gear is therefore more compact, lighter in weight, and lower in cost.
As an interesting side note, the Vespasian had a Kfactor of 78, and the Neptune 125, attesting to the genius of the early inventors. The upper limit for the K-factor loading is presently uncertain and cannot be firmly established by analysis, or even by laboratory testing. The service experience of the considerable number of gears in merchant service with higher K-factors and harder materials will have to determine if still higher loadings can be applied with confidence. Pinion and gear materials must be of high quality, and heat-treating must be carefully controlled so that the tooth portions can safely accept the high stresses imposed on them. Case-hardened or through-hardened and ground materials and nitrided materials offer the use of considerably higher K-factors, and more compact and lighter weight gearing. They have been used in many diesel engine and turbine applications, mostly outside of the U. S. Ground gears are particularly suited for planetary gearing. Heat-treating and grinding requirements impose conditions upon the design of ground gearing that do not apply to through-hardened gears of hardnesses that can be machined without grinding. Where ground elements are applied to conventional gearing, they are generally made single rather than double helical.
Fig. 6(b)
Locked-train reduction gear
This is done mainly because of the clearance required at each end,of the tooth face for the grinding wheel. 3l9 Journal and Thrust Bearings. Journal bearings must carry the weight of the gear elements and also transmit the large tooth meshing forces to the casing structure. Hydrodynamic bearings have been used almost exclusively in this application, the conventional babbitt-lined, steel-shell sleeve bearing being extremely long-lived, with a high tolerance for abnormalities, such as dirt and rust, in the operating environment.
325
As may be noted from Fig. 4, the tooth meshing forces for ahead and astern rotation are in nearly opposite directions and generally in different directions than the weight reaction. It is necessary, therefore, to select an angular position for mounting the bearing in the housing so that the bearing areas are in the best possible relationship to the applied bearing forces. First reduction pinion bearings operate at high speeds. When starting and at low speeds, the bearing forces are low, consisting almost entirely of the weight components. At higher speeds, however, the bearing reactions continuously increase. These conditions are favorable for hydrodynamic bearings and permit the safe use of high unit loadings in these bearings. Unit pressures of 225250 psi of projected area are generally acceptable for high-speed journal bearings in commercial service, and considerably higher pressures are used successfully in naval service. The clearance ratio for these high-speed bearings should be 0.002 to 0.003 in. per inch of journal diameter. This clearance ratio is important in defining the difference in radii of curvature of the journal and bearing surfaces, and thus the degree of convergence and divergence of the load-carrying oil a m . The clearance opening, as such, in the unloaded half of the bearing affects only the quantity of oil passing through the bearing. Second reduction gear bearings operate at lower speeds and have relatively high static loadings due to the weight of the bull gear. These conditions are less favorable and require lower unit pressures of 150-175 psi. The clearance ratio should be about 0.001 in. per inch of journal diameter. Intermediate-speed bearings fall between the high and low-speed bearings, with loadings of 175-200 psi and clearances of 0.001 to 0.0015 in. per inch of journal diameter. In addition to carrying a load, the journal bearings must accurately position the gear and pinion journals to keep their axes precisely parallel. Replacement of bearings, therefore, must be made so as not to alter the journal position. To facilitate bearing replacement, a common practice is to stencil on the bearing shell its shell thickness at several points. Then, a replacement bearing with the same shell tbckness will maintain the original jqurnal position. The main propeller 'thrust bearing is generally either integral with the gear unit or immediately adjacent to it. Its main purpose, of course, is to transmit the propeller thrust to the hull, but a secondary purpose is to hold the second reduction gear wheel in its proper axial position. With double-helical gearing the main thrust bearing also holds the second reduction pinions in their axial position, and further, by an axially restricted coupling the first reduction gears can also be positioned. It is also common to couple the first reduction gear to its second reduction pinion with a coupling which permits endwise motion. Then, positioning thrust bearings must be provided for the first reduction elements. This can be
326
MARINE WGNEERING
/
TURBINE ROTOR
1
either a pivoted shoe or plain collar thrust bearing applied to either the first reduction pinion or its gear wheel. This bearing must have sufficient capacity to overcome the frictional forces in the couplings which act on the first reduction elements. Where adequate foundation structure can be provided, it is convenient to locate the main thrust bearing forward of the second reduction gear, with the thrust housing an integral part'of the gear casing. This location has two advantages; (a) the diameter of the thrust bearing can be smaller because the shaft portion does not have to transmit torque, and (b) the thrust collar can be a separate piece that can be readily removed over the end of the gear shaft in the event that it ' is necessary to replace or refinish the collar surface. For higher powers, and where greater stiffness is required for the thrust bearing foundation, the thrust bearing is located aft of the second reduction gear. An installation in which the main thrust bearing is located immediately aft of the slow-speed gear is shown in Fig. 5. The thrust housing structure is independent of the gear case and joins it by a flexible oil-tight connection. The bolted attachment to the foundation which transmits the propeller thrust to the hull is independent of the bolting attachment of the gear base to its foundation, so that the thrust bearihg and its foundation can deflect as a result of the propeller thrust with no distorting effect on the gear casing. In either location; the thrust bearing shares the lubricating system with the gear, and its oil drain discharges into the gear base. Only a single shaft oil seal is required on the output shaft. The thrust bearing is of the pivoted-shoe type, with two sets of shoes acting on opposite sides of a thrust collar to accept thrust in either direction with pressures of about 375 psi. All journal and thrust bearings are force-fed from a central lubricating oi1 system. Each journal and thrust
bearing is generally provided with a sight-flow and thermometer fitting in a visible location so as to provide an indication of performance. As a sample of oil leaving the bearing passes through the sight flow (or bubbler) it provides a visible jet of oil that can be seen at some distance, giving assurance that the bearing is being properly lubricated. This oil also passes over a thermometer well installed integral with the sight-flow fitting for sensing and indicating either locally or remotely the temperature of the oil leaving the bearing. 3.10 Couplings. The coupling of each gear and pinion to its connecting shaft can be of a number of types depending upon the degrees of freedom of movement that the service requires. The second reduction gear to line shaft coupling is usually a "soli&couplingnwith the flanges integral with the shaft sections. This coupling provides no freedom of movement within itself either axially, angularly, or torsionany. As discussed previously, the line shaft bearings and the second reduction gear bearings cannot be held in absolute alignment due to thermal and other distortions in the hull and foundations; however these movements are predictable and the bearing arrangement can be designed such that, when properly aligned, the shafting can bend elastically without imposing objectionable stresses in the shaft or altering the bearing reactions in an unacceptable manner. Thermal distortions in the gear and turbine casings and their supporting structure create a relative movement of the turbine rotor and high-speed pinion axes that introduces an angular offset at one or both coupling elements. In addition, the turbine rotor is positioned axially by its thrust bearing; consequently, the thermal growth of the turbine rotor due to the high-temperature steam creates a considerable end motion which must be accommodated by endwise sliding and clearances in the coupling. For steam turbine-driven gears the coupling to the first reduction pinion is usually a gear tooth (dental) type flexible coupling, with two gear tooth elements separated by a length of shafting or a sleeve. Figure 7 is a typical coupling of this type. The engaging tooth elements at each end use internal and external spur gears of involute form, for convenience of manufacture, which mesh with backlash in the circumferential direction, but with closely controlled radial clearance between the tips of the external teeth and the roots of the internal teeth. Under torque the axes of the two elements are held in line by the contact on the involute tooth faces. With no torque transmitted, the axes are held in line within the limit of the radial clearance. When running under angular misalignment, each meshing pair of teeth will slide back and forth a small amount. The angular misalignment which this type of coupling can accept without significant wear is limited and is dependent upon the coupling size, speed of rotation, torque, and hardness and finish of the tooth surfaces. It is obviously desirable to avoid excessively short coupling lengths which impose high angular movements on the tooth elements. For longer coupfing lengths it may be
)N GEARS
I
I
I 1
(
I
preferable to use a "single-ended" flexible coupling with a tooth-element coupling at one end and a solid coupling at the other end. In t h i case the long shaft can deflect elastically as a cantilever beam to accommodate the lateral offset. Couplings between the h t reduction gears and second reduction pinions have smaller misalignments to accommodate, but otherwise resemble the turbine-to-pinion ooupling. In the single-case gears, Figs. 5 and 6, the misalignment is limited to the clearance in the journal bearings since the bearings themselves are held rigidly in line. Like the turbine to pinion coupling, they can be "single-endedn [as shown in Fig. 6(b)] or doubleended [as shown in Fig. 5(b)]. When the first redudion elements have their own thrust bearing, at least one flexible element is needed for endwise freedom. When the first reduction elements are positioned by the second reduction and the main propeller thrust bewing, the flexible coupling elements are made with a olose end alearance. Lubrication of the coupling teeth is important even though the reciprocating sliding velocity is entirely too low to support an oil fdm between the surface in contact. Oil is held in the tooth portion by centrifugal force and an oil retaining ring keeps the sliding surfaces submerged in oil. Oil is fed to the annulus at one end of the teeth and leaves from the other end, forcing a flow endwise through the teeth for lubrication, cooling, and purging of the sludge which tends to centrifuge and collect. For diesel engine drives it is usually necessary to have a coupling with torsional flexibility to minimize the transmission of torque variations to the gearing. Several types of these, using rubber or other elastometem in compression or shear, are effective in adding both torsional resilience and damping to attenuate the torsional oscillations which are inherent in the reciprocating engine. Hydraulic couplings, now familiarin automotive transmissions, had one of their earliest applications in marine drives. They are effectivein smoothing the torque input to the gear. However, their slip repreaents a direct power loss. Electric couplings have characteristics similar to the hydraulic coupling but are dependent on a source of electric power for their operation. Both hydraulic and electric couplings have the capability of providing a convenient means of disconnecting, synchronizing, and reconnecting engines in a multi-engine arrangement. 9.1 1 Clutches. Propeller drives that use either prime movers in combination or a prime mover which is uni.iotational may require a clutch to disconnect and reconnect or synchronize and reconnect the main engines from the propeller. An assortment of devices using mechanical, frictional, bydraulic, or electrical schemes is available for these purposes. Each device has,its peculiar characteristics, so that the selection of the best coupling arrangement depends on the requirements in a specific case. The hydraulic coupling transfem torque by the passage
327
of oil between two halves of a torus. -The torque trammitted can be controlled, therefore, by controlling the volume of oil in the coupling. It can be arranged to quickly diseharge the oil in the coupling to disconnect the load, and to reconnect the load by readmitting oil. A hydraulic coupling will absorb energy to bring the shafts into near synchronization; however, there are h i t s to the hydraulic torque available for synchronization, and limits on the amount of energy that can be absorbed during the period of high slip. An electric coupling has characteristics very similar to a hydraulic coupling. It too can act as a disconnect and as a synchronizing clutch, by controlling the current to the rotating field. Like the hydraulic coupling, there are limitations on the synchronizing torque, and the energy absorption during synchronizing. Chapter 10 contains a description of electric couplings. Friction clutches use friction elements which slide under eontrolled pressure to bring the shafts into synchronization and then, once synchronized, transmit torque without slip by the same friction surfaces. Figure 8 shows the application of clutches of this type in a reversing gear train. In this case the friction material is attached to the inside of an inflatable tube and is made to bear on the cylindrical drum which it surrounds by admitting fluid, usually air, under p m u r e to the tube. The inflatable tube, called the clutch gland, is made of fabric and rubber similar to an automobile tire and is bonded to a steel outer ring. Both ahead and astern clutch glands are driven by the engine. One engages with the drum driving the ahead gear train or the other engages with the drum driving the astern gear train. Reversing j~ accomplished by alternately admitting fluid and inflating the a h a d and astern clutches. When the idle clutch is deflated, the friction surfaces are removed from contact. The operating air is admitted to the rotating shaft through a shaft seal. This is conveniently done at a shaft end as illustrated but can also be done on any available shaft portion. With synchromesh couplings the normal torque transmission is through sets of engaging internal and external tooth elements similar to those of a "dental" flexible coupling. In addition, the coupling includes a mechanism for shifting the tooth elements d y to engage and disengage the teeth, a friction element to brieg the shafts intQ synchronism prior to engagement,. a balking mechanism to prevent tooth contact while a Merentid speed exists and, in the case of high-speed applications, a "transition torque control" to maintain shaft synchronism during the shift from friction t o gear tooth drive. Referring to Fig. 9, the clutch is engaged by admitting air or oil to the operating cylinder. This applies pressure to the friction disks, and the torque so developed acts to synchronize shaft speeds. When synchronism is reached the accelerating torque is reduced to zero. The balking mechaniwn then automatically releases the coupling sleeve, and as it approaches engagement with its mating hub the spring-loaded pins act as lowe keys to maintain synchronous speed and align the hub and sleeve teeth
b
i
MARINE ENGINEERING
r(
Fig. 8
Reverse gear with friction clutchw
DISENGAGED POSITION
Fig. 9
EN-GAGED POSITION
Synchromesh. coupling
for final engagement. After engagement the torque is transmitted through the dental coupling elements; the friction disks and balking mechanism perform no further function. The clutch is disengaged by simply shifting the sleeve endwise so that the hub and sleeve teeth at one end move out of mesh. Once engaged or disengaged, the clutch is maintained in the desired position by mechanical means and hydraulic pressure is no longer reauired. ?he torque-transmitting capability of the synchromesh clutch is determined by the tooth elements. The torque capacity of the friction disks need only be sufficient to bring the shafts into synchronization. A synchro-self-shifting clutch, like the synchromesh coupling, is a positive coupling in the engaged position. It is self-engaging when passing through synchronism; that is, immediately upon synchronizing the speeds of the inputand output shafts, the input shaft engages the output. The clutch disengages automatically as soon as the torque reverses, that is, when the output shaft tends to drive the input shaft. In this coupling, engagement and disengagement are brought about by the relative rotation of the driving and driven ends. The primary action of the clutch, in fact, resembles that of a ratchet which will lock up to transmit torque in one direction, but will turn freely under a torque in the other direction. The basic clutch, Fig. 10, has a torque-transmitting sleeve which can shift axially. One end of this sleeve is in constant engagement with the output shaft through a helical spline. The other end of the sleeve has dental coupling type teeth which engage and disengage with mating teeth in the clutch ring which is secured to the input shaft. The position of the sleeve is controlled by a ratchet-and-pawl arrangement which senses the relative speeds of the input and output shafts. When the speed of the driving half overtakes the speed of the output half, the pawls engage so that further rotation of the driving half forces the sleeve to move axially on the helical spline to bring the coupling teeth into engagement. The coupling can also be made with two sets of spur dental type coupling teeth so that in the engaged position it also acts as a flexible coupling. It can be arranged with a manual shift that will prevent engagement of the pawls and allow its driving engine to be tested without driving the propeller. 3.12 Lubrication. The main reduction gears are normally provided with a source of lubricating oil by a system which is separate from the gear itself. In the case of steam turbine drives, the same source of oil also serves the turbine requirements. The distribution system for leading oil under pressure to each bearing hnd to the tooth sprays, and for containing and leading the drains to the oil sump, is contained within the gear. All the gear requires is a continuous supply of clean oil at a pressure of about 10-15 psi and a temperature of about 120 F. The oil distribution system integral with the gear is designed to provide each bearing and tooth spray with the proper quantity of oil.
329
GEARS CLUTCH RING
HELICALLY SPLINED OUTPUT SHAFT
Fig. 10 Synchro-self-shifting dutch
The oil supply to the gear can range from 110 to 130 F in temperature. The discharge from the high-speed journal bearings may be as high as 180 F but the average temperature rise in the total flow through the gear unit is of the order of 20-30 deg F. Oil in the viscosity range of 380-510 SSU at 100 F is generally suitable. It represents a slight compromise between the lighter viscosity oil which would be optimum for high-speed journal bearings and the somewhat heavier viscosity which would be more favorable for the tooth meshes. Moderately loaded gear units with relatively fine pitch teeth will operaie satisfactorily with a good grade of straight mineral oil; however, more heavily loaded gears, and particularly gears with teeth of coarser pitch, require an oil having good "extreme pressure" or antiscuffing qualities to prevent scuffing or galling. The "EP" quality of the oil is evaluated by tests such as the "Ryder" or "Three-ball" test in which the ability to resist scuffing between two sliding steel surfaces is measured. Generally, the "EP" quality is given to the oil by the addition of chemical agents. A secondary function of the circulating lubricant is to carry away the heat losses of the gearing and its bearings. As indicated by Fig. 11, the efficiency of reduction gears varies with the horsepower rating. In addition, the efficiency of a gear depends on the type of gear and the particular bearing and tooth pressures; that is, a conservatively designed gear with low bearing pressures and K-factors will have a lower efficiency than its more heavily loaded counterpart. Another secondary function of the lubricating oil is the prevention of rusting of the interior surfaces of the gear. Iaodern oils have good antirust qualities even in the presence of the small amounts of moisture which cannot be avoided. They are completely effective during operation when all inside surfaces are thoroughly washed with oil. For prolonged shutdowns, however, the normal lubricating oil will drain from the steel parts and become ineffective. As a result, special precautions
330
MARINE ENGINEERING
must be taken to prevent rust damage to the gear during prolonged shutdowns. The lubrication system is the only service that is essential to the performance of the reduction gear; therefore, for remote control, it is necessary to monitor only the pressure and temperature of the lubricating oil supplied to the reduction gear. Most remote-control systems, however, will monitor the temperature of the oil leaving each bearing to give an indication of the performance of each of the reduction gear bearings and the main thrust bearing. It is general practice to provide a vent to avoid a build-up of pressure within the gear. A single vent
PERCENT HORSEPOWER
Fig. 11
Efficiency of double reduction geon
opening from the entire gear oase and sump space is sufficient. When more than one vent connection is made, all vents must be interconnected to avoid a circulation of outside air through the gear interior. 3.1 3 Accessories. In addition to performing its primary function of transmitting power, a number of accessory features axe often provided as a part of the gear unit. For turbine applications, a motor-driven turning gear is provided with the main reduction gear. Its primary purpose is to rotate the turbine rotors slowly during warm-up and cool-down operations. It is essential that the turbine rotor be rotated when heating or cooling as the rotor will otherwise not be of a uniform temperature and will bow. The turning gear also serves the useful purpose of providing a means of turning the shaft for other purposes such as inspection of the gear teeth. While the turning gear is a low-powered device (2.5 to 10 hp), it is mecha'nically capable of developing large torques in the propeller shaft by virtue of its high gear ratio. On occasions therefore, the turning gear serves the additional purpose of being a locking device that prevents rotation of the turbine and gears. This is a desirable feature under casualty conditions when the ship is being towed or, in the case of a multiple-screw ship, when driving with other shafts. The turning gear is generally a double-reduction gear with two worm and wheel reductions to connect the driving motor to the after end of one of the first reduction pinions. The ratio of the turning gear is selected so that the propeller shaft turns at a rate of about one revolution in ten minutes. Figure 5 shows a turning gear which drives the aft end of the first reduction pinion on the low-pmure side. A disconnect clutch, usually lever operated, uncouples the turning gear for normal operation. The turning gear requires lubrication and this is pro-
.. (4) (5)
5000
10000
TURBINES K-FACTORS OF 125 FOR FIRST REDUCTION, 100 FOR SECOND REOUCTION WEIGHT OF MAIN THRUST BEARING NOT INCLUDED
15000 20000 25000 HORSEPOWER
30000
Rg. 12 Approximob weight of reductkn geon
35000
40000
vided by the main lubricating system since the main the loss of oil supply due to an interruption of electric system must be in operation even for the low turning power or the inadvertent securing of a motor-driven speed. pump. This feature, however, is seldom applied in Interlocks with the turbine throttles or warning plates merchant service. are necessary to prevent the inadvertent starting of the 3.14 Weight Estimates. An approximate weight of propulsion turbine with the turning gear engaged; other- the reduction gear unit for conventional articulated and wise, extreme damage may be done to the turning gear locked-train gears for steam turbine-driven merchant and motor. ships can be obtained from the curves in Fig. 12. These The main gear also may serve as a convenient mount- curves show that the gear weight varies in nearly direct ing for the shaft revolution counter and tachometer. It proportion to the horsepower and inverse proportion to is usually mounted just forward of the slow-speed gear the propeller speed. Within reasonable limits, the weight shaft and is driven directly or through gearing from the of the gear will also vary inversely with the K-factor. main shaft. For more complex gear arrangements, there is no simFor naval applications, it is standard practice to drive ple procedure f o i determining weights and dimensions. the main lubricating oil pump by a train of gearing taking These must be determined by first establishing the prinits power from one of the intermediate shafts. This cipal dimensions of the rotating parts, and roughly so-called "attached pump" furnishes oil for all purposes detailing their weights. A good approximation of the during normal operations; however, separately driven total weight can then be made by doubling the weight pumps axe required for low-speed, stand-by, and astern of the rotating parts. operation because the attached pump cannot supply an A procedure which is useful in establishing an approxiadequate oil supply under these conditions. An at- mate weight, as well as dimensions, of a reduction gear tached pump has the advantage of protecting against is outlined in reference [ti].
Section 4 Applications 4.1 Articulated DoubleReduction Gears. Most of the gear arrangements that have been used for ship propulsion are described in Section 1. The vast majority of turbine-driven merchant ships built in the U. S. in the past several decades are double-reduction with the rotating elements arranged as in Fig. l(d). In most of these the arrangement of the rotating parts can be further categorized as "three-plane. " The number of planes refers to the number of horizontal planes which contain the pinion and gear axes (with small differences in elevs tions ignored), and to the number of horizontal joints in the gear housing to provide for assembly. The section drawing shown in Fig. 5 is representative of a typical double-reduction, articulated, three-plane reduction gear in a single housing structure. It is widely used with cross-compound steam turbines up to horsepowers of about 30,000 with the propeller speeds that are conventional for merchant ships (i.e., 105 to 120 rpm for single-screw ships). Variants of this arrangement are the single-plane and two-plane gear casing constructions which use similar rotating parts, but with modified dispositions of the turbine and propeller shaft axes. Figure 13 compares these categories. Whereas the three-plane gear is generally a single structure supported by the ship's foundation, the single and two-plane gears are more conveniently built as separate first and second-reduction units, each independently mounted on the ship's structure. The three-plane gear requires maximum headroom
but, in return, provides good spacing between the turbines and ample space beneath the low-pressure turbine to install the condenser. At the other extreme, the singleplane gear requires a minimum of headroom and foundation structure underneath the turbines. On the other hand, additional overall length of the propulsion machinery is necessary because the condenser must be located forward of the low-pressure turbine. The twoplane gear is intermediate in its space requirements as regards headroom, length, and width. The space available for the condenser under the LP turbine usually is less than convenient. The choice of one, two, or three planes is, therefore, largely one of arrangement. 4.2 Locked-Train Gears. For higher powers, the diameters, face widths, and tooth bending stresses of the gears in a conventional double-reduction arrangement increase to the point where a locked-train arrangement becomes the most practical. By dividing the power paths from each input pinion as illustrated by Fig. l(g), the diameters, face widths, and tooth loading factors become more favorable, but at the price of a greater number of parts and a more complex casing structure. The division of power between the two intermediate shafts driven by a common high-speed pinion must be nearly equal, and to accomplish this they must be "timed. " That is, the driving pinion must contact both driven gears when all backlashes are taken up. There are a number of ways in which this may be conveniently done. Once done the meshing gear and coupling teeth
332
MARINE ENGINEERING
fig. 13
Double-reduction gear arrangements
Figure 6 is a section drawing of a typical locked~train gear for merchant marine service. It follows the construction that has been almost standard for naval combatant ships from destroyer escorts to aircraft carriers since the mid 1930's. It differs slightly in proportions since the gear ratio is higher for merchant ships than for the faster-turning propellers of naval ships. Note from Figs. 6 and 13 that the locked-train gear is more restrictive in pinion locations (and gear arrangement) than the conventional double-reduction gear. 4.3 Reversing Gears. There are many applications of single-reduction gears for diesel engines in the moderate-power range. The single-reduction reversing gear, Fig. 8, is a typical arrangement using friction couplings for alternately driving the ahead and astern gear trains. Double-reduction reversing gears employing similar principles have also been used with a gas turbine as the prime mover. 4.4 Gears for Contrarofating Propellers. Contrarotating, coaxial propellers are often given consideration because of their improvement in propulsive efficiency. They require special gear arrangements, and many variations are possible [9]. Contrarotating gear arrangements fall into two distinct categories; one drives the two propellers a t equal or other predetermined ratio of revolutions, and the other drives the two propellers with equal or other predetermined ratio of torque. An example of the first category is shown in Fig. 14. Here a cross-compound steam turbine arrangement, with oppositely rotatkg turbines, drives the oppositely rotating propeller shafts through double-reduction gears. Note that the two first reduction gears are in mesh to assure that the two propeller shafts will make the same number of revolutions even though the power inputs of ithe two turbines may not be equal. As an illustration of the second category, also for a cross-compound steam turbine, each turbine drives the sun gear of a planetary first-reduction gear. Referring to Fig. 15, the planet carriers rotate in the direction of the turbines to drive one second-reduction gear. The ring gear rotates in the opposite direction to drive the other second-reduction gear. The ratios of the two second-reductions can be made to differ by the amount of the torque difference in the output shafts of each planetary gear so that equal torques are applied to each propeller shaft. References
are match-marked so that the timed assembly can be repeated. However, if a pinion, gear, or coupling is replaced, it 'is necessary to retime. The quill shafts connecting the first reduction gears to the second reduction pinions have considerable torsional flexibility and thus act to divide the torques equally despite slight inaccuracy in timing. Without this flexibility, timing and machining would have to be perfect to obtain equal power division.
1 "Guide to Propulsion Reduction Gear Alignment and Installation," T & R Bullet+ 3-10, SNAME Panel M-12. 2 "Reduction Gears, " Naval Ships Technical Manual, chapter 42, Department of the Navy. 3 "Reference Information-B~ic Gear Geometry," American Gear Manufacturers Associa;tion, Standard 115.01. 4 "Gear Assembly, Propulsion (Naval Shipboard
REDUCTION GEARS
Fig. 14
Fixed-axis gean for contrarotation, equal rotation
t
Fig.
IS
Planetary reduction gear for contrarotation, equal torque
Use), " Military Specification MIL-G-17859, Department of the Navy. 5 "Nomenclature of Gear-Tooth Wear and Failure, " American Gear Manufacturers Association, Standard 110.03. 6 D. W. Dudley, Practical Gear Design, McGrawHill, New York, 1954. 7 H. C. Andersen and J. J. Zrodowski, "Co-Ordi-
nated Alignment of Line Shaft, Propulsion Gear, and Turbine"slJ'Trans. SNAME, vol. 67, 1959. 8 "Ship Design Computer Program-Locked-Train Double-Reduction Gear Design, " NAVSHIPS 0900-00& 5070, Naval Ship Engineering Center, July 1966. 9 W. I. H. Budd, "Main Reduction Gears for Contrarotation," Marine Technology, vol. 6, no. 4, October 1969, pp. 440448.
ELECTRIC PROPULSION DRIVES
CHAPTER X
I Electric Propulsion Drives
w* E*
c
o
Section 1 lntrbduction
1.1 Signiflcant Features. Electric propulsion drives offer a number of important advantages which, in certain applications, more than overcome the inherently higher first cost, increased weight and space, and the higher transmission losses of such systems. The ease and convenience by which propeller speed and direction are controllable in an electric drive are among its foremost features. This control can be effected from a number of remote locations, and thus put directly in the hands of the vessel navigator or operator. The electric control system lends itself to any desired speed of response and to the incorporation of time constants or limits, so that the machinery is not subjected to abusive or dangerous operation. The electrical connection between the generator and the propulsion motor provides a freedom of installation arrangement not offered in a mechanical drive system. The prime mover and its generator can be located wherever desired and constitute a mechanically independent enginbgenerator set. The propulsion generator sets are usable as a source of power for other functions when not being used for propulsion. Cargo pumping, fire-fighting, cargo-refrigerating, and dredge-pump driving are illustrative of such supplemental service. In cases where the development of the desired propeller power requires the use of multiple prime movers (a typical situation in high-speed diesel drives), the electric drive provides a convenient means of coupling the several units to the propeller without the use of mechanical clutches or couplings. An electric drive system will normallv be arranged so that vessel operation at less than full p w e r will be &complished with minimum number of prime movers in service. This contributes to more efficient operation and affords downtime for routine maintenance on units not then required for propulsion. Some prime movers, gas turbines, and high-speed diesels, for instance, are unidirectional machines. An electric drive can provide the required reverse rotation of the propeller by relatively simple control means. It is possible and oftentimes convenient to select a low-speed motor that matches the desired propeller speed directly without the use of reduction gearing. This motor can be supplied with power from high-speed generators, thus providing the necessary speed reduction
a
between the prime mover and the propeller without the use of mechanical speed-reducing means. Electric drives have been built in a wide range of sizes (up to 45,000 shp pet shaft). However, such high power applications are exceptional, and most electric drives are of more modest ratings. 1.2 UtilizatKon. Electric drive propulsion systems are applied when their ease of control, flexibility of arrangement, and adaptability to multiple use of the prime mover generator sets can be utilized to advantage. In general, propulsion electric drives are employed on vessels falling in the following broad categories: (a) Vessels requiring a high degree of maneuverability. Vessels such as ferries, icebreakers, tugs, oceanographic vessels and cable-layers are typical of this group. Here changes in speed and in direction of propeller rotation are frequent and of vital importance to the successful operation of the vessel. Often these electric drives will utilize d-c machinery because of the superior speed control inherent with such machines. (b) Vessels requiring large amounls of speciaLpurpose power. Vessels such as self-unloaders, fireboats, selfpropelled dredges, large tankers, and tenders represent applications where effective use can be made of the propulsion generating sets for other special purposes. The nnturc of the service encountered by such vessels permits the generating sets to be applied to supply large amounts of power for pumping or cargo handling when propulsioll power demands are low or nonexistent. Thus a saving is made in the amount of power-producing equipment that must otherwise be installed for these purposes. (c) Vessels utilizing nonreversing, high-speed, and multiple prime movers. High-speed diesel engines and gas turbines are generally of the unidirectional type and frequently are employed in multiple to produce the required prime mover power. Here electric drive provides the means for reversing the propeller, converting the high speed of the prime movers to suitable propeller speeds, and electrically coupling the multiple units to a single drive. (d) Deep submergence vehicles. Such vehicles usually employ relatively small amounts of propulsion power and move at low speeds. The energy is obtained from batteries and is applied directly to multiple motors or
converted by static means to variable-frequency a-c power and applied to a-c motors. 1.3 Types of Electric Drive Systems. From an electrical point of view, electric propulsion systems may be classified aa direct-current or alternating-current systems. A further significant aapect is the type of prime movers used, with the major alternatives being diesel engines, gas turbines, or steam turbines. In addition to the foregoing, it should be noted that the rapid progress made in the field of high-capacity silicon rectifiers has permitted the development of a hybrid system in which the generators are of the a-c type and the drive motors are of the d-c type. Bridging the gap between these two types of machines are banks of silicon rectifiers which convert the a-c power to d-c. This is a highly significant, development because it permits the use of a-c generators which are compatible in size and speed with high-speed prime movers, and retains the favorable speed control characteristics of the d-c drive system. The steadily increasing speeds and powers of diesel engines have outstripped the corresponding capacity of conventional d-c generators, but the availability of practical rectification devices now permits the generation of src power in machines capable of matching any foreseeable engine speed and power rating. Turbines, either steam or gas, are incompatible with conventional d-c generators as to speed and powers. Invariably, reduction gears must be employed to adapt turbine speeds to d-c generator speeds, and such systems are rarely employed. Hybrid systems are expected to find increasing use where turbineelectric drive systems are advantageous. 1.4 Description of Direct-Current Drives. Directcurrent systems have been used on by far the greatest number and variety of installations. The most common d-c system consists of a multiple number of high-speed diesels driving direct-coupled generators, which in turn feed power to one or more motors connected directly to the propeller shaft or through a speed-reducing gear. The usual system involves one, two, or four enginegenerator sets arranged to drive one propeller. Twinscrew vessels usually consist of duplicate systems with various degrees of ability to transfer power from one side to the other, or to drive both screws at reduced power from the diesel-generator sets of one side. Directcurrent propulsion systems find application primarily in the-low and moderate power ranges, and where flexibility of setup and ease of control are of high value. The usual power size of a d-c drive system is from 1000 to 6000 hp per shaft. Exceptions can be cited on both sides of this range down to about 400 hp and up to the 10,500 hp per shaft of the icebreaker Glacier and the 19,600 hp of the center shaft of the nuclear-powered icebreaker Lenin. It should be noted that this latter power is developed by a double-armature motor, and thus each unit develops 9800 hp. The voltage selected for d-c systems is mainly a motor snd generator design consideration with a somewhat
Fig. 1
TWO-enginebasic aptem
arbitrary upper limit of 1000 volts. This limit is considered prudent for d-c machines that are to operate under shipboard conditions of vibration, high humidity, and frequent maneuvering, and has been established by the IEEE Committee on Marine Transportation [I].' Higher-powered systems benefit from the use of higher voltages and in the case of the very high-powered icebreaker Lettin, 1200volts was selected. More commonly, however, medium voltages in the range of 500 to 800 are appropriate. For any given power, a lowering of the voltage level requires a proportionate increase of the current, and higher currents require larger commutators, more brushes, larger cables, and higher-capacity contactors or swikhes. Where possible, therefore, voltages are selected so that required currents do not exoeed available handling devices, and 3000 amperes is a commonly observed maximum on all but very large systems. It is common practice in multiple-engine, d-c propulsion systems to employ a series loop system and to intersperse the motor armatures between the generator armatures so as to reduce the maximum voltage existing between parts of the system and to ground in the event that some point is a t ground potential. A two-engine system of this type is illustrated in Fig. 1. The important features of this electrical arrangement are: one, it utilizes full engine torque and power capability without subjecting the machines to overloads of any type, even when only one engine is in service; and two, it avoids problems as to governing, load divi$on, and shortcircuit protection which are inherent in systems that require parallel operation of independently driven d-c generators. If the generators were operated in parallel, the utilization of one engine a t full power would require overcapacity in the generator or the motor. This is because the power of a fixed-pitch propeller varies approximately as the cube of the rpm. Thus, a t 50-percent power, the rpm will be 79 percent, and at that rpm the motor will be able to absorb 79-percent volts. If the generator is to develop full power a t this lower voltage, overcurrent must be drawn so that the product of volta and current
1
Numbem in brackets -te
References at end of chapter.
336
MARINE ENGINEERING
-
PROPELLER RPM, % RATED Fig. 2
Propeller torque-rpm characteristics
is 100 percent. Alternatively, the motor could be provided with extra flux capacity so that it could accept the full voltage at the lower speed. In either case, overcapacity must be provided in order to permit full utilization of the one engine mode. It is not always possible or desirable to employ an equal number of generator and motor armatures. This brings with it some further complications which, in the extreme cases encountered on some high-powered systems, forces the use of parallel operating generators. If in the case just considered there were only one motor armature in the propulsion motor, the voltage of each generator could be only half that of the motor. This is not objectionable if the motor is not over about 4000 hp and of the direct-drive type, because 1000 volts could be used on the motor and each generator then rated at 500 volts. If, however, the ratio of generators to motors exceeds 2: 1 and becomes 3: 1 or 4: 1, a parallel system may become the better choice when consideration is given to the high ratio of voltage between the motor and its supplying generators [2]. The d-c electric drive system, in addition to providing the usual electric drive features, offers several important additional control and operating advantages which make it the preferred system for many applications. These advantages are as follows: (a) Erne of control. Basic control of a d-c propulsion drive is accomplished by varying the generator voltage through field control. This process is smooth, simple, efficient, and lends itself readily to remote control. (b) Multiple control stations. The simplicity of the process by which propeller speed and direction of rotation are controlled facilitates the use of multiple control stations when these are desirable from a ship-operating standpoint. Control can be delegated to any desired remote location by selector switches located at the engine room control station. The engine room station can retain or regain control at any time. (c) Adaptability to varying propeller-hull charaeter-
istics. A unique feature of the d-c electric drive is its ability to adapt to a range of propeller power-versus-rpm characteristics. The propeller power-rpm characteristic changes somewhat with hull condition and vessel loading, but these changes are not very large and can be accepted without major effects on the propelling machinery. However, vessels subject to the restraining effects of heavy tows or passage through ice cause their propellers to exhibit wide variations in power-rpm characteristics. A propeller on a stalled vessel (bollard condition) will develop full-power torque at approximately 70 percent of the free-route propeller rpm. Diesel engines, which are basically constant-torque machines, must run a t full speed to develop full power, and thus cannot by themselves be fully utilized over a variety of propeller power-speed characteristics. By proper selection of the d-c propelling motor and by utilizing field control, the constant-torque characteristic of a diesel engine can be adapted to a range of propeller characteristics. This is illustrated in Fig. 2' based on a two-engine installation as in Fig. 1. The adaptability here illustrated is obtained by selecting the propelling motor so that it can develop full power in the bollard condition and then weakening its field strength to match any other full-power propeller. speed up to that of the free-route condition. This permits the utilization of full rated engine capacity through a varying propeller characteristic speed range without installing excess engine or electrical capacity. In effect, the system provides a variable speed ratio between the prime movers and the propeller to suit a variety of operating conditions. 1.S Description of Alternating-Current Drives. The use of a-c electric drives is generally associated with the use of turbine prime movers. The usual system consists of a single turbine driving a direct-connected, high-speed generator which furnishes power to a single, low-speed, direct-drive motor of the synchronous type. A multiple number of turbine-generator sets can be used in this type of system if required by the power requirements and available turbine ratings. When high-speed diesels are used as prime movers in an a-c electric drive, multiple units will normally be required to develop the desired propulsion power. With the use of synchronous machines, there is a fixed ratio of speed between the turbine-generator and the propeller motor. This ratio is governed by the ratio of the number of poles on the motor to the number of poles on the generator. Thus a two-pole generator operating at 3600 rpm supplying power to a 60-pole synchronous motor would cause the motor to operate at SOof its own speed or 120 rpm. In effect, the a-c drive accomplishes the same result as the use of reduction gears, and it is useful to think of such generators and motors as being electrically "geared" together. I n addition, the electric drive provides the ability to reverse the propeller by electrical means rather than mechanical. Alternating-current drives are advantageous relative to direct-current drives from the standpoint of size, weight, cost, maintenance, and simplicity. However,
ELECTRIC PROPIULSlON DRIVES
these advantages are often outweighed by the control flexibility and superior torque performance of direct current for vessels of moderate power and special service. Alternating-current propulsion is particularly adaptable to ships requiring high power since both a-c generators and a-c motors can be readily built for powers of 50,000 hp or more if required. Alternating-current drive systems are most suitable for ships that spend a high proportion of their operating time a t or near full power rather than at severely reduced power or in service requiring frequent maneuvers. Electric drives of the a-c type tend to fall into the moderate and high-powered range. Thus, 10,000 hp per shaft might be considered as a reasonable lower limit and 60,000 hp per shaft as a likely upper range value. A notable exception is the very popular T2-SE-A1 tankers of World War 11, which were 6000 hp; it is doubtful, however, that such a low rating would again be applied in this form to oceangoing ships. At the high end of the scale, there is no practical limitation on rating for a-c propulsion. The voltage levels used on a-c propulsion systems range from about 2300 to 7500 volts, with the larger power units being associated with th'e higher voltages. The voltage is selected on the basis of motor and generator design considerations and on the basis of available switching devices. It was natural that the induction motor should be used on early installations because its torque performance afforded the censer-vatism necessary on an application where little was known about the actual torque requirements of a propeller under maneuvering conditions. Furthermore, the induction motor made practical the arrangement of pole-changing windings to obtain two different speed ratios between the prime mover and the propeller. This was important on warships where the cruising power requirements were low and where an alternate speed ratio contributed to better economy of prime mover operation. After experience was gained with the induction motor drives, the more desirable synchronous motor was applied with complete success. By providing the synchronous motor field with a pole face winding, so that it may be operated as a squirrelcage induction motor during maneuvering, it has been possible to provide characteristics which enable it to satisfactorily handle the maneuvering requirements of the propeller.' Therefore this type of motor is applied on practically all a-c propulsion systems. The advantages which make the use of the synchronous motor desirable as compared with the induction motor are: (1) better efficiency-a typical slow-speed synchronous propulsion motor has a full-load efficiency of 98 percent as compared to the 94 percent of a correspondinginduction motor; (2) higher power factor-the synchronous motor can be operated at 100 percent power factor as compared with the 70-75 percent of the slow-speed induction motor, resulting in a lower weight and cost of the generator; (3) lower weight and cost of the motor;
337
(4) larger air gap-this is conducive to more satisfactory installation and maintenance. The following features are associated with the alternating-current form of electric drive: (a) High eficiency. A reasonable average figure for the loss in the synchronous a-c electrical transmission between the prime mover and the propeller shaft is 6 percent. The overall transmission system efficiency must also reflect the excitation losses. (b) Flexibility of installation. The direct-drive motor can be located aft in the vessel to reduce t4e shaft length, and the generator sets can be installed at any location desired. (c) Use of unidirectional prime movers. Since revereing of the propeller is accomplished by electrical means, the prime movers can be unidirectional units. Thus if steam turbines are applied, no reversing stages are required and the valving is simplified. Gas turbines are inherently unidirectional machines, and some externd means must be provided for obtaining reversed propeller thrust. (d) Multiple prime movers. Any number of prime movers can be combined (paralleled) electkally to provide power to the propulsion motor. (e) Dual use of propulsion power. The propulsion generator set(s) is available for other functions when not being used for propulion. (f) Speed reduction--prime mover to propeller shaft. By choosing a motor with a large number of poles as compared to the generator, a speed reduction is effected electrically without the use of mechanical meaiw. Almost all turbine a-c electric drives use generators directly coupled to the prime mover and motors driving directly to the propeller. The ratio of poles on the motor to poles on the generator will usually be in the range of 30 or 35 to 1, producing a speed ratio that is the inverse ratio of the number of poles. (g) Available in large power ratings. Alternatingcurrent machines can be built in very large ratings. From a practical standpoint, they can be built with any rating required for ship propulsion applications. 1.6 The Alternating-Current Direct-Current System. An a-c rectified d-c drive, by utilizing a high-capacity silicon rectifier, makes possible the use of high-speed a-c generator sets to provide power to d-c propulsion motors. The power size of this hybrid system is now limited by the magimum d-c motor size or to about 15,000 hp per motor armature, although a higher propulsion shaft horsepower can be obtained by the use of multiplearmature d-c motors as illustrated in Fig. 3. This system retains the favorable speed-torque and control characteristics of the variable voltage or conventional d-c system and permits the use of large high-speed prime movers with direct-connected generators. This system has been applied to many industrial applications with good success and it is considered a suitable and reliable transmission system for marine propulsion. High-speed locomotive-type diesel engines in the 4000-hp range commonly employ direct-connected
338
MARINE EN
the kva taken from the generator, and the actual kva rating of the a-c propulsion generator should be somewhat higher to compensate for a-c voltage distortion (harmonics). For a 6-phase rectifier, the displacement power factor should be multiplied by approximately 0.96 to obtain the required generator volt ampere rating. For a 12-phase rectifier, the displacement power factor should be multiplied by approximately 0.99 to obtain the generator volt-ampere rating. The switching action of the'rectifiers in commutating the d-c load causes some distortion in the a-c wave shape, and the d-c voltage wave contains a ripple component superimposed on the average d-c voltage. Excessive ripple voltage can result in significant extra motor heating and to deterioration of commutation performance. However, the total rms d-c voltage ripple produced in a 6-phase rectifier is about 8 percent, and in a 12-phase unit about 5 percent. The extra hating produced by such ripple voltages is negligible and no Fig. 3 Two-generator, single-shaft, a-c reZtiRd d-c system special motor design is necessary from this standpoint. For good d-c motor commutation, the current ripple should be limited to a maximum of 2 percent and, in most cases, normal d-c motor inductance is sufficient to give satisfactory commutation with 6-phase and 12a-c generators and rectify this power to d-c for use in the phase rectifiers. In special cases, or when less than driving motors. Much larger units are employed as 6-phase units are used, some de-rating, atering, or other sources of excitation for very large electric utility compensations may be necessary. The speed of the d-c propulsion motor is controlled generators. The system voltage is selected with the same criteria in the same manner as the variable-voltage d-c system; as the conventional d-c system. Some additional that is, by varying the magnitude of the voltage supplied flexibility is possible in the selection of the a-c generator to the propulsion motor armature. This is accomplished voltage by using a transformer between the a-c generator by varying the a-c generator excitation or by a combinaand rectifier. This arrangement may be necessary in tion of a-c generator excitation and generator set speed very large ratings where a single turbine-generator control. The direction of motor-propeller rotation is supplies power to a double or triple-armature d-c motor; changed by reversing the polarity of the propulsion however, for most applications the a-c generator voltage motor field. Unlike the conventional d-c system, the rectifier blocks can be selected so that power can be supplied directly from the a-c generator to the rectifier to the d-c motor. the transfer of power from the propulsion motor to the The rated voltages of the a-c and d-c units, when using a-c generator set. Each drive system must, therefore, 6-phase and 12-phase systems, will be in the ratio of 1.0 be checked to assure that the pump-back energy during to 1.3, and if the d-c motor is limited to 1000 volts, the a propeller reversal can be absorbed in the motor maximum a-c voltage is approximately 775 volts, armature-rectifier circuit. If necessary, a resistor can be switched into the motor armature circuit to absorb line-to-line. The power factor of the rectifier will be less than unity this pump-back energy and limit excessive d-c armature because of the harmonic components (distortion factor) current. The losses in the a-c rectified d-c electric transmission and because of the generator reactance, transformer between the prime mover and the propeller are less than reactance, if used, and the d-c circuit inductance (displacement power factor). The displacement power in the d-c system because of the higher efficiency of the factor for 6 or 12-phase rectifiers connected directly to high-speed a-c generator. A reasonable figure for the the a-c generator will be between 0.91 and 0.96, depend- loss in transmission systems where transformers are not ing upon the generator reactance. The displacement required is 12 percent, and with systems requiring power factor gives only the fundamental component of rectifier transformers a reasonable figure is 14 percent.
ELECTRIC PROPULSION DRIVES
339
Section 2 The Diesel Direct-Current Drive System 2.1 System Composition. A diesel d-c type of electric drive permits the use of several small, high-speed; lightweight, nonreversible engines to drive a single propeller. The system frequently consists of a pair of engine-driven generators and a single propulsion motor but in some cases, double-armature motors, or two motors driving through reduction gears, are used. The use of high-speed engines results in a considerable reduction in size and weight of machinery, and it is desirable in every case to use engines with rotative speeds as high as is consistent with reasonable maintenance and life. To reduce wear on the engines, it is usual practice to operate at reduced engine speeds when less than full power is required. Thus an engine having a full speed of 900 rpm will commonly have a babe speed of about 450 rpm and be operated at that speed until torque and power demands require raising its speed. Naturally aspirated diesel engines are basically constant-torque machines and are therefore capable of developing power proportional to speed over their operating range. However, prudent operation dictates some reduction in torque at reduced speeds and, therefore, system operation is arranged for reductions in both torque and speed whenever less than full power is required. Supercharged engines are more restricted at reduced speed, but this characteristic is in no practical way detrimental to their marine application in a d-c system where the engine speed is independent of the propeller speed. Generators are normally of the shunt-wound type, separately excited and arranged for direct connection to the engine. In small plants the generators are usually arranged for self-ventilation with a fan attached to the armature, the ventilating air being taken from the engine room and exhausted back to the engine room or through a duct to the outside. In large installations, where the heat loss is considerable, generators are usually fitted with a closed ventilating system, with the shaftmounted fan forcing the air through a water-cooled air cooler. In this case the machine is totally enclosed and is independent of the environmental conditions of the engine room. Propulsion motors are usually of the same type and general construction as the main generatora. For some installations, where the diameter of the motor is restricted, double-armature motors may be used. Such a motor consists of two duplicate units usually mounted on a common shaft with two supporting bearings and two separate field frames. Single-armature motors are lower in volume, weight, and cost than double-armature units and should be used whenever conditions permit. For applications involving very low propeller speeds and rather small plants, it is sometimes desirable to use one or two high-speed motors connected to the propeller shaft through a reduction gear. The excitation subsystem must be capable of providing
controllable amounts of excitation power for the separately excited main motors and generators so that changes in propeller speed and direction of rotation can be accomplished through variable voltage and polarity control of the generators. The amount of excitation power is small, being of the order of 1.5 to 2 percent of the machine rating, but its continuity and ease of control are vital to the system. In many installations, particularly those of small and moderate-size plants, the power for excitation is supplied by exciters driven by the main engines so that the propulsion plant is self-contained and independent of any auxiliary engines. A further extension of this concept of self-sufficiency that has been used in many small installations involves the use of main engine-driven generator exciters and a small auxiliary generator that operates at constant potential and is capable of supplying underway auxiliary power as well as motor excitation. The full range of main generator voltage and polarity control is achieved by varying the very low power of the field of the generator exciter. This use of double amplification permits the use of small rheostats and switches in the control stations and small control cables. Main engine-driven exciters or auxiliary generators must be capable of developing the required power over the semice speed range,of the main engines.. If an auxiliary generator is so driven, it must be controlled by a voltage regulator to hold constant voltage. With larger plants, excitation power is derived from the vital bus of the ship's semice power system. The dependence placed upon the ship's service power system for a variety of vital services has brought about the 'inclusion of features and redundancies which assure continuity of auxiliary power under virtually all circumstances. Its use, therefore, as the basic source of excitation power is not inconsistent with the importance of maintaining propulsion power control [3]. Since most vessels, except the very smallest, use a-c ship's service systems, conversion to d-c and preferably controllable-voltage d-c is necessary. This is often done with motor-generator sets but, in recent years, increasinp; use is being made of static conversion devices. The development of silicon-controlled rectifiers in larger sizes permits the supply of easily controlled excitation power to generator and motor fields. These excitation systems lend themselves readily to the addition of current regulating, limiting, and other protective features. A spare excitation conversion unit is normally installed to safeguard this vital service. On twin-screw vessels, a single spare is considered adequate for the two propulsion systems. The control of a diesel d-c electric system is accomplished basically by varying the voltage and polarity of the main generators. This voltage variation is effected by a combination of generator speed and excitation
MARINE ENGINEERING
ELECTRIC PROPULSION DRIVES SPWJAL
"i i
SERVICES DC BUS
MTR FLD
CENTRAL OPERATINO qATION Fig. 4
Schemalc diagram, two-engine d-c electric ryrtem
changes so as to produce a full range of motor speed control in either direction of rotation. Motor excitation is normally held at a constant value for each operating mode, and changes are necessary only when the number of generators in service is changed or when the propeller speed-rpm characteristics are changed (e.g., when towing). A basic schematic diagram for a two-engine generator, two-motor system is shown in Fig. 4. This schematic shows the d-c loop with power switches 1s and 2s in the normal operating position. These "setup" switches can be either manually or electrically operated, and no "load-break" capability is required since protective devices operate to cause the loop current to go to zero, or very nearly zero, before the main contacts of the setup switch can be opened. On switches 1s and 2S, all three sets of contacts are mechanicallv interlocked. If it is necessary for maintenance or operating economy to remove generator GI, for example, switch 18 is operated so that contacts 3-4 close when 1-2 and
STOP, one of the generator field contactors, AH (ahead) or AS (astern), will close to supply d-c power through the silicon-controlled rectifiers (SCR's) to the generator fields and full field current would be provided to the motor field by the closing of contactor FW. Generator field voltage is increased with throttle movement off STOP by gradually gating the SCR's fully on to provide full field at about half throttle while the engine speed remains near idle. During the remainder of throttle travel, engine speed is increased by controlling the fuel rack through a current&-air-pressure transducer located on the engine. Thus, over the latter part of throttle movement, armature voltage varies directly with engine speed. Tracking of the generator excitation and engine speed control is assured by having the controlling rheostats for both engine-generator sets on the same o h.-.$+ 0ua1U.
~h~ loop ammeters on the pilothouse (PH) console and on the C0S control panel provide a continuoue display of the operating condition of the propuls~on ~ ~ n ~ ~ e ~ r o $ ~ ~ ~ e ~ ~ ~ ~ ~ system ~ ~ for ~ ~both c the ~ ~helmsman a l & and . ~the~COS e fpersonnel. , " c ' ~ ~ ~ Loop current and propeller speed are indicators of loadarrangement whenever switch 1s is positioned to bypass ing of the electrical system and of the power being this generator. -. . . to the propeller. The loop protected from D h n g a s t a h u p of the system, the throttle would be excessive current by two devices. The first of these is the in the STOP position and the switches Control would be in the central operating station (COS), thermal overload alarm, set at 110-115 percent of normal the diesels would run at idle meed. and the field current as a warning to the operator of an abnormal -.motor . weakening contactor FW wo;ld be open to prevent operating condition requiring his corrective action. The overheating of the motor field. After starting the diesels, second is the magnetic overload trip, which is set for control could be transferred by the selector switch in the 200-300 percent of rated loop current and removes COS to the pilothouse, if desired, and the system brought motor and generator excitation. This trip is fund* up to a higher speed. As soon as the throttle is moved off mentally a short-circuit and back-up protective device -
,
and is purposely set above any normal operating peak condition. When operating with. one generator and two motor armatures in the system, the motor field must be reduced to 63 percent of normal in order for the propeller to absorb all the available power. Field current is reduced by manual operation of the motor field rheostats. The special services bus shown in Fig. 4 can be a load bus, such as is often found on drilling rigs or dredges, or it can be a separate source for d-c propulsion power. When propulsion power is not required, switch 1T or 2T is closed and the generators are then used to supply d-c power to other equipment. Additional circuitry and instrumentation will be required for such service, and it is possible to operate in this mode a t voltage lower than in the propulsion mode if desired. Cables for inteiconnection of the main generators, motors and setup switches are usually varnished cloth or asbestos-varnished-cloth insulated, and of the leaded and armored type or the impervious sheathed and armored type. Because of limited space usually available for making bends, the individual single-conductor cables should not be larger than 1,000,000 circular mils, and a suflicient number of such cables or smaller cables should be connected in parallel for each lead. It is desirable to run cables as directly as possible between units and in well-ventilated and accessible locations. Cables should be continuous betweeh terminations with no splices, and be well supported in metal racks. Where propulsion motors or generators require forced ventilation by separate blowers, these are generally supplied and mounted on the main units. When such blowers are separately mounted, they are connected to the machines by ducts supplied by the shipbuilder. Circulating water for machines fitted with air coolers is normally supplied by motor-driven pumps. Propulsion motor bearings on small, low-speed units may be the self-contained disk lubricated type and require no external cooling. Larger motors, however, will use force-lubricated bearings supplied by a lubricating oil system consisting of a sump tank, motor-driven oil pumps, and oil cooler. The propeller thrust bearing is usually supplied from the same system. Propulsion generator bearings are normally forcelubricated from the diesel engine system, the bearings being provided with necessary inlet orifices to limit the flow of oil to proper values. 2.2 System Design Features. I n laying out a suitable diesel-electric d-c propulsion system, careful consideration must be given to a number of items. The first step is to select a suitable type and size of engine. This will, of course, depend on the type and service of the vessel involved and the power requirement. I n estimating the total engine brake horsepower required, it is satisfactory to use a value of 85 percent for the transmission efficiency. This allows for the losses of the generators, motors, and cabling. Thus for a system requiring 4000 propeller shaft horsepower, the engine brake horsepower should total 4700. This could be
34 1
supplied by four 1175-hp engines, but a more likely choice would be two 2350-hp units. If the engines are to drive exciters or auxiliary generators, the engine brake horsepower should be increased accordingly. I n order to minimize the size and cost of the currentcarrying elements of the main system, it is desirable to use moderately high voltages on the generators and motors. The desirable voltage will be influenced by the ratings of the machines and the ratio of the number of generators to the number of motors when the usual series loop connection is used. IEEE marine rules establish 1000 volts as a maximum for a single armature, and if this voltage is used on a single-armature motor, the permissible generator voltage will be 500 if two generators are employed and 250 if four units are used. I n systems that employ an equal number of motor and 'generator armatures, there is considerable freedom of choice since, by electrically interspersing generators and motors, the maximum system voltage is no higher than the voltage of a single unit. Although from a system standpoint 1000 volts per unit could be used, practical machine design and operating conditions will tend to reduce this voltage to the 500-750 volt range on all but the very largest installations. If the installation is such that the main generator sets are to be used in port to supply power for special auxiliaries, their voltage demands may influence the system voltage choice. This influence would be toward a lower voltage but the separately excited main generators can be operated a t reduced voltage, rkduced speed, or both, and still supply special auxiliary power, provided its current rating is not exceeded and provided the engine is not overloaded a t the chosen operating speed. Generators and motors of the shunt-wound type arranged for separate excitation are the general standard. A series loop system of such machines is inherently stable, even under the weak-field motor conditions encountered when the number of in-service engine-generator sets is reduced to a minimum. Good load-sharing between paralleled d-c generators requires that their terminal voltage decrease with increasing load. A drooping voltage regulation characteristic of at least 5 percent (and preferably more) should be provided to insure good parallel operation. Two or more motors operating in mechanical parallel and supplied in parallel from a common power bus must have drodping speed versus load characteristics to share load without depending on an excitation-regulating control system. Inherent load division between motors is entirely dependent on their load-speed characteristics; these must droop for stable operation and they must be identical throughout the range of operating conditions for equal load sharing. The availabilitv of materials such as silicones and epoxy resins has resulted in the appearance of classes of insulation that permit high safe temperature rises. Class A insulation is seldom used on propulsion machines and virtually all units use Class B or Class F insulation systems. Class H insulation is occasionally employed,
but care must be taken to avoid adverse brush-wear problems if silicones are used. With a 50 C ambient temperature, the allowable temperature rises for Class B, F, and H insulation are 60, 80, and 100 deg C, respectively. More complete tabulations will be found in reference [I]. The use of the higher temperature insulations results in smaller and lighter weight machines for a given power. They are also better &ted to the marine environment since they utilize practically no cellulose materials. Totally enclosed machines equipped with surface air coolers create their own ambient conditions and are affected only to a very minor degree by the ambient air temperature of the room in which they are located. I n some cases it may be desirable to provide coolers that will deliver air to the machines a t 40 C rather than 50 C and allow a corkesponding increase in the allowable temperature rise. The maximum seawater temperature is considered to be 85 F. Propulsion generators and motors are normally rated for the full-power continuous shaft horsepower requirement of the vessel, and no overload ratings are provided. This practice is consistent with the fact that the diesel engine usually has a maximum rating with no overload capacity. However, in designing the electrical equipment, it is necessary to take into account the fact that current ' peaks are encountered during maneuvering. These peaks are of short duration and their magnitudes, with well-designed excitation and control systems, can be kept to values within the inherent commutating ability of the main units. Their effect on heating is not appreciable. For small and medium-sized propulsion plants, it is generally satisfactory to use motors and generators of the protected, self-ventilated type and provide adequate room ventilation. For large installations, where heat loss from the electrical machines is considerable, or where engine room or motor room space is limited, the units should be ~rovided with a forced-ventilation system or with air coolers and closed systems. While the totally enclosed machines with individual coolers are somewhat less accessible and somewhat more costly, they lend themselves to a simpler ship installation layout and eliminate the need for large ducts, outside ventilators, and motor-driven blowers, at least for the high-speed generators. Low-speed, direct-connected motors will still require motor-driven blowers, but no ducts to the outside are necessary. Ventilating systems should be laid out on the basis of circulating 70 to 100 cfm of air per kw of loss in the machine. Pressure drops in machines of this type are of the order of 1.5 in. of water. Main propulsion motors and generators should be fitted with 0; arranged for connection to fire extinguishing systems of t h e COz type. On other than closed recirculating air systems, means should be provided to shut off the-circulation of air when the fire extinguishing system is used.
From the standpoint of minimum cost and weight, a single-armature motor should be used where space permits. A double-armature motor can be built in a smaller diameter and may be used where the diameter is limiting or where the added reliability of two separate electrical units is considered desirable. Another arrangement, which may be used where space and weight limitations are severe, consists of one or more high-speed motors connected to the propeller shaft through reduction gears. While this system saves weight and space for propeller speeds below approximately 150 rpm, it increases the complications of the installation and involves additional high-speed, precision equipment. When direct-connected propulsion motors are used, it is common practice to include the main thrust bearing as a part of one of the motor's main bearings. This can be done a t either end of the motor, but the use of a thrust bearing at the forward or free end of the motor contributes to a smaller size thrust collar and one which can be removed readily from the end of the motor shaft. When geared motors are used, the thrust bearing will ordinarily be located in the reduction gear. The propulsion control normally should be mounted on a platform in the upper part of the engine room, preferably so that the operator has a view of the engine room space. This arrangement makes a simple cable layout with the cables coming out of the top of the various machines and into the bottom of the control board. It also avoids dangers associated with cables entering the machines from below in close proximity to the bilges. For pilothouse control, a simple desk incorporating the necessary master controller instruments and indicators is sufficient for propulsion purposes. On every installation, it is necessary to check the engine-generator combination to make certain that there will be no serious torsional vibration within the operating speed range of the unit. This is a responsibility of the engine builder and requires a careful calculation of the mass-elastic system and its response to engine excitation frequencies. I t is also desirable to make a careful calculation of the mass-elastic system comprising the propulsion motor, the lime shaft, and the propeller to make certain that there will be no critical speeds induced by the propeller in the operating range. This is normally a responsibility of the shipbuilder. The d-c electric system does not contribute to any vibratory excitation since the d-c power flow is smooth and continuous, and the torque produced is virtually without variation. 2.3 Physical Characteristics. With the high-speed diesel engine, the generating set can be fitted readily into any reasonable engine room layout. Parts are relatively small and easy to handle, and spaces necessary for disassembly and maintenance are considerably less than for large, direct-connected engines. Foundations are small and simple. As examples of the size of sets, typical engine-generator units would have approximate overall dimensions as follows :
343
ELECTRIC PROPULSION DRIVES
MARINE ENGINEERING
140 0
t
cn
2120
40
/
',
BASED ON 900 RPM DIESEL ENGINES.700 VOLT GENERATORS SINGLE-.ARMATURE MOTOR FOR DIRECT DRI& AND TWO MOTORS FOR GEARED DRIVE
SHAFT HORSEPWER-THOUSANDS Fig. 6
~~ do0
A00 .dm sd00 SHAFT HORSEPOWER
Weight of diesel d-c electric drive equipment at various power ratings
dm dm
The curves in Fig. 6 show, for several propeller speeds, the approximate weight in pounds per shp of a singlescrew propulsion plant for ratings up to 10,000 shp. Flg. 5 Diameter of d-c propulsion koton for varlous ratingr and speeds These weights include engines, generators, motor, and control. They do not include cable or auxiliaries. These latter items vary considerably with different layouts but represent a very small percentage of the weight of the main machinery, and therefore are of minor Width Rating Length Height (generator) importance. For a propeller speed of 100 rpm, curves are given for both direct-connected and geared motors. 12 f t 2000 kw, 900 rpm 29 f t 7.5 ft A comparison of these curves shows the saving in weight 1000 kw, 900 rpm 20 ft 12 ft 6 ft with the geared drive for low propeller speeds. O :
lob0
2.4
Applications
These dimensions will, of course, vary somewhat with (a) Oceanographic Survey Ships. The oceanographic different makes of engines. survey ship Wilkes (AGS-33) is one of a group of specialEither single-armature or double-armature propulsion service vessels using diesel d-c electric propulsion. This motors can be used as dictated by the machinery rating ship has been designed to be highly maneuverable in and space. The motor diameter is usually the limiting order to fulfill the wide variety of oceanographic, dimension; Fig. 5 shows, for several propeller speeds, the hydrographic, and other research tasks for which it is approximate diameters of single-armature motors of intended. The ship has a single screw rated a t 3000 shp various powers. Multiple armatures can be used to and 200 rpm. The propulsion plant consists of the folobtain higher power. These curves are for Class F lowing major components: insulated machines. Main Propulsion Engines: Propulsion control cubicles require relatively little space and usually can be easily located on a flat above the Tu$o, 1800-bhp, 1000-rpm units main generators or at one end of the engine room. On Main Propulsion Generators: single generator plants, a desk-type control station can Two 1260-kw, 1000-rpm, 700-V units. be used since power leads are not brought to the control Variable voltage field excitation (7.2 kw, 0 to cubicle. As examples of the size of control boards, the 180 V) from silicon-controlled rectifier following approximate dimensions are given for single(SCR) in main propulsion control cubicles. screw drives: Main Propulsion Motor: -No. of One double-armature, 3000-hp (1500 hp per Sbp Engines Width Depth Height armature). 200-r~m.700-V unit. Constant5000 2 8ft 5 ft 8ft voltage fihd kxLit*tioi (2.5 kw, 250 V) 1000 1 5 ft 3 ft 3 'ft (desk) from SCR in propulsion control cubicles.
(
MARINE i ENGINEERING GI = G2 = 1 2 6 0 KW IM)OAMPS. 7OOVOLTS
,$,.
;rl PHASE
A.
ELECTRIC PROPULSION DRIVES
C
4 6 0 V. I PH. 6OCPS ROOM THROTTLE
DRIVEN 300 K W DC GEN FIELD
1 0-60CPS TRANSF
I' 460 VOLTS I PH 6OCPS
(PERMISSIVE)
Jig. 8
Cl
C2
- LOOP CONTACT0
MANUAL SETUP SWITCHES - - - - - INDICATES MECHANICAL CONNECTION SI 8 S 2
Fig. 7
1
/
4 6 0 V 3PH 6OCPS C
Elechical schematic for d-c propulsion loop and bow thruster
The auxiliary propulsion and navigation equipment consists of a bow thruster driven by a 350-hp d-c motor and a propulsive rudder driven by a 250-hp a-c motor. Either of two 300-kw d-c generators, one diesel-driven and one gas-turbine-driven, can supply power to the bow thruster. Figure 7 is the basic schematic arrangement of the electrical circuits. As will be noted from a study of Fig. 7, considerable redundancy is built into the d-c loop circuit. With two generators and two motors operating, full power at 200 rprn will be realized. Switching to operate with only one generator and the two motor armatures is easily done by operation of the contactor-type setup switches. If it is desired to remove either of the main propulsion motor armatures from service, it is necessary to transfer the links to the bypass position. This is intended only as an emergency measure in case of damage to a unit, and all normal operations are performed with both motor units in service. Also the gas-turbine-driven, d-c generator can be switched into the loop in the event that it is desirable to have the diesels shut down. The primary propulsion control equipment is located in the central operating station (COS) and it can override either of the two above-deck stations; i.e., the pilothouse or the aft control station. This vessel includes a machinery control console which performs display, alarm, and data logging functions for most of the machinery on the ship. Starting and stopping of the main propulsion diesels are the only actual propulsion control functions on this console.
A common throttle handle is used for engine speed and generator field excitation. Duplicate rheostats are located at the pilothouse and at the engine room consoles. Engine speed is kept at approximately 400 rpm through one quarter of the throttle travel while generator excitation is gradually increased to 100 percent. Beyond one-quarter throttle, the engine speed is increased linearly to full throttle. The generator fields are controlled by varying the firing time of the silicon-controlled rectifiers, and the engine speed is controlled by varying the current into an electric-pneumatic transducer which controls the fuel rack position. The throttle control includes a one-way tracking d-c servo system whereby the engine room throttle follows the pilothouse (PH) throttle. The one-way tracking feature permits central operating station (COS) takeover without consultation with the PH. Transfer from the COS, however, can be accomplished only upon operation of a permissive circuit and matching of the two controllers. Voice communication and observation of the COS throttle remote position indicator on the PH console are necessary to effect transfer to the PH. Local control of engine speed is also available, with two air pressure gages on the engine panel to allow "bumpfree" transfer from the COS control panel. In the operation of a d-c loop with two independent prime movers, loss of power from one prime mover, when both are in service, can result in reversing the prime mover which has lost power and may cause severe engine damage. To avoid this, a system of speed-
Generator ekcitation circuitry
activated interlocks is utilized.' The field circuit of each generator can be energized only when two sets of speedsensing contacts are closed. One set operates off a voltage-sensitive relay operated by a tachometer generator, and the other operates from a mechanical input. Two separate takeoff points on each engine assure independent operation. Both sets of contacts are in series, and close at about 300 rpm on increasing speed and open at about 200 rpm on decreasing speed. If both sets do not close on the way up, no field excitation can be applied. Upon loss of power, as long as either set of contacts opens, the field circuit will be opened, thus preventing the generator from motoring and driving the engine backwards. The probability of loss of the prime mover and failure of both sets of contacts to open is extremely remote. Should one set fail, the second will protect the unit and still not interfere with normal operation. If one engine loses power, the second unit will be able to carry the load since the motor voltage will be reduced by 50 percent and the speed will drop proportionally. Thus the operator will be able to continue operation at half-speed until convenient to change the setup contactors. I n addition to the normal indicators on the PH console, a loop ammeter is included. This is a direct indication of power into the propeller, and enables the helmsman to operate at full load on the propulsion plant without constant PH-to-COS communication. The system is protected from overspeed trip-out and from excessive loop current during a quick reversal operation (defined as immediate FULL AHEAD to FULL ASTERN or conversely, although the latter tends to be less severe due to the relatively low astern ship speeds) by a time-delay relay activated as the throttle goes through the STOP position. This reduces engine speed for a few seconds, removes the generator field excitation, and inserts a field-weakening resistor into the motor field. This time delay, plus the inherent machine time constants, allows the ship to slow down to approximately 70 percent speed before applying full generator field in the reverse direction, neutralizes the effect of the acceleration
torque transmitted to the diesels due to the waterwheel action of the propeller, and serves to limit the loop current to less than the instantaneous magnetic trip setting of 200 percent of normal. This magnetic protective device, which performs strictly a backup function, has an automatic reset feature to permit resumption of normal operation as soon as the throttle is returned to the STOP position. Silicon-controlled rectifiers are used for the variable voltage power supplies to the generator fields. The reference signal to the rectifiers is varied from 0 to 15 volts DC by the throttle movement over the first quarter of travel. Reversing is accomplished by contactors which change the polarity applied to the field connectors. Fixed output-voltage silicon rectifiers are supplied for the main motor field excitation, and automatic field weakening of the main motor is provided in the STOP position of the throttle. Each solid-state rectifier is fed from the 440-V a-c ship's service system by circuits as shown in Figs. 7 and 8. Duplicate breakers, transformers, and silicon rectifiers are supplied for the main propulsion circuits. The field circuits of the generators are arranged for suiciding (residual armature voltage applied in reverse across the field in the STOP position). Once again care has been taken to avoid feedback between the armature and field excitation circuits by using NO and NC contacts on the same contactor for the two functions, as shown ifi Fig. 8. The 350-hp d-c bow thruster operates over a wide speed range by varying the excitation to the SCR from 0 to 15 volts from the pilothouse throttle rheostat, thus varying the generator field voltage from 0 to 180 volts. The bow thruster motor field is supplied with 250 V d-c from a silicon rectifier. The reversal of thrust from port to starboard is accomplished by reversing the voltage applied to the generator field, after a short time delay to avoid excessive plugging. The a-c propulsive rudder circuit is designed to start the motor at % voltage and % frequency. The unit system approach-one 300-kw generator and one 250-hp
346
MARINE ENGINEERING
ELECTRIC PROPULSION DRIVES
4-hour maximum power a t 810 rpm. The main motor is capable of developing its power over a speed range, and a t its maximum 4hour rating this range is from 120 to 175 rpm. These propeller speeds correspond to the bollard and free-route conditions of the vessel. The main propulsion control station for each of the two propulsion system is located in the corresponding motor room. This station is designed to
.
(1) Select any combination of one to five dieselgenerator sets with which to supply power to. the associated main motor. (2) Supply excitation power to the motor and the generators from the assigned motor-generator set or from the standby unit, if necessary. (3) Start, stop, and regulate the speed of the propulsion motor from zero speed to maximum speed in either direction of rotation. (4) Provide means for dividing the load equally between generators. (5) Provide visual indications of machinery performance and alarms for dangerous operating conditions. (6) Provide for transfer of main motor speed and direction control to the pilothouse.
The pilothouse control station is designed to provide for remote operation of the speed controls for both Fig. 9 Simpllfled schematic d i a g r a d l a c i e r motors. This station also is mechanically connected to two bridge-wing control pedestals so that propeller control can be accomplished from these outside positions when in closely confined conditions. The pilothouse motor-allows this reduction in voltage and frequency control function can also be transferred to the aloft without a reduced voltage starter, since the voltage sags conning station when it is desired to handle the propellers as soon as applied. The special constant volts-per-cycle from that higher vantage point. A basic schematic regulator helps this operation, and also permits easy diagram of this system is shown in Fig. 9. operation from % to full speed by varying the prime The electric-drive propulsion system of an icebreaker mover speed by means of a throttle control lever operat- is designed so that full use can be made of its power ing a rheostat from either the PH or aft control throughout the range of propeller speeds encountered stations. between free-route and bollard conditions of the vessel. (b) USCG Icebreaker Glacier.The high-powered The propeller power-speed characteristics for this vessel, USCG icebreaker USS Glacier is an interesting example together with key power plant data, are shown in Fig. 10. of a very large diesel d-c electric system capable of Fundamentally, control of propeller speed is obtained developing a total of 21,000 hp on its twin screws over a by just one variable, generator voltage. Of course, speed range of 120 to 175 rpm. This is by no means a adjustment of the motor shunt field will also affect the typical diesel d-c electric drive system but, because of propeller speed but adjustment of the motor field is to be its size and special features, it merits consideration as an made only for the purpose of compensating for a new illustration of the application of this type of system to propeller power-speed characteristic. For any one the demanding duty and rugged conditions of polar condition of operation, the motor field should be held icebreaking service. constant and the motor speed changed by varying the The ship is twin, screw with each system consisting of bus voltage. Even the direction of rotation of the the following principal elements: propeller is controlled by generator voltage, since the motors are reversed by reversing the generator polarity. 5 diesel engines-1920 hp cont.; 2400 hp for 4 hr The generator voltage on the Glacier is controlled by 5 d-c generato-1340 kw cont. ; 1700 kw for 4 hr changing the engine speed and changing the generator 1d-c prop. motor-8450 hp cont. ; 10,500 hp for 4 hr field strength. Both of these functions are controlled 1 excitation and control subsystem from a master speed controller handwheel on the motor room control board. All of the generators driving one It should be noted that the diesel-generator sets propulsion motor are connected in parallel through develop their rated continuous power a t 720 rpm and the electrically operated circuit breakers. These breakers
K W
3
B W
ln K
0
I C
u
a I
ln
I
0 1 0
I 0
64
80 100 I20 PROPELLER RPM 1 1 1 1 1 1 1 1 1 100 200 300 400 500 600 700 800 900 MOTOR VOLTS (FULL FIELD ) I 1 1 1 1 1 1 1 1 1 1 1 1 100 200 300 400 500 600 MOTOR VOLTS (WEAK FIELD) 40
140
160
1 1 700
1 1 800
180
1 900
Fig. 10 Propeller power requirem>nh-+lacier
will trip on overcurrent, short circuit, or engine overspeed, and they can also be tripped manually by either the engine room or motor room operator. The overcurrent trip is set for currents higher than any peaks encountered in normal service, and protection against moderate current overload is not necessary since the engines cannot exceed their maximum rating long enough to seriously overload a generator. At low propeller speeds the engines are run a t their idling speed of 300 rpm and the generator voltage is controlled by varying the generator excitation. This is accomplished by means of a potentiometer rheostat connected in the field of the generator exciter which provides excitation for all generators on the same bus. This rheostat is connected mechanically to the governor control cam and both are driven by the speed control handwheel. A pilot motor driving the handwheel makes
it possible to operate the speed control from any desired remote location. Remote control of the propulsion motor speed consists basically of a system for remotely operating the master speed controller on the motor room control board. The master speed controller in the motor room is the device that actually controls propulsion speed a t all times. Regardless of where the remote control may be originating, the motor room operator can take over control a t any time merely by disconnecting the remote control signal. Rather comprehensive tests were made during the trials of the Glacierto show the performance of the propulsion system. Oscillographic records were made during a series of maneuvering operations, and the data are summarized and reviewed in reference [4]. This interesting art5cle reveals a great deal about the operation of this well-engineered, -high-powered electric-drive system.
Section 3 The Turbine Direct-Curre~~tDrive System 3.1 System Composition. A turbine d-c drive system usually comprises a single, high-speed, nonreversing steam turbine driving a propulsion generator through a reduction gear, a propulsion motor, a control
system, connecting cable, and various auxiliaries. The turbine-generator set may be operated a t constant speed, and it may be convenient in such a case to add a ship's service generator to the same set.
MARINE ENGINEERING 11 0 0 1000 900
800 700 600
d
W
500
W
%
400
a
0
5
300
W 2
5 0 0 TO 1 0 0 0 VOLTS
1000
Fig. 11
.
2000 3000 4000 6000 GENERATOR RATING. KILOWATTS
10,000
Maximum d-c generator speed venur rating
The turbine-generator set requires the use of a reduction gear to reduce the inherently high speed of the turbine to a speed that is acceptable to the d-c generator. I n some cases it is considered advantageous t o utilize twin- or double-armature generators so as to take advantage of the higher speed a t which the smaller double unit can be run. Figure 11 shows the relationship between kw rating and maximum speed a t which normal d-c generators can be operated with reasonable design, performance, and maintenance. 3.2 Differences from Diesel Electric System. From an electrical standpoint, the operation of a turbine d-c electric drive is basically the same as that of the diesel system. Speed and direction of rotation of the propeller motor are controlled by varying the magnitude and polarity of the propulsion generator voltage through
The system differs in that it is usual to employ only one turbine per screw since it is'more economical to build a single turbine of full rating than a number of smaller ones. It is also possible and often desirable to operate the turbine-generator set a t constant speed and use the same turbine to drive a ship's service generator as previously noted. On high-speed diesel-electric systems, it is highly desirable from an engine-operation standpoint to reduce the engine speed whenever less than maximum power is required. 3.3 Applications. The turbine d-c drive is limited to cases where the extreme flexibility of the variable voltage control system is desirable, and where conditions are such that a turbine power plant is more desirable than a diesel plant as a prime mover. Applications of this form of drive are very limited. The U. S. Corps of Engineers has a number of turbine d-c electric dredges in operation but no new ones have been constructed in recent years. A noteworthy example of the application of,turbine d-c electric drive to a special service vessel is the cable ship Long Lines [5]. This large oceangoing vessel is a twin-screw ship with a total shaft horsepower of 8500. Each of its twin systems consists of a constant-speed steam turbine and reduction gear driving a doublearmature propulsion generator, an a-c ship's service generator, and an auxiliary generator which supplies power to the bow cable drive. The propulsion generator is of the double-armature type so that it can be operated a t the desired speed of 900 rpm. Each double-unit generator supplies the power required by its associated single-armature, direct-drive propulsion motor which is capable of developing a maximum power of 4250 hp a t 135 rpm. The a-c auxiliary generator and the bow cable drive generator are each rated at about 500 kw. The twin-screw turboelectric propulsion system provides excellent maneu~erabilit~y,speed control, and remote operation desirable during cable laying and recovery. The propulsion motors can be operated from the main motor room, three locations on the bridge, a location near the bow sheaves, and from the aft steering station.
Section 4 The Turbine Alternating-Current Drive System 4.1 System Composition-Synchronous. All of the discussion in this section on a-c electric drives will be based on the use of synchronous machines. These systems will invariably be of the three-phase type and operate a t unity power factor. They will not, however, always be 60-cycle systems since prime mover considerations may influence the use of a higher frequency and the propulsion system is normally independent of other power systems. Frequently, however, 60 cycles
per second is a good selection, and this facilitates the use of industrially equivalent designs [6]. I n the case of steam turbines, this form of propulsion makes possible the use of a single nonreversing highspeed turbine operating at or near its most efficient speed. The system normally comprises, in addition to the steam plant, a single, variable-speed, nonreversible steam turbine driving a direct-connected generator, a propulsion motor, a source of excitation power, control
'
ELECTRIC PROP'ULSION DRIVES
equipment, interconnecting cable, and certain auxiliaries. If the prime mover is a gas turbine, it will normally be of the two-shaft type so that wide speed range operation, essential for ship propulsion, will be possible. The power turbine output speed on large-size gas turbines that might be fitted to moderate and large-size vessels is in the 3600 to 5400-rpm range, and is thus similar to the normal speeds encountered on steam turbines of similar rating. The output torque-speed characteristics are also similar, and thus the a-c electric drive that would be employed is basically similar for either steam or gas turbine prime mover systems. In either steam turbine or gas turbine applications, the generator is normally of the direct-connected, high-speed turbo-type. The generator is similar to the ususl central station unit and uses a distributed field winding placed in slots in a cylindrical steel rotor. It will normally have two poles although four-pole units are also feasible. The rotor carries ventilating fans at each end which circulate air in a closed system through the machine and water-cooled surface air coolers. This type of totally enclosed machine is now universally employed and has the advantages of keeping the windings clean, simplifying the installation, and making a quieteroperating machine. The generator is provided with temperature-detecting coils inserted in several locations of the stator winding so that observations can be made of running temperatures. Arrangements are also made, usually by electric space heaters, for heating the machine when idle to prevent condensation of moisture. Figure 12 illustrates typical physical characteristics of a-c propulsion generators. The directdrive synchronous motor is of the salient pole type and is characterized by its large number of poles (60 to 72 being common) in order to operate a t the low speed required by the propeller. The motor is therefore large in diameter and short in stacked length. It must be provided with a heavier than normal pole face winding, so that it can operate successfully as an induction motor under heavy torque loadings produced by the propeller under reversing conditions with headway on the ship. Typical physical characteristics of a-c synchronous propulsion motors are shown in Fig. 13. Motors usually have forced-air circulation in a selfcontained system with surface air coolers. Because the rotative speeds are so low, motor-driven blowers are used to provide the necessary air circulation. These blower units are mounted on or adjacent to the main motor. The motor is normally provided with pedestal-type bearings bolted directly to the ship's.foundation, and one of these units can be combined with the propeller thrust bearing if desired. To facilitate major maintenance or repair, the span between motor bearings is increased sufficiently so that the motor stator can be shifted axially to expose the stator windings and the rotor pbles without necessitating bearing movement or shaft disassembly. This is an important feature that contributes
349
significantly to the ease with which inspections can be made and maintenance work accomplished with the motor in place. As in the case with generators, stator temperature detectors, fire-extinguishing connections, and space heaters are provided. The pxcitation requirements for the a-c system are considerably more severe and difficult to meet than for the d-c system. When starting or reversing the motor, it must operate as a squirrel-cage induction motor until its speed is electrically close enough to that of the generator so that it may be synchronized. During this out-of-synchronism mode of operation, the motor power factor is very low and its current demands high. To maintain generator voltage and provide the current needed to develop proper motor torque, the generator must be over-excited on a short-time basis. These conditions are particularly severe when the motor is reversed from a full-speed ahead operating condition because the ship continues to move through the water a t considerable speed and the water flow to the propeller causes it to resist motor efforts to stop and reverse it. Typical propeller torque-rpm characteristics are shown for various ship speeds in Fig. 14. With the a-c drive, it is necessary to have a separate source of excitation power. It is not practical to use generator-driven exciters because of the wide speed range of the turbine-generator set (100 percent to 20 percent speed) and the fact that the exciter must provide approximately 2.5 times normal voltage during maneuvering, which is done a t the minimum generator speed. Excitation is commonly supplied from separate motorgenerator sets operating from the ship's service system. A pilot exciter of the rapid response or of the static regulator type is also used so that regulating and limiting control functions may be provided conveniently. A standby excitation set should be provided to safeguard the availability of this vital auxiliary. I n the case of twin-screw ships, a total of three excitation sets would be supplied, the third unit being arranged as the spare for either of the two propeller systems. Under steady running conditions, sufficient excitation must be maintained on the main generator so that the main units do not pull out of step because of torque variations of the propeller caused by ,turning or sea conditions. , This can be done by maintaining a degree of overexcitation, or regulators may be employed which act automatically to provide increased excitation when torque variations require it. I n the turbine-electric a-c system, speed control of the propulsion motor is obtained by frequency control, or, in other words, by varying the speed of the prime movergenerator set. The turbine is under the control of . a governing system which permits its speed to be varied over the range of from about 20 percent to 100 percent speed. All steady-state running is performed with the main motor in synchronism with the generator and the speeds of the motor and generator proportional to each other.
350
MARINE ENGINEERING
ELECTRIC PROPULSION DRIVES C
REMOVE COOLERS
I
+ =PLAN A VIEW R O ~ RR&OVAL
Fig. 12
A-c proplion generatorsdimemions and weights
- -
HP
WEIGHT IPOUNDSI DIMENSIONS (INCHES I ISQRPM I A I B I C I D I E I F I G I H I I I J I
WEIGHT
3 5 0 0 0 ~ 2 4 8 ~ 2 8 4 ~ 2 8 5 ~ 1 2 5 ( 1 3 5 ~ 1 2 5 ~ 1 7 2 ~ 1 31620)080 2 0~74~ 14511351195 115) 701 821
450000
I 0 3 RPM WOO0 )31b(3281338 ~ l 5 8 ~ 1 6 7 ~ 1 5 6 ~ 2 0 06) 0 1 21 8 7 4~1
525000
r weigh Fig. 13 A-c propulsion m o t a r 4 i m ~ s i o n and
Figure 15 shows a schematic wiring diagram for a typical single-screw turbine-electric a-c propulsion system. The direction of rotation of the propulsion motor is controlled by reversing contactors (S1 to 85). This reversing switch group can be remotely controlled switches, contactors, or breakers, depending upon the control arrangement and system size. Excitation power is supplied by a main or standby excitation motorgenerator set driven from the auxiliary (ship's service) power system. Other excitation systems can be used, such as auxiliary turbine or diesel-driven exciter or static exciters supplied from the ship's service power system. The propulsion motor field is controlled by an excitation contactor with contacts E l , E2, and E3. The field of the propulsion exciter is automatically controlled by a propulsion regulator. This regulator automatically adjusts the excitation power level in response to the direction control lever and throttle control lever movement. The control station, depending upon the control arrangement, will have either one or two control levers. With a two-lever station, one lever controls the direction switch (S1 to 85) and supplies the logic signals for automatic starting and synchronizing of the propulsion motor. The other lever provides speed control logic to the prime mover governor. Interlocking between the two levers is provided to (1) allow moving the reverser lever only when the throttle lever is in the maneuvering speed position, and (2) to allow movement of the throttle levers from the maneuvering position only when the reverser is in the run position and propulsion motor has started and is synchronized with the generator. With a two-lever control station, the following sequence would be automatic in response to the lever movements: A. REVERSER lever is in STOP THROTTLE lever is in MANEUVERING 1. The turbine-generator is running a t minimum speed (15 to 25 percent rated). 2. Excitation bus is a t zero volts with the M-G set running. 3. Motor field is shorted through its discharge resistor ( E l and E2 open, E3 closed). B. REVERSER lever is moved to AHEAD 1. Direction contacts S1, 52, and 53 close. 2. The propulsion regulator increases generator excitation to maintain rated volts per cycle or maximum level required by system design, usually between 6 and 7 times rated generator field power. 3. Propulsion motor starts and accelerates as an induction motor. 4. The field of the propulsion motor is automatically excited by the closing of the motor field contactor. ( E l and E2 close and E3 opens). The automatic synchronizing system should be designed to apply motor field at the proper slip and a t a phase angle to obtain a high pull-in torque. C. THROTTLE lever is moved to the desired ' propeller speed 1. The turbine speed governor resets to call for the speed indicated by the throttle lever position signal.
35 1
Fig. 14 Typical propeller rpm-torque curves for various h i p speeds
2. Fuel-power increases until the system speed. reaches the turbine governor set point. The following sequence would be followed in reversing from full AHEAD to full ASTERN: A. THROTTLE lever is moved to the minimum speed position 1. The turbine governor is reset to call for minimum speed. B. REVERSER lever is moved from AHEAD to ASTERN 1. Excitation voltage goes to zero. 2. Motor field contactor shorts motor field ( E l and E2 open, and E3 closes). 3. Reversing switch contacts S1, 52, and 53 open and+contacts52, 54, and S5 close. 4. The propulsion regulator increases excitation to the generator to maintain rated volts per cycle or to the maximum excitation power limit. 5. The propulsion motor is now operating as an induction motor. It will reverse and accelerate to a speed close to the synchronous speed of the turbinegenerator. 6. The field of the propulsion motor is automatically excited and it pulls into step as a synchronous motor. C. THROTTLE lever is moved to the desired propeller speed
MARINE ENGINEERING
EXCITATION BUS
SHIP SERVICE POWER
Fig.
IS
Typical schematic circuit diagram-turbine electric a-c drive
1. The turbine speed governor resets to call for the speed indicated by the throttle lever position signal. 2. Fuel-power increases until the system speed reaches the turbine governor set point. Control systems with higher degrees of automation, programmed sequencing, and remote actuation can be employed, and the trend is to move in this direction. Protection against faults or short circuits in the main circuit is provided by phase balance relays or a differential relay, or both. The functioning of these relays causes excitation to be removed from the propulsion motor and generator. Ground protection is provided by a ground current relay circuit in the propulsion motor neutral. Functioning of this relay also removes excitation from the propulsion motor and generator. Systems which utilize multiple prime movers must incorporate means to obtain proper kw load-sharing between prime movers, kvar sharing between generators, and provisions for adding or removing a generator set from propulsion duty. Real load-sharing between propulsion engines is usually accomplished by speed droop. Since good speed regulation (low droop with load) is not required or even desirable, this droop can be set much higher than normally considered for ship's service generator sets. A speed droop setting of 10 percent should provide good load-sharing and stable operation. Reactive load-sharing between generators can be accomplished by connecting the generator fields in parallel from a single exciter, or by reactive droop if individual exciters are used. Various methods can be used for adding or removing generator sets, depending upon the type of prime mover and generator. The most straightforward method is to provide adequate switchgear and automatic paralleling between propulsion generator sets. Where multiple prime movers are used, it is desirable
to provide for operation on less than the full number of propulsion sets. For instance, if three turbine-generators are used to provide power to one propulsion motor, it is desirable to incorporate coritrol features that enable operation on 1, 2, or 3 generator sets. Operation on a reduced number of generators (reduced power capability) increases the propeller reversal time and electrical machine load because the propeller torque during a reversal from high ahead ship speeds exceeds the plugging motor torque available with reduced power input. The maneuvering (reversal time) performance is improved and overloading required of the generators and motor is reduced by the addition of a dynamic braking resistor. This resistor is connected to absorb the propulsion p u m p back power due to the forward motion of the ship and the resultant water action on the propeller. When the ship has lost sufficient headway, the dynamic brake is disconnected and the propulsion generators in use are connected in reverse phase rotation to the propulsion motor. The power from the reduced number of propulsion generators should now be sufficient to accelerate the propulsion motor to near synchronous speed in the astern direction. The general requirements for connecting cables as given in Section 2 for d-c propulsion apply. Cables should preferably be of the single-conductor type and should have nonmagnetic armor. The armor should be grounded at approximately the mid-point of the cable run. Single-conductor, a-c cables should not be located closer than 3 in. from parallel magnetic material and, where cables pierce a bulkhead, all conductors of the same set should pass through a common nonferrous plate to prevent heating of the magnetic bulkhead. Single conductor cables should be supported on insulators, and where cables are arranged in groups they should be transposed when lengths exceed about 100 ft.
ELECTRIC PROPIJLSlON DRIVES
Air coolers are supplied with the propulsion motors and generators and are normally built in as a part of the enclosed ventilating air system. The necessary motordriven vent fans for the main motor are also considered a part of the motor and are normally mounted as a part of the main motor. Lubrication of the generator bearings is provided from the turbine system. Propulsion motor ,bearings are normally force-lubricated from a separate system consisting of a motor-driven pump, oil coolers and strainers, and a sump tank. 4.2 System Design Features. For electrical losses in the transmission between the turbine and the propeller, an average figure of 6 percent can be taken. This figure does not include the excitation power of the generator and motor fields. This loss will amount to about 1 percent of the kilowatt rating of the generator and about 1.5 percent of the kilowatt rating of the motor. This loss is included separately since it is supplied from a source separate from the main turbine. The propulsion system, being an independent system, enjoys considerable freedom of choice as to voltage and frequency. As noted previously, the choice of voltage level is based on motor and generator machine design considerations, and on the availability of needed switchgear or control apparatus. As a guide, a 10,000-hp .system might well use about 2400 volts and a 50,000-hp system would find 6600 to 7500 volts advantageous. The minimum frequency of a turbine-electric, a-c system is largely determined by the speed of the turbinegenerator unit and the use of a two-pole generator design. Since the turbine will invariably have a rotation speed of 3000 rpm or more, this results in a minimum frequency of 50 cps. The corresponding motor, if its desired operating speed is 100 rpm, would have 60 poles. Higher frequencies would require more poles on the motor and tend toward a less satisfactory and less economical design. Frequencies, then, tend to be in the 50-90 cps range, with the lower frequencies favoring the slow-speed motor and the higher frequencies being accepted only as necessary from the turbine standpoint. The inherent torque characteristics of an a-c ship propulsion motor require coordination of the system design. The a-c motor and generator combination will not carry overload torques under steady running conditions a t full power unless the machines are designed with considerable torque margin or unless provision is made in the control and excitation system to automatically increase the excitation of the machines when the overload torque is imposed. Unless these provisions are made, the machines will pull out of step and require resynchronizing. The torque requirements on the motor during a fullpower, full-speed reversal are the most exacting and usually largely fix the design of both the motor and generator. Since synchronous motors are almost universally used in a-c propulsion plants, only this type of motor will be considered in the discussion of this problem. The rotor is provided with a substantial
353
induction winding which must be carefully designed so that it will not detract appreciably from the purely synchronous motor characteristics for steady running, and so that it will be able to develo~sufficient induction motor torque for stopping the propeller and bringing it up to speed in theastern direction while the motion of the ship through the water is still attempting to drive the ~ r o ~ e l l eand r the motor in the ahead direction. The ind;ction winding must be carefully proportioned so as to have sufficient thermal capacity to handle the heavy currents induced during the maneuvering cycle. In order to studv further the characteristics necessarv to accomplish a Ifull-power, full-speed reversal, it is desirable to review briefly the sequence of operations during the maneuver. On signal for full-speed astern, the operator first moves the turbine speed control lever to the IDLING position, which sets the governor for about 20 percent speed. He then moves the field lever to the OFF position, removing excitation from both the generator and motor. He next moves the direction lever from the FULL AHEAD position to the START position astern. While these operations have been going on, the ship has been slowing down because power was removed from the propeller. However, it is still moving ahead through the water at a considerable speed and thus driving the propeller and motor in the ahead direction. In the last o~erationjust mentioned. two of the three-phase connections between the generator and the motor are reversed; and approximately double excitation is applied to the generator field. Power therefore is applied immediately to the stator of the motor in the reverse direction, causing large currents to circulate in the damper winding of the motor and therefore developing a heavy torque which acts to stop the propeller and then to reverse it and bring it up astern, close to synchronism with the main generator. The operator then moves the control lever to the RUN position, which applies field to the motor, pulling it into synchronism with the generator as a synchronous motor and a t the same time reducing the generator excitation to normal. While the motor is operating as an induction machine, it has a rather low power factor and places a current demand on the generator of from three to five times normal. If no provision were made for overexciting the generator field, its terminal voltage would collapse, and the motor would fail to deliver the necessary torque. This is the reason that the excitation system must be closely coordinated with the machine design. The motor design must be carefully proportioned so that it will develop sufficient synchronizing torque to pull into step with the generator although the ship still is going ahead and causing ahead torque to be developed by the propeller. As each type of ship and propeller will have different maneuvering characteristics, the first step in determining the proper design of the propulsion motor and generator is to calculate the expected maneuvering chmacteristics. Figure 16 shows the full-power reversal speed-torque curves for a typical single-screw, turbine-electric-pro-
354
ELECTRIC PROPULSION DRIVES
MARINE ENGINEERING Table 1
virtually all cases the main motor will be located as far aft as possible. This results in important reductions in length of line shaft, shaft alley, and number of bearings. Motor foundations should provide for axial space in which to shift the motor stator to uncover the windings of stator and rotor without disturbing the motor rotor and its bearings. As in the case of any propulsion plant, a careful analysis of the complete plant must be made to ensure that there will be no dangerous critical speeds within the operating range. Although the electrical machines provide damping action, it is possible to obtain vibration frequencies due to a combination of the propeller and the synchronous motor which will resonate with the natural frequency of the system. It is also possible to reflect such pulsations back into the turbine-generator set. It is therefore necessary that a careful analysis of the entire system be made. 4.3 Physical Characteristics. Because of the great variation in requirements and arrangements, precise data on equipment weight and size cannot be presented; however, general outline dimensions are given in Fig. 12 for a-c propulsion generators of the turbine type, and Fig. 13 gives similar data for a-c propulsion motors of the direct-drive type. These figures also give overall weights for these machines. The speed chosen for these machines is arbitrary but nevertheless typical and illustrative of units that would be used. Control or excitation equipments are small and light in weight when compared to the main motor and generators and lend themselves to mechanical packaging that facilitates convenient installation.
Allowable Temperature Rises for A-C Machines, Deg C
-
INSULATION Ambient .................... Armature windings by imbedded detectors.. . . . . . . . Salient pole fields b resistance. ~urbine-typefieldsty resistance
PROPELLER SPEED, O/o RATED
Fig. 16 Typical propeller and motor torque characbrirtiu
I
I
pelled oceangoing vessel. Curve A shows the calculated maneuvering torque capacity of the propulsion plant. Curve B shows the propeller torque requirements during maneuvering, starting with the ship going ahead at full speed. As the ship slows down, the propeller torque is reduced so that there is a family of curves similar to curve B to cover various ship speeds (see Fig. 14). Curve C is a portion of one of these curves and represents the propeller torque-speed conditions existing when the motor speed has come within the range from which, at point Dl it can be synchronized with the generator. The motor can be synchronized with its generator when the slip has been reduced to about 2 percent or less, and this ability is enhanced by timing the application of motor field so that the phase of its angular slip position is most favorable for synchronization. During induction motor operation, the motor field will .be short-circuited on itself or through a resistor chosen to aid the synchronizing process and to keep induced voltages.within acceptable values. The size, weight, and cost of a-c electric propulsion equipment are increased if inherent torque margins are increased. The torque margin of a propulsion system is defined as the increase in torque, above rated torque, to which the system may be subjected without having the motor pull out-of-step with the generator. A torque margin of 10 percent is sufficient when combined with an automatic control means for raising excitation momentarily whenever torque increases occur, such ss can be
Class B 50
Class F 50
70 70 80
90 90 100
Class H 50
110 110 120
produced when maneuvering. I n the case of twin-screw vessels, a sudden hard-over rudder movement can appreciably increase the loading of the inboard propeller and, in such in~tallations,load limit as well as excitation increase may be desirable. The comments made in Section 2.2 regarding the availability of improved insulations and the demise of Class A insulation apply to a-c machines as well. Class B insulation systems are the usual standard but the use of Class F is increasing. Allowable temperature rises are given in Table 1 and more complete details will be found in reference [I]. Again it should be noted that the ambient of totally enclosed machines fitted with air coolers is that of the air delivered by the coolers and, if designed for less than 50 C air delivery, corresponding extra temperature-rise allowances can be made. Propulsion motors and generators are rated for the maximum full-power, continuous shaft horsepower requirements of the vessel. Electrical equipment so designed, with temperature ratings as indicated in the preceding paragraph, obviously can carry some overload without distress. Such overload will result in higher operating temperatures and some reduction in overall insulation life of the machine. The amount of such life reduction depends on the severity and frequency of the overloads and on the length of time of such service. Experience indicates that it is advisable to provide both the generator and motor with closed ventilating systems and water-cooled air coolers. The preserving of clean windings in such systems, and the elimination of extensive air duct systems, makes the overall cost of the closed system favorable. I n cases where added safeguards against air-cooler water leakage seem prudent, double-tube type air coolers are used. For installations where the propulsion motor is located in a room by itself, it is not necessary to provide a completely closed ventilating system on the motor itself. In this case it may be advantageous to provide for the ventilating fans to exhaust the warm air from the motor through the air coolers into the motor room. The cool air is then drawn back into the motor through openings at both ends of the motor. The machinery layout selected will give consideration to many factors and these will vary for different types of ships. The mechanical independence of the main motor and the turbine-generator set, however, is useful and in
'
355
TZSE-A1 Tankers Turbine Generator. . . . . . . . . . . .5400 kw, 3715 rpm 2370 volts, 3 phase 62 cycles, 1.0 pf Main M ~. . . . ~ . . . . .~. . . . ~. . .6600 . hp, 93 rpm 2370 volts. 3 ~ h a s e 62 cycles, l.oSpf T2-SE-A2 Tankers Turbine Generator. . . . . . . . . . . ,7650 kw, 3715 rpm 3610 volts, 3 phase 62 cycles, 1.0 pf Main Motor. . . . . . . . . . . . . . . . . .10,000 hp, 106 rpm 3610 volts, 3 phase 62 cycles,. 1.0 pf ,
This type of drive is discussed further in [8]. b. Passenger Vessel :Canberra The Canberra is a large passenger vessel built for service between England, Australia, and the West Coast of North America [9]. The choice of steam-turbine, a-c electric drive was based on a number of factors, and was no doubt influenced by very favorable operating passenger experiencevessels. of its owners with three other turboelectric
The Canberra has a twin-screw propulsion plant with a combined rated shaft horsepower of 85,000. Each screw is driven by a 42,500-hp double-unit motor at 147 rpm full speed. The motors are of the synchronous type and operate at unity power factor and a maximum voltage of 6000 volts. Each of the double-unit motors is supplied with power by a single generator directly coupled to a high-performance steam turbine. The generators are somewhat oversized in order to permit a single unit to 4.4 Applications supply the desired amount of power to both propulsion a. T2 Tankers. Over 500 turbine a-c electric drive motors for certain legs of the vessel's itinerary. It is tankers were built during the 19401s,mainly in response interesting to note that the tested efficiency of the main to World War I1 needs. Many were converted to other motors is over 98 percent throughout the power range service such as ore-carriers and self-unloading colliers [7]. of 50 to 100 percent and is 98.4 percent at full power. An unusual aspect of this application is the emphasis Many of these vessels were "jumboized" by adding placed upon quietness of operation. The Canberra is longer midbodies and in some cases only the stern sections were retained. Invariably the propulsion plant probably the quietest and smoothest-running ship of its was retained and, in many cases, overhauled and type in service and her electric drive, while contributing uprated. There were numerous instances' when the to her quietness directly, also enables the two propellers propulsion turbine-generator sets were used to feed to be run in synchronism and in the phase position that power ashore in times of disaster or other critical need. produces the greatest neutralization of vibratory forces; Popularity and longevity are not the main reasons for the 42-pole motors permit 21 different synchronized citing this application but rather the fact that the basic operation relationships shaft-to-shaft. I n addition to plant is typical of a steam-turbine, a-c electric drive for a the actual vibration reduction, the elimination of the single-screw ship. The plant consists of two boilers usual "twin-screw beat" is a particularly important which supply steam to a single propulsion steam turbine psychological improvement on a passenger vessel. The direct-connected to a two-pole generator, one 80-pole twin screws of the Canberra are normally run in synchrosynchronous motor direct-connected to the propeller nism with each other except when in confined waters shaft, and an excitation and control subsystem. The where maneuvering is expected and independent propeller control is important to the handling of the vessel. rating of the main units is as follows:
.
120
Section 5 The Diesel Alternating-C~rrrentDrive System 5.1
Differences from Turbine a-c Electric Drive.
Few diesel a-c electric drives have been built but they are of particular interest from an electric-drive . standpoint. The diesel a-c electric-drive system consists of a multiple number of diesel engines, each driving a directconnected, salient pole, a-c generator, a single slowspeed, direct-drive propulsion motor, an excitation subsystem, and a control system. The main motor will most likely be of the synchronous type, and subsequent portions of this section will be based on that type of motor. The fundamental principles of speed control and reversal of the propulsion motor are the same as for a turbine-electric drive. However, the torque characteristics of the diesel engine a t reduced speed are less favorable than those of a turbine ank there is risk of stalling the engines during critical maneuvering if proper control safeguards are not provided. I n addition, a number of generating units must be operated in parallel over a range of from 30 to 100-percent speed. These factors combine to require a carefully designed control system to ensure that the operating conditions can be met within the capabilities of the equipment. 5.2 Description of a Typical System. As a means of illustrating a diesel a-c electric drive, consider the Navy Submarine Tender Hunley. The Hunley is a diesel a-c electric drive of 15,000 shp [lo, 111. This application is typical of those special-service vessels in which use can be made of the prime mover generating sets for auxiliary
service when they are not required for primary propulsion. The propulsion plant consists of six 850-rpm diesel-generator sets rated a t 2655 bhp each with a 2000-kw generator output a t 3300 volts and 70.8 cycles. These six generators operate in parallel to supply power to the 15,000-hp direct-drive propulsion motor. The ship may be propelled by any combination of 3, 4, 5, or 6 generators, and up to 3 generator sets can be used to augment the ship's service power supply when tending power demands are heavy. The basic propulsion system is shown in Fig. 17. A single exciter is arranged to supply field power to all of the propulsion generators and the propulsion motor. The main exciter is under the control of a pilot exciter of the amplidyne type, and it in turn is controlled by the propulsion regulator. This regulator automatically maintains proper excitation for all operating modes. Any of the generators can be connected to augment the ship's service system for special duty such as cargo handling, pumping, or overside power supply. As shown, excitation for this mode of operation is from static exciters. Typically, each of the gix diesel-generator sets consists of a high-speed engine direcbconnected to a salient pole generator. The diesel-generator sets must be varied in speed in order to obtain propeller speed control. This can be done from 100-percent speed down to about 30percent speed under steady-state operating conditions. At this low speed, load transients must be avoided to prevent danger of engine stalling, and as a result maneu-
-
PROPULSION. GENERATOR EXCITATION BUS
Fig. 17 Schematic diagrum-dienl a-c electric drive 0
357
ELECTRIC PROPULSION DRIVES
MARINE ENGINEERING
I/
440 VA? POWER
1
I
I
I
I 1
AHEAD ASTERN PROPELLER RPM,XRATED
Fig. 18
Propulsion motor torque characteristics with power supply set for 40-percent astern rpm
0 0
vering operations are carried out a t higher engine speeds, i.e., 4 0 4 5 percent, so as to have increased torque capability and assurance of nonstalling performance. The propulsion motor is directly connected to the propeller shaft and is rated a t 15,000shp and unitypower factor. A voltage of from 3000 to 4000 volts is indicated for this machine. The motor is equipped with as heavy an amortisseur winding as is practical for a machine of this type without requiring an unbalanced design or reflecting adversely on normal synchronous motor operation. The motor is totally enclosed with surface air coolers and motor-driven blowers. A motor of this type is shown in Fig. 13. The motor torque characteristics under maneuvering conditions are of paramount importance. Figure 18 shows typical motor speed-torque characteristics at 40percent rated frequency (the assumed maneuvering speed selected for the diesel-generator sets). Curve A is the motor performance under full rated volts-per-cycle supply conditions, as would be expected on most industrial applications operating from an unwavering power supply. On a ship application, the motor must be started from a generator of equal capacity and the motor torque must be determined on this basis. The generator characteristics, motor characteristics, and excitation system must be selected so that optimum system performance is obtained. Curve B is the system motor torque exclusive of any prime mover limitation. This curve is based on the volt-ampere output from the generators with their field forced to about 2.5 times normal excitation. Producing this motor torque, however, requires the generator to demand more torque from the engine than is available over a part of the speed range. Curve C, a modification of curve B, shows the motor torque available within the limits of engine capacity. The torque demand of the engine has been reduced so as to just equal, but not exceed, the available torque. This is done by regulating the excitation on the
20
Fig. 19
40 60 80 SHIP SPEED, % RATED
100
Maximum operating speeds
generators in response to a speed signal from the enginegenerator set. Curves D and E are similar motor torque curves but with only half of the six dieselgenerator sets supplying power to the motor. The free-route propeller power-speed curves are shown in Fig. 19 as the usual cubic curve. To this has been added the power capabilities represented by 3, 4, 5, and 6 diesel-generator sets. The intersections of these curves represent maximum operating speeds for the various combinations. The propeller characteristics of importance during maneuvering operations are the dynamic relationships of propeller torque and speed for a series of vessel speeds. Such characteristics are shown in Fig. 14. The power requirements shown by Fig. 19 represent the steadystate duty as seen by the motor, and the curves of Fig. 14, when augmented by an appropriate acceleration component, represent the transient duty as seen by the motor. These then are the output requirements of the system, and all control and system designs are directed to meet these needs while staying within the basic capabilities of the primemovers, thegenerators, and the propulsion motor. The propulsion control is divided into three sections. One controls the two excitation motor-generator sets and provides for selection of the in-service set. Another section provides for generator control and connection to either the propulsion bus or to the special ship's service duty bus; and a third section provides for control of motor speed and direction of rotation. The propulsion motor speed and direction of rotation are controlled by three levers: REVERSING, FIELD, and SPEED. These, together with the major interlocking, are shown in the functional diagram of Fig. 20. The arrangement of mechanical interlocks will prevent damaging operation. I n general, the REVERSING
I
I
358
MARINE ENGINEERING ENGINE SPEED TRANSMIlTER MECHANICAL INTERLOCKING
I
I
AHEAD OFF DYNAMIC BRAKE ASTERN
It I
(
I
R VERSING 'LEVER
FULL
RUN NO. 2
I
I
FIELD LEVER Fig. 20
I
I
CEE~ (
Motor control leven
% OF RATED TORQUE
Fig. 21
Propulsion system characbristict-rtarting and mnning
lever cannot be moved unless the FIELD lever is in the OFF position and the SPEED lever is in the MANEUVERING position. The FIELD lever cannot be moved unless the REVERSING lever is in the AHEAD or ASTERN and the SPEED lever is in the MANEUVERING position. The SPEED lever cannot be moved unless the FIELD lever is in the RUN position. Each generator is controlled by a POWER SELECTOR lever having four positions: SPECIAL SERVICE, OFF, PROP 1, and PROP 2. This lever permits any generator to be added or removed from propulsion duty a t any time, and to be used for special service power supply when not used for propulsion. In order to start the plant, the generators which are to be used for propulsion (at least 50 percent of full capacity) are first switched from local control in the engine room to remote control a t the main propulsion control board. This connects the engine governors to a master trans-
mitter, and all engines being readied for propulsion service will run a t approximately the same speed. The generators are connected to the propulsion bus by moving their individual POWER SELECTOR levers to the PROP 1position. When the first POWER SELECTOR lever is moved to the PROP 2 position, sufficient excitation is applied to energize the bus and synchronize all of the connected generators. Each will in turn then be moved to the PROP 2 ~ositionand excited a like amount. The propulsion b;s is now energized by all of the in-service generators but at a reduced frequency and very low voltage. Moving the motor REVERSER lever to the AHEAD position connects the motor to the propulsion bus in the ahead phase rotation ready for induction motor operation with its field shorted through a discharge resistor. The motor may not start in this position because of the low generator excitation but when the FIELD lever is moved to position No. 1, approximately 250-percent generator excitation is applied and the motor starts. The heavy overexcitationof the generators, subject to relief by any necessary engine torque-limiting action, assures the production of maximum system torque and the motor accelerates to near-synchronous speed as an induction motor. The FIELD lever is now moved to position No. 2 and the motor synchronizes with the generators when its field is energized. This motor synchronization can readily be made automatic under the action of a slip sensing relay, which not only makes certain that the motor speed is sufficiently close to assure successful synchronization, but signals the application of its field a t the most favorable instant of its slip cycle. The final movement of the FIELD lever to the RUN position latches the motor field contactor in the closed position. The propulsion regulator now will maintain the bus voltage a t a constant volts-per-cycle and the machines will operate close to unity power factor. The SPEED lever can now be moved to any speed in the SLOW to FULL range with the entire system (engine generators and main motor) moving together in synchronism. The starting and running characteristics of the system are shown in Fig. 21 for six engine-generator sets in service and for three sets in service. Since typical diesels cannot be depended upon for reliable operation a t less than about 30-percent speed, the minimum propeller speed is also 30 percent under normal synchronous generator-motor operations. From a practical standpoint this is a satisfactory condition in almost every application, but if some lower speed must be provided, a subsynchronous mode can be utilized. The main motor, operating as an induction motor, is run a t high percentage slip by controlling the generator excitation to permit such action. Operation in this mode is within the capabilities of the electrical equipment because the propeller power demands below 30percent speed are very low. Once the ship is moving at relatively high speed (above 60-percent speed or so) considerably higher
ELECTRIC PROPULSION DRIVES
torques are encountered in reversing the propeller and stopping the ship. A reversal from high ahead speeds requires that special steps be taken to slow the vessel down part way before the engines can assume the burden within their capabilities. The situation can be seen from an examination of Fig. 22. The motor torque developed under the system constraints is unable to reverse the propeller until the ship headway has been reduced to about 65 percent. If the motor were unwisely put into such a process, it would be subjected to a long period of heavy plugging action and excessive heating of its amortisseur winding would result. A much more satisfactory situation results from the use of a dynamic braking process during the necessary ship deceleration period and the subsequent use of the motor in its induction motor mode. A dynamic brake is a resistor which can be connected to the motor terminals to absorb power generated in the motor by the action of the water passing through the propeller due to the ship's headway. This action does not produce heat in the amortisseur winding, and it is a highly effective method for developing astern thrust, particularly in the upper and more critical high-speed range. When the ship's velocity has been reduced to the necessary 65 percent, the dynamic brake can be disconnected and the propulsion motor connected to the propulsion bus in the reverse phase rotation. The motor can now be plugged and accelerated in the reverse rotation within the capabilities of the engines and synchronized when it gets close to the 40-percent speed of the generator sets. After synchronization, the motor speed can be increased quite rapidly and the vessel stopping procedure completed. This process is illustrated by Fig. 23. From a control sequence standpoint, the foregoing reversing actions are carried out from a full AHEAD condition in the following manner: (1) Return the SPEED lever to the MANEUVERING position. This positions the engine governors at 45-percent no-load speed. However, the complete system continues to run at 65-percent speed, due to the forward movement of the ship and resulting water action on the propeller. (2) Return the FIELD lever to the OFF position. This disconnects the propulsion motor field and reduces the field supplied to the generators. (3) Move the REVERSING lever to the DYNAMIC BRAKE position. This connects the propulsion motor to the brake and applies the correct motor field. The control lever is left in this position until the ship's headway has decreased sufficiently to permit the motor to be pulled into synchronism astern. (4) After the REVERSING lever is moved to the ASTERN position, the FIELD lever and SPEED lever are moved in the same fashion as described for starting. The generator field is forced to the maximum permitted by the propulsion regulator, the motor is automatically synchronized as it accelerates to near-synchronous speed
Fig. 22
Revenal characbristics
70RPM OR
% SPEED
MOTDR-PROPELLER AND GENERATOR RPM
fig. 23
High-speed rwcmal performance
astern, and the SPEED lever is moved to the desired astern propeller speed. The propeller is accelerated to the desired astern speed at a rate determined by the number of engines in service. The automatic load limit feature of the engine governors limits the fuel, and consequently the torque, to the maximum capacity of the system.
360
MARINE ENGINEERING
Section 6 Electric Couplings 6.1 -General Description. The electric coupling is a device for transmitting torque by means of electromagnetic forces without having any mechanical contact between the driving and driven members [12, 131. The electric coupling consists of two steel spiders with rims and flanges. The inside of the outer rim carries a number of poles which can be excited from an outside source through collector rings. The inner element, a laminated core, surrounds the rim and carries a squirrel-cage winding similar to that of the usual squirrel-cage induction motor. Both the inner and the outer elements are supported for rotation and separated radially by an air gap of about % in. One element, usually the inner one, will be connected to the prime mover, and the other to the driven device. Coupling ratings are usually in the 1000 to 4000-hp range. A typical coupling is illustrated in Fig. 24. The fundamental principle of the electric coupling is that of developing torque by inducing current in a squirrel-cage induction-motor-type winding by rotating a magnetic field around the squirrel-cage winding. The coupling-driven element rotates in the same direction as a
!I
Rg. 24
the driving element but a t a slightly slower speed, and the amount of this "slip" is just that required for the development of the necessary driving torque. I n order for electric couplings to be suitable for maneuvering a ship, they must be capable of producing large amounts of torque a t high slip. Thus they are normally equipped with double-layer, bar-type, squirrelcage windings and are designed to produce 150-percent pull-out torque, as well as a minimum of 75-percent normal torque up to 140-percent slip. Such high values of slip will be encountered during reversal duty when the prime mover is operating in the reverse rotation and the propeller is yet operating in the ahead rotation. The double-layer, squirrel-cage winding enables the coupling to produce high torque a t high slip conditions such as occur during maneuvering operations. When the slip is high, the induced voltage is of high frequency, causing the higher reactance deep bars to force the current into the outer high-resistance bars. This results in maximum torque. When operating at normal slip the frequency is low, and a major portion of the current flows in the .deeper, low-resistance winding,
Electric coupling
ELECTRIC PROPULSION DRIVES
resulting in high efficiency. Efficiency is usually above on Shipboard," IEEE Publication No. 45, February 97 percent, including excitation loss of about 1 percent, 1967. slip of 1% percent, and some windage loss. 2 J. A. Wasmund, "Series- Versus Parallel-Con6.2 Applications. A number of installations of nected Generators for Multiple-Engine D-C Dieselelectric couplings have been made on cargo vessels and Electric Ship-Propulsion Systems," Trans. AIEE, 1954. on large tugboats. The typical arrangement utilizes 3 W. E. Jacobsen, "Marine Power Applications," a pair of moderate-speed reversible engines to drive a Standard Handbook for Electrical Engineers, 10th edition, single propeller via electric couplings and a combining McGraw-Hill, New York. reduction gear. Here the couplings provide for the 4 J. A. Wasmund, "How Trials Prove Design engagement or disengagement of the engines simply by Theories," Marine Engineering/Log, August 1956. energizing or deenergizing the field winding. 5 "CS Long Lines," Marine EngineeTinglLog, July I n the case of a twin-engined ship, the usual procedure 1IYVV. OR9 when maneuvering in very close quarters is to run one 6 D. W. Drews, "Turbine-Electric Propulsion for engine ahead and the other engine astern. The ship may Ships," ASNE Journal, August 1963. then be maneuvered readily in either direction by 7 M. Mack Earle, "The Conversion of T2 Tankers operating a lever which applies field to the proper coupling, and thus connects the propeller to either the for Great Lakes and Seaway Service," Trans. SNAME, ahead-running or astern-running engine. All ahead and vol. 68, 1960. 8 "Sun-Built T2-SE-A1 Tankers, " Marine Engineerastern thrusts, within the capability of one engine, are ing and Shipping Review, July 1947. then attainable without further reversing of the engines. 9 T. W.Bunyan,P.D.Morris,andD.D.Stephen, Electric couplings act as torsionally flexible members and torsional dampers. The pulsations in torque from "Canberra," Trans. IME, October 1962. 10 W. E. Jacobsen and R. L. Koch, "Diesel-Electric the engines are smoothed out, reducing gear wear and Propulsion for Polaris Submarine Tender," ASNE noise and minimizing v torsional vibrations in the drive ~ou~rnal, August 1962. system. 11 H. M. Burford, R. L. Koch, and J. D. Westbrook, The propulsion control system, in addition to the usual engine starting, stopping, and reversing features, requires "Performance of a Diesel Electric A.C. Propulsion Plant only the integrated arrangement of engine-speed govern- (Based on the Design and Sea Trials of USS Hunley ing and coupling excitation control. It can be arranged (AS-31))," SNAME Hampton Roads Section, October to suit virtually any particular requirements and is well 1962. suited to remote or pilothouse control. 12 M. R. Lory, L. A. Kilgore, and R. A. Baudry, "Electric Couplings," Trans. AIEE, August 1940. References 13 M. R. Lory, "Electric Couplings for Great Lakes 1 "Recommended Practice for Electric Installations Ships," SNAME Great Lakes Section, September 1950.
1 I
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
I
CHAPTER XI I*
C. L. Long -
f
II System Propellers, Shafting, and Shafting Vibration Analysis
Section 1 Introduction 1.1 General. A main propulsion shafting system transom sterns. The distinguishing characteristic of (including the propeller) consists of the equipment neces- this arrangement is that the shafting must be extended sary to convert the rotative power output of the main outboard for a considerable distance in order to provide propulsion engines into thrust horsepower, suitable for adequate clearance between the propeller and the hull. propelling the ship, and the means to impart this thrust One or more strut bearings are required to support the to the ship's hull. In the following pages, the design of outboard shafting. a main propulsion shafting system will be discussed from A shafting arrangement typical of single-screw merthe viewpoint of a shipbuilder undertaking the task of chant ships is shown in Fig. 2. The arrangement illuspreparing a detailed design. I t will, however, be as- trated corresponds to the so-called Mariner or clear-water sumed that the propeller hydrodynamic design has been stern design (there being no lower rudder support); developed; the hydrodynamic design of propellers and Powever, the shafting arrangements of most merchant other propulsion devices is thoroughly covered in Prin- ships are very similar. The major difference between the ciples of Naval Architecture [I]' and therefore will not be shafting arrangements of various merchant ships is the pursued here. Although the fundamentals outlined in location of the main engines. When the main engines the following sections apply to all types of propulsors are located well aft, such as on tankers, there may be as andc prime movers,. the discussion has been primarily few as one or even no inboard bearings at all. When the directed towards a conventional arrangement with a main engines are located approximately amidships, as fixedLpitch propeller and a geared steam turbine main on dry cargo ships, a considerable length of inboard engine. This was necessaw in order to reduce the range shafting is required. of variations which had to be considered. The shafting located inside the ship is termed line Due to the nonuniform wake field in which a ship's shafting. The outboard sections of shafting (wet shaftpropeller operates, the propeller is a source of potentially ing) are designated differently depending upon their dangerous vibratory excitations. The shafting system location. The section to which the propeller is secured itself, which is inherently flexible, is extremely vulnerable is the"propel1er shaft or tail shaft. The section passing to these vibratory excitations; consequently, an analysis through the stern tube is the stern tube shaft unless the of the dynamic characteristics of a shafting system is an propeller is supported by it (as is the case with most integral aspect of the design process and is discussed in merchant ships) in which case it is designated as the this chapter. propeller shaft or tail shaft. If there is a section of 1.2 Description of Shaftfng System. The main pro- shafting between the propeller and stern tube shafts, it wpulsion shafting system must accomplidh' a number would be referred to as an intermediate shaft. of objectives which are vital to the ship's operation. Shafting sections are connected by means of bolted These objectives are: (a) transmit the power output from flange couplings. The coupling flanges are normally the main engines to the propulsor; (b) support the pro- forged integrally with the shafting section; however, pulsor; (c) transmit the thrust developed by the propulsor when required by the arrangement (e.g., stern tube to the ship's hull; (d) safely withstand transient operating shafts which require flanges on both ends and also require loads (e.g., high-speed maneuvers, quick reversals); (e) corrosion-resistant sleeves to be fitted to the shaft in way be free of deleterious modes of vibration; and df) provide of bearings), a removable coupling, sometimes referred to reliable operation throughout the operating range. as a muff coupling, is used. Figure 1 is a shafting arrangement typical of those Bearings are used to support the shafting in essentially found on multishaft ships and single-shaft ships having a straight line between the main propulsion engine and the desired location of the propeller. Bearings inside the ship are known by several names with line shaft bearings, steady bearings, and spring bearings being the most Numbers in brackets designate Reference8 at end of chapter.
I
iI t
MARINE ENGINEERING STATE PERFORMANCE REQUIREMENTS ESTABLISH DESIGN CRITERIA DEVELOP SHAFTING ARRANGEMENT
ESTABLISH SIZES
DETERMINE DYNAMIC CHARACTERISTICS DEVELOP DESIGN DETAILS
DESIGN COMPLETE
Fig. 3
Shafting system design sequence
popular in that order. Bearings which support outboard sections of shafting are called stern tube bearings if they are located in the stern tube and strut bearings when located in struts. Outboard bearings may be lubricated by either seawater or oil; high-quality seals are required in the event the latter is used. I n order to control flooding, in the event of a casualty, bulkhead stuffing boxes are installed where the shafting passes through bulkheads. A more substantial seal is installed a t the forward end of the stern tube where the shafting penetrates the watertight boundary of the hull. The propeller thrust is transmitted to the hull by means of a main thrust bearing. The main thrust bearing may be located either forward or aft of the slow-speed gear. If located forward, the thrust collar is detachable so as to permit the ii~tallationof the gear on the shaft and, secondarily, to permit replacement of the thrust collar if ever required. If located aft, the collar may be forged integrally with either the slow-speed gear shaft or a subsequent section of shafting. Since one purpose of the main thrust bearing is to limit movement of the slow-speed gear, the main thrust bearing is usually installed close to the gear. Installation of the thrust bearing close to the gear also facilitates lubrication of the thrust bearing. 1.8 Design Sequence. The design of a shafting system is, by necessity, an iterative process because the various system design parameters are, to some extent, mutually dependent. The iterativedesign process usually followed is illustrated in Fig. 3. As indicated by Fig. 3, the first step in the design of a shafting system is to state the performance requirements;
that is, the type of propul$ve system, number of shafts, type of service, and the like. Next, the design criteria to be employed must be fixed; i.e., one of the various classification society rules could be followed, oil-lubricated stem tube bearings may be selected, hollow shafting may be ruled out, etc. I n establishing the design criteria, it must be recognized that the shafting interfaces with thepropulsor, the main engines, and the ship system as a whole. After the design criteria are established and the general ship arrangement is available, an approximate shafting arrangement can be developed. This entails a t least tentatively locating the main engine, propeller, and shaft bearings with due regard given to arrangement restrictions, clearances required, shaft rake, construction restraints, and overhaul and maintenance requirements. Before the design can progress further, the shafting diameters, corresponding to the preliminary arrangement, must be computed along with the length of shafting sections, flange dimensions, and preliminary propeller data. With this data the bearing reactions can be approximated and the bearing dimensions and loadings can be checked. At this point, it will generally be desirable to adjust the bearing arrangement tentatively selected so as to obtain more equal bearing reactions or to alter the number of bearings. Variations in bearing loads due to thermal expansion of the shafting bearings, particularly those in the way of the main engines, are investigated to ensure satisfactory bearing performance under all operating conditions. There are three basic types of vibration which can occur in a main propulsioq shafting system; these are torsional, longitudinal, and whirling vibration. It is essential that a preliminary vibration analysis of the shafting system be made in the early design stages because the shafting vibration characteristics are largely established by the ship parameters that are fixed a t that time. Specifically, the shape of the hull afterbody, type of propeller, propeller aperture clearance^, number of propeller blades, length of shafting, shaft material, position of the m+in thrust bearing, type and configuration of prime mover, spacing of the aftermost bearings, and type of aftermost bearings largely establish the dynamic characteristics of a shafting system. The subsdquent development of design details has a relatively secondary effect as compared with these major parameters. In addition, an analysis of the system's response to shock loadings is required for naval combatant ships. An analysis of the dynamic characteristics of a shafting system can be one of the more complex aspects of the design process. Once the arrangement, component sizes, and dynamic characteristics have been shown to comply with the d e sign criteria, design details are developed. This entails designing flange fillets, flange bolts, keys, keyways, sleeves, and the like.
PROPELLERS,SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
Section 2 2.1 Location of Main Engines. The engine flange location and the propeller location are essential information in establishing the shafting arrangement. The foreand-aft position of \he main engines is generally established during the preliminary design stages after studying the ship cargo stowage, ship trim, and shafting system. The cost and n~eightof shafting are significant; for these reasons, and also to minimize the use of prime cargo space for the main machinery and shafting, the main machinery is located as far aft as practicable. With vessels such as oil tankers, the main machinery is confined to the stern end of the s h i ~such that a short run of shafting is required and the ;umber of line shaft bearings is minimal. On the other hand, dry cargo vessels do not have the ability to adjust their operating draft by taking on ballast; therefore, in order to provide satisfactory light load draft conditions, it is necessary to locate the main engines (and associated weight) well forward of the stern. Normally the main engine should be set as close to the inner bottom as the configuration of the main machinery will permit. It is possible, and it is the usual case, to have limited projections of the main machinery (e.g., the slow-speed gear lube oil suqp) below the inner bottom when such projections do not excessively weaken the inner bottom (see Figs. 6, 8, and 9 of Chapter I). The main engine location in the athwartship direction is on the ship centerline of single-screw ships. On multiscrew ships the engines are set off the ship centerline approximately the same distance as the propellers, but the shaft centerlines usually do not parallel the centerline of the ship. The location of the engine in the athwartship direction is controlled by the propeller location, main engine details, and the machinery room arrangement requirements. 2.2 Location of Propeller. The location of the propeller is determined by the propeller diameter, the acceptable clearance between the propeller and the baseline of the ship, and the acceptable clearances b e tween the propeller and the hull in the plane of the propeller. Although the propeller diameter selected should theoretically be the one corresponding to optimum efficiency for the propeller-ship system, in practice the optimum propeller diameter is usually larger than can be accommodated. As a result, the propeller diameter selected is a compromise. In locating the propeller in the aperture of a singlescrew ship, a clearance of 6 to 12 in. is normally provided between the propeller tip and the baseline with clearwater sterns-or to the rudder shoe with a closed stern (Fig. 4). With high-speed ships, which are generally characterized by shallow draft and multiple screws, propellers are often permitted to project below the base line in order to provide adequate clearance between the propeller and the hull. This is satisfactory provided
maxiinum draft limitations for service routes or drvdocking are not exceeded. One of the most effective means of ensuring a satisfactory level of vibration aboard ship is by providing adequitte clearance between the propeller and the hull . surface. For this reason, the subject of providing clearances is one of overriding importance. Generally speaking, the greater the clearances, the better the performance from a vibration standpoint. There are three types of vibratory forces generated by the propeller: (a) alternating pressure forces on the hull due to the alternating hydrodynamic pressure fields caused by the propeller blades; (b) alternating propeller shaft bearing forces which are primarily caused by wake irregularities; and (c) alternating forces transmitted throughout the shafting system which are primarily caused by wake irregularities. If the frequency of the exciting force should coincide with one of the hull or shafting system natural frequencies, very objectionable vibration can occur. A further breakdown of the forces generated by the propeller is given in reference [2]. When selecting propeller clearances, the perfprmance of similar ships should be an influencing factor. Of course, differences between the important parameters of the ships under comparison must be assessed. Important parameters to consider are the unit thrust loading on the propeller blades, number of propeller blades, amount of propeller skew, length of the ship, and t& ending angle of the water-plane forward of the propellet. References [3,4] discuss the influence of the more important of the foregoing parameters and summarize test and analytical data on this subject. Figure 4 may be used as guidance in assessing the aperture clearances of single-screw ships. Figure 4 shows the
Fig. 4
Propeller aperture clearances
. MARINE ENGINEERING athwai-tship rake angle, both ~f which are measured relative to the ship centerline, It is rare for 19 to exceed 3.75 deg or 4 to exceed 2.5 deg. From rake alone the reduction in propulsion efficiency will normally not exceed0.3pei-cent. Aside from the efficiency penalty, there is no objection to moderate amounts of rake. 2.4 Shaft Withdrawal. Occasionally shafting sections, particularly those outboard, must be withdrawn to be inspected or repaired. Consequently, provisions for removing shaft sections from the ship must be considered when developing a shafting brrangement. . On singlescrew ships with shafting arrangements similar to Fig. 2, the propeller shaft is almost without exception withdrawn inboard for inspection. If repairs are necessary, the shaft is removed from the ship by cutting a hole in the side of the ship and passing the shaft through it. This technique u.ould be used for removing line shaft sections as i-ell. Ftfi. 5 Ueamnm of a propelk supported by strut bearing With shafts having struts as shown in Fig. 1, a check ahaft can be must be'made to ensure that the withdrawn from the strut after the propeller is removed. range of eexperience wIdah hw been obtained in cannec- Withdrawal can be accomplished by removing the beartion dth I a q p single-screw &ips. When the propeller ing bushings so that the shaft can be inclined sufficiently is supported by a strut bearing, i.e., multiscrew and to aiIIow the forward end of the shaft to clear the ship's bmam-&ern vesseh, two clearance dimensions 11-amnt structure, mating shaft flange; etc. This consideration careful skudy. These dimensions rand the range of ex- can govern the length of the propeller shaft and the size of the stmt barrel. Figure 1 shotvs the removal position perience with them are shown in 1%. 5. 2.3 Shaft Rake. In order to provide ~atiiudein of the prcrpeller shaft., Removal of the stern tube shaft, which must hhve locating the position of the pmpeller and the main engines, it is usually necessary to wke the shaft cehterline. flanges on both ends, requires a decision regarding the The &aft is generally raked downward going aft as this type of flanges to be provided on the shaft. If the shaft permits the main engines to be located higher in the is manufactured with integral flanges on both ends, the ship. In mdtiscren- ships the shaft is generally raked in stern tdbe barrel and bearing bushings inust be sized to both the ve$ical and hori~onG1planes, usually donm- pass the flange diameter. Since i t is desirable to pass the shaft outboard, sufficient clearance should be proward and autbalbrd going aft. Large rakes s h a W be a~oidedsince a reduction in the vided to incline the shaft such that it will clear outboard propulsive efficiency is associated with rake. The intm- struts, etc. I n order to use smaller stern tubes and duction of rake incurs a reduction in the propulsion M a g bushings, the stern tube shaft can be manufactured with a removable Aange mupling on the forward efficiency equal t o erid. Prior to unshipping the shaft, the removable coupling is removed so that it is not neeesary t o disturb where B k. the shaft vertical rake angh and is the the stern tube bearings. (I
PROPELLRZS, SHAFTiNG, AND SHAFTiNG SYSTW VlSRAMON ANALYSIS
dimtly obtained from the min engine t o q u e and the p r o p e k thrust. On the other hand, vibrabry loads emanating from the propeller do not lend t h m m i ' ~ to s a precise evalwtion and are diieuit tn tseat in an absolute sense. 3.2 P r o p e l l d n d u d Loads. Aside from the alternating bending s h due to Ithe weight of the propeller, the ckumferentially n o n d o r m velocity of the water inflow to the propeller (wake) is the most important qouni.e of the alternating i d s in the shafting system. If is, howeverS important to $i&hguiish between the importance of the chumfereniaa nonuniforrnity of water i d o w a t a particular propeller radius and the nonunifomity of the average flow at, one radius as compared with mother. W e the former 1eads to vibratory propeller fothe latter does @. A propeller blade section w o r e in a constant d o c i t y field & a particular radius has a steady flow and force p-hkrni. The average axial velocity at each radius can be dierent without @awingalternating loads. I n such a ease the pmpeHq design can be adjusted for radial variations in the inflow vdocity to aehieve optimum efficiency. However, a propeller can only be designed t o satisfy average eond%ions a t each radius. Variation in the axial component of the inflow velocity a t each radius gives rise to the p e r i d c fluctuating forces genembd by the propeller. The variation in the inflow water =loci@ at a particular radius results in a change in the angle d attaek of the psopeI.3~blade sections as the propeller makes one ~vo1ution,thereby creating a1krnati.g propeller forces. Figure 6 b an example of the axial, VA,and tangentid, VT,i d o w velocities in the plane of the propeller for a single-screw ship. The tan- . gential velocity component is symmetric on both sides of the vertical mnterhe and is g&erally upward. The symmetry of the tangenttd ve1ociQ component would
+
rw VA V,
tend to suggest that its effect is uniform, but such is not the case. For a propeller bhde rotating clockwise looking forward, the tangedtiaJ veJoci*y component, eRwtively reduces the angle of attack on the bhde sections as they pass up the port side (reducing thrust) and increases the angle of attack of the blade sections as they pass down the starboard side ( i n c b n g thrust). Figure 7 iliustmtes how the variable axial and tangential velocities give rise t o variable loads. Also, another very important fact is t h t the tangential velocity components shift the center of propeller thrust to the shrboard side of the
speed of advance
= resultant of V, and V, = minimum advance angle ,P = maximum advance angle E . = variation in blade advance Ya
Section 3 f ftafting Leads 3.1 Dwkn Coaaidemt?ons. In general, the dimensions ~f s h d t b g are predic@ed on the basis of strength requirements; however, it is ocxasionally nebee+ eary to modify an otherwise sstipfactory shafting system design due to vibration considerations. Shafting &rametemurnally have only a minor impact on the longitudind vibration chmcteri~tim~ but the wh&:ling and torsional modes rtre sensitive t o &a& diameters. Shafting vibration, m mch, is d k u d in Sections 7-9.
Propulsion shafting is subjected to a variety of steady and alternating Ioads which induce torsional shear, axial thrust, and bending stressesin the shafting. In addition, there are radial compressive stressesbetween the shafting and mating elements (such as the propeller and sleeves) which, when coupled with axid strains from bending stress, tape v q important from a fatigue standpoint. The steady loads represent average conditions ~nnd ean be &mated w i a~degree of certainty as they are
P,,
Fig. 7 Typical warMan in udvdvence onale of a Made sedan &kg one revdution
Fspd ~ l o w o f w u t c r i n p ~ a n c o f p r o ~
blade tangential veloci'ty = axial speed of advance
= tangential
367
angle during one revolution
Locus of the resultant of tspeeds h e axirri tangential of and odvonce
I(
368
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRA'I'ION ANALYSIS
MARINE ENGINEERING
much greater for the five-bladed propeller than for the Table 1 Ratio of Shaft Torque Measured During High-Speed four or six-bladed propeller. For a single-screw s h i ~ Maneuvers to Normal Torque having a propeller with an even number i f blades, thk No. Torque Ratio fluctuating forces of two opposite blades give rise to a hi^^^^^ Shaft Inboard Outboard larger total t h r u ~ tand torque amplitude because op- ~~~~l 4 1.2-1.4 1.2-1.3 posite blades simultaneously pass through the slow water Naval 1 1.1-1.2 velocities at the top and bottom of the propeller disk. 2 1.2-1.3 1 1.1-1.3 The transverse force and bending moment,developed by one blade tend to be compensated by similar loads on the opposite blade. For propellers having an odd number of blades, the Table 2 Propeller Variable Torque Excitation Factors blades pass the upper and lower high-wake regions alternately. The total thrust and torque variations are No. of Torque Excitation Factor, r Propeller Blades 3 4 5 therefore smaller as compared with a propeller having an 0.07-0.12 0.10-0.15 0.06-0.10 even number of blades. However, due to the alternate Single-screw vessels Twin-screw vessels 0.02-0.05 0.02-0.05 0.02-0.04 loading of the propeller blades, the transverse forces and with struts bending moment do not cancel. Therefore, larger bend- Twin-screw vessels 0.04-0.08 0.04-0.06 0.04-0.05 with bossings ing moments occur with a propeller having an odd Note: Excitation torque = 4, where Q = mean torque. number of blades. The nonuniform character of the water inflow to the propeller can be resolved into Fourier components with the propeller rotational frequency (shaft frequency) as the fundamental [10]. Since it may be assumed that in power; this results in a higher shaft torque. As the linearity exists between inflow velocity variations and hull becomes foul, the ship speed reduces and full power propeller blade force variations, the Fourier components is developed a t a lower rpm; consequently, the torque of the inflow velocity are also the Fourier components loading on the shafting correspondingly increases. Such of force of a single blade making one revolution. Only torque increases are normally not considered in merchant those harmonics of loading which are integral multiples practice because merchant ships do not engage in extenof blade frequency ( M )contribute to the unsteady thrust sive high-speed maneuvers. The torque increase (which and torque, and only those harmonics of loading adjacent is relatively small) due to hull fouling is accepted as a to multiples of blade frequency (kZ f 1) contribute to reduction in the factor of safety. the unsteady transverse forces and bending moments [5]. The torque increases measured during trials of singleAll other harmonics of shaft frequency cancel when screw and multiscrew ships in high-speed turns are given summed over the blades. The selection of the number in Table 1. The torque ratio shown is the peak torque of blades can be based on the relative strengths of the value observed during steering maneuvers divided by harmonics in the inflow water velocity to the propeller the torque a t the start of the tests. to minimize the alternating thrust and torque and bendAlternating torsional loads on the shafting are gening moments. erated by the propeller and occur at predominantly blade Variable propeller forces, in addition to those resulting frequency as a result of the wake as discussed in Secfrom a nonuniform water inflow, are generated as a result tion 3.2. Although alternating loads can be generated of the proximity of the hull to the propeller. Hull by other sources, the propeller is the only one of practical surface forces generated by the propeller are of the ut- importance, except in diesel propulsion plants, where the most importance when evaluating hull vibrations. cyclic engine torque is significant. Shafting systems are 3.3 Torsional Loads. The torsional load. on the carefully designed to avoid torsional resonant frequencies shafting, which results in the steady torsional stress, is a t full power; therefore, alternating torsional loads are calculated from the output of the main engine. If the not congidered to be amplified by resonance. The range full-power shaft horsepower output, shp, of the main of magnitude of the forced torsional alternating loads is engine is developed a t N rpm, then the steady torsional given in Table 2. I t will be noted that the variable load, Q, on the shafting is: torque can be of a significant magnitude even without magnification. 3.4 Thrust Loads. The magnitude of the steady thrust load on the shafting system is equal to the towed I n the design of naval shafting systems, it is common resistance of the ship a t the speed corresponding to maxipractice to increase the torque calcu!ated with equation mum design power, corrected by the interacting effect (2) by 20 percent. The increase in design torque is an of the propeller and hull as the propeller pushes the ship. allowance in recognition of the additional torque de- This interaction effect is known as the thrust deducveloped during high-speed maneuvers, rough-water op- tion [I.]. The value of the design thrust can be obtained erations, foul-hull conditions, etc. During turns, the from the powering calculations or from model basin tests propeller rpm reduces without a corresponding reduction of the ship. For preliminary design purposes
-
.&
I 0'1
E
I
I
90. (a1
180. PROPELLER POSITION
270.
360.
h w
( c ] VERTICAL BENDING MOMENT IN PROPELLER SHAFT (PROPELLER WEIGHT INCLUDED)
$
a w
PROPELLER TORQUE VARIATION
E~:EE
PROPELLER POSITION
z
+
!j I5 J
I
I
3+10 4
K w
p+
5
I
0
I
0.
90.
180' 270. PROPELLER POSITION
360'
( d l HORIZONTAL BENDING MOMENT IN PROPELLER SHAFT
0 I'.
5
0.
0
w
a- 5
i*'
b
,
'b'
L"
I
0.
'v' I
90'
180' 270' PROPELLER POSITION
360.
(b) PROPELLER THRUST VARIATION
B
Fig. 8
I I
i/
Typical single-screw propeller alternating thrust, torque, and bending moments from nonuniform water inflow velocities
propeller centerline of a clockwise-turning propeller on a of the propeller, the thrust and torque can be determined single-screw, ship. This off-center thrust gives rise to a and plotted as shown in Fig. 8. If the K r K e - J diagram bending moment which is imposed upon the propeller is not available for the propeller, the step-by-step calculation in reference [7] can be used. This method is shaft. Analyses can be made to predict the magnitude of the based on that given by Burrill in [8]. Since the slowest axial inflow velocity (highest wake) alternating components of torque and thrust including the eccentricity of the resultant thrust relative to the of single-screw ships is generally in the region above the shaft centerline [5]. Four basic methods are available to propeller centerline, the greatest thrust tends to be calculate the unsteady forces and moments on marine developed when the propeller blade is in the upper part propellers aaused by circumferential nonuniform inflow. of its orbit. The effect of the tangential inflow velocity These are quasi-steady, two-dimensional unsteady along is to shift the resultant thrust to the starboard side bea strip, combination quasi-steady two-dimensional un- cause the propeller blades develop greater thrust moving steady along a strip, and three-dimensional unsteady. against the tangential velocity, as discussed in the foreEither the quasi-steady or the two-dimensional unsteady going. This subject is given a detailed discussion in technique may be used to obtain approximate estimates reference [9], and it is noted that as the shape of the stern of the fluctuating thrust and torque; but if close predic- sections change from a V to a U shape, the resultant tions are required the three-dimensional unsteady tech- thrust center tends to move down because the inflow nique should be used. Application of the quasi-steady velocities over the bottom region of the propeller disk method is much simpler than the three-dimensional un- become more nearly equal to those in the upper region. steady approach. Due to its simplicity and the fact that The position of the resultant thrust is also sensitive to it produces results which are generally accurate enough the ship's draft. For instance, when a cargo ship operin a relative sense for most practical applications, the ates lightly loaded with the propeller blades breaking the water surface, the center of thrust obviously shifts quasi-steady approach is a very useful method. A quasi-steady analysis is conducted by making an lower in the propeller disk. Figure 8 shows that a single-screw ship with a four or instantaneous examination of the flow velocities relative to the propeller blades a t discrete angular positions of a six-bladed propeller (that is, an even number of blades) propeller blade [6]. The inflow velocities are regarded has larger torque and axial thrust variations than one as constant (quasi-steady) at each blade position. By with a five-bladed propeller. However, the thrust eccenusing the open-water characteristics (KT&-J diagram) tricity (propeller shaft bending moment) is shown to be
1 t
,'
369
-
PROPELLERS, SHAFI'ING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
Table S
Propdler Variable Ruwst Exc8atiorr F a c m That EHcitation No: Bk&s ~EEEW~-&~W k B h Frotar, J Behind struta O.W.85 %%or& Behind ekege 0 . ~ 4I% . 3 4ar5 w i d akee 0.05-0.W R Behind hossinea O.WU.12
Behind b & i
Behind bmaing~t
0.w.10
Q.W.08
Q.OEHl.12 a U-seotine tend to emphsiae the e v e m d e r compomntg and B-wotiona the & d e r eamponents. Nate: M a t i o n thrask = p,where T = m e n thru&. s
sional the imps-ce of the k t stis reduced even furthepep 3.5 B e d L d s . Lmds which cause bending &mes to omur in the shaP%iqm e the d t of gravity, shmk, off-center t M loarfsBand whirling shaft vibmtion. With the exceptionof once-per-revolution whirling vibration. d are &ernratinn hack relative to a point on t i e shaft'a~~cl occur at e&h& shaft rotative fre&eney or Qnce or twim propeller blade frequency. The weight of the shafting iW (a gravity l o d ) * k
are
shaftini unless there =US& w*ht ~ ~ ~ l ~ e f t t m t i ~ ~ , such as a S h d t locking device or brake d r m , or an exceptionably long span between bearings. When the shaft mans beheen bearilarrs are essentially equal, the maximbrn static bending moment a e c t m b ~ aihe t shaft bearings as a result of shaft weight cae be d e f e r m i d
Table 4
Ship Name OWsp Chrysai Robimon Jamestwn Observaiion zslund
II
heh hue
of f mnges from about 0.16 to 0.23 for singIescrew &hipsv e n g frarn fine to full lines, respectively. Twin-screw ships have t values ran& from about 0.1 t o 0.2,2,& larger value applying t o &ips with bossings and tihe sm&w value mmspon&ng to &ips with struts. PC values of 0.73 for Singlewrew ships and 0.68 for mdtiscrew ships sre average v a b and normally found to be suitable for preliminary estimates. Reference tII contains methods of estimating t and PC and should be consulted €0 obtain a more a m r a t e &mate of the thrust I d if m d d test results a~ ship's performance etiYenleiti0~~ are n ~ avaihble. t Alternating thrust h a & ltre generated by the propeller. The p d o m i n m t dtemating t h r a 1-d o m m at pmpdler bMe frequency abi a consequence of the nonunifwm inflow water velacity 1;6 the propeller as discussed in Election 3.2, The magnitude af the variable thrusC M a is dependent an the ntmbes of prnpeller blade& For single-mrew ships, an even number d bJr8des will m l t in greater d t e m t i q thrust lmds than am odd n u m k as dimussed in 3.2. For preEminmy e&imates, fhe magnitude of the alternating thrust as a percentage of steady thrust can be taken from Table 3. Insofar tm the strength of the shafting is concernedr neither the steady nor J k m a t i n g thrust lmds are major design consi~ations. With merchant ships, the steady camp&ve stress is 1008 to I500 @; even In highly stressed shafts in navd skips the steady e o m p d v e stress d d o m reshes 2500 psi, Torsional shear stresses me of predominant importance; and since the s t m s due to thrust EEO nut combine additively with the tar-
Tanker Mariner
20,300 38,100 15,200 50,200
16,400
2
M,
=
Ib/h.
@ are mt approxirnateIy If the spans bekeen equal, such a simple appmmh cannot be used; instead, continuously supported beam analytieal techniques, such a5 the three-moment equation or Hardy C m methad, must be used. During recent y e w , the practice has been to use the digital mmput;er to mIculate the bending moments a t all efiticaI &a& sections, utilizing mntinuous b a r n forrnuIa%ians. Weight 1 4 s on the outboard shafting tend to be of more imporbnee due to the large mncentraW weight of the propeller. Standard eonfirnous beam equations ean be used ha determine the magdude of the bending moments. Howeuw, because of the long bearings used o u t b d , the lacakians of fhe bearing peaetions we not cleady defined. The pactice is ta assume the reaetion a t the e n t e r of d1 bearings except the bearing just fopward of the propder. Because af the ham weight of the propelIer, the pmpeller shaft has a s i ~ i 6 c a n slope t a t this bearing; therefore, the rwltanh W i n g reaction tends to be in the after region af the Wing. WaterIubricated bearings of Emurn vitae, micarby or rubber have L / D ratios of about 4 f o r this bearing and the resultant reaction is usually assumed to be one shdk diameter forward of the aft bearing face. Od-Eubrieated bearings h&veLID ratios of about I to 2, and a review of the shaft contact in these bearings indicates that hard contact is confined to the after region of the bearing for a length appmximateIy eq;uaI to the diameter of the shaft. Current practice is t o assume that the resultant bearing reaction in oil-lubricated b e d g a Is one-half shaft diameter from the after bearing face. Generafly the most signifimnt weight moment ia due to the overhtmg moment of the propeller. The maxi-
Thrust,
Thrust Eccent. C, it
Thrust
Eccent. Factor C / D Reference 0.047 [11]
shp
rpm
503 68 615 84 436.5 62
0.74 0.77 0.69
10,000 15,000 8,500
95 112
184,000 246,000 156,000
19.5
0.91
22
20.5
1.31 1.67
0.060 0.082
B2] [13]
93 76
0.75 0.60
26,500 22,000
108.5
385,000 271,000
23 22
1.26
0.055 0.076
1141
685
528
WpLp
85
110
Table 5
(5)
where =
Equation (5) is the moment at the bearing reaction point assuming that the reaction is a point support rather than a distributed reaction over a region of the shaft. The point support assumption is justified in that the exact load distribution on the bearing is unknown and the moment calculated in this manner is somewhat in excess of the actual value when the position of the resultant reaction is estimated reasonably well. There are a number of influences in addition to the gravity moment of the propeller which can have a significant impact on the propeller shaft bending stress. These are the eccentricity of thrust, water depth, sea conditions, and ship maneuvers. Under the general guidance of SNAME Panel M-8, the propeller shafts of a total of five ships have been instrumented to measure the bending stresses under actual operating conditions. Data obtained from these tests are reported in references [ll-151. Table 4 summarizes the characteristics of the ships tested. The tests were conducted to show the significance of the ship loading, sea conditions, ship maneuvers, and thrust eccentricity. Eccentricity of the propeller thrust produces a significant propeller shaft bending moment. The propeller resultant thrust is eccentric from the propeller shaft centerline under almost all operating conditions and is usually in the upper starboard quadrant when looking forward. Therefore, it does not combine directly with the propeller gravity moment. Light draft operating conditions and "U" shaped stern sections tend to bring the thrust and gravity moments closer together and make them more additive. Table 4 shows the thrust eccentricity factor, C / D , determined from full-scale test data for heavy-displacement, calm-sea conditions. The thrust eccentricity, C, shown in Table 4 is the resultant of the eccentric thrust and the gravity components.
ib
Prop. Dia. D, ft
cn
A
propeller overhung moment in propeller shaft, in.-lb W , = weight of propeller assembly including shafting aft of reaction point, Ib L, = distance from CG of propeller assembly to aftermost bearing reaction, in.
M L
= bending moment at bearing*in.-lb = span between in. w = weight per unit length ef shaft&
Rhin ."---=
Lyl
Design
mum static propeller shaft bending stress is computed as
M,
V = ship s p d at maximurn power, knots W = ship's resistance at Tr, Ib ehp = ship's bull &wGm horsepowe~a t TP, hp s h= ~ maximum &aft horsemwer, h~ *t =. thnrst cE&uction fraction PC = propulsive coefficient
Type Ship TZSE-A2 Tanker - -. . --. Tanker Victory
Ships Instrumented to Determine Tailshaft Bending Stresses
Load Heavy Heavy Heavy Light Light Light Heavy Light Heavy Light
1.M
[16]
Increase in Propeller Shaft Bending Stresses Due to Various EfFects
Sea Condition Calm Calm Calm Calm Calm Calm Stormy Stormy Calm Calm
Operation Ahead Ahead Maneuvering Crash Back Ahead Ahead Ahead Ahead
Water D e ~ t h Factor
Deep Deep Deep Deep Shoal Shoal
3% 9%
2%
1% 1%
lhll-scale tests on the Esso Jarnestown I141 permit an evaluation of the influence of ship loading, sea conditions, and maneuvers. These factors are summarized in Table 5. The factors presented in Table 5 are the ratios of the bending stresses for the various conditions described to the bending stresses under full-load, deepwater, calm-seas, and straight-ahead operations. The extrapolated results from the Observation I s l a d tests [15] generally support the factors in Table 5. It should be noted that maneuvers such as era&-backs rarely occur, and that the shaft need not be designed to withstand stresses three times the normal value on a continuing basis. Shock loadings, considered in designing naval shafting for combatant ships, are akin to the gravity loading and are frequently determined by multiplying the gravity force loads by a "shock" factor; however, more sophisticated methods are available for determining the shock loads through the application of dynamic analysis techniques. References [16,17] treat the procedure for conducting dynamic shock analyses of shafting systems. Misalignment in shafting systems can produce very significant bending loads and this factor is probably responsible for the majority of inboard shafting failures. The sensitivity of the shafting to misalignment should be reviewed particularly as regards water-lubricated stern tube and strut bearings which are subject to wear in service. The sensitivity of the shafting to misalignment can be assessed by calculating the shafting bearing reactions and moments with the shafting in various misaligned condi-
372
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
MARINE ENGINEERING
I
'
tions. Digital computer programs have been developed for making this analysis; one such pmgram is described in Section 5. Lateral or whirling vibration of the shafting can result in inqreased bending loads in the shafting. However, since the shafting system is designed to avoid whirling criticals in the upper operating range, bending loads from shaft whirling vibration are not considered when designing the shafting. 3.6 Radial Loads. Radial loads in shafting are caused by driving the propeller onto the shaft taper, shrink-fitting sleeves on the shafting, and shrink-fitting removable flange couplings. The radial compressive stresses resulting from these loads are normally of in-
significant magnitude and are not considered in determining the shaft factor of safety. However, these radial loads can be of importance in that they give rise to fretting corrosion when coupled with bending loads and alternating torsional loads that cause minute relative movement of the mating surfaces. Fretting corrosion can be controlled by limiting the relative motion and by cold-rolling the mating shafting surface. Cold-rolling of shafting surfaces is discussed in Section 4.3. Another consideration is that if the radial load is applied abruptly, a stress concentration can occur. Therefore, design details should be developed so as to minimize sudden changes in radial loads caused by shrink or press fits.
Section 4 Shafting Design 4.1 Shaft Materials. With the exception of naval vessels and merchant vessels of very high power, mild steel is used for both inboard and outboard shafting. In the case of high-powered ships, the inboard shafting may be made of high-strength steel; however, high-strength steel is not recommended for outboard applications. Because of the seawater environment and fretting corrosion conditions that exist at shaft sleeves and the propeller interface, the fatigue limit of high-strength steel is not reliably greater than that of mild steel, nor is the endurance limit in a fretting corrosion condition better than that of mild steel. Considerations in the selection of shafting materials are: fatigue characteristics, weldability, the nilductility temperature, and the energy absorption capability. An array of chemistry and physical property standards has been established for marine shafting materials that provides a range from which shafting materials can be selected. Chapter 22 contains more specific information regarding shafting materials. 4.2 Computation of Shaft Diameters. Shafting for merchant vessels is required to meet the minimum standards set by the classification society which classes the vessel. Classification societies use rather simple formulas to compute the minimum shaft diameters. These formulas normally contain coefficients which are changed from time to time in recognition of experience or advancements in technology. The American Bureau of Shipping (ABS) line shaft diameter formula is of the following form [IS]:
M
d = C(K?) where
d
=
minimum line shaft diameter, in.
c = constant for type of shaft = constant relating to operating environment
K
It may be noted that equation (6) neglects bcnding loads, alternating loads, and stress risers. Furthermore, it presupposes that the shift will be a solid forging of mild steel and that no unusual circumstances exist. However, despite the factors not explicitly considered, equation (6) does provide,a sound basis for designing line shafts. This is because the predominant torsional shear stress is properly considered. The level of torsional shear stress corresponding to equation (6) can be determined by observing that
By substituting equation (6) into equation (7) and setting K = 64, which is the case for oceangoing vessels
If c were specified as 0.95 (as was the case for a period before 1965), the corresponding allowable steady torsional shear stress would be 5849 psi; with a c value of 0.875 (which was adopted in 1965), the corresponding torsional stress becomes 7486 psi. When the allowable torsional shear stress of 7486 psi is compared with the material minimum tensile yield stress of 30,000 psi, it is seen that adequate margin is allowed for the secmdary duences which appear to be neglected. Only vessels to be navigated in ice require special consideration. (6) The stern tube shaft diameter required by ABS for merchant ships is directly proportional to the line shaft diameter. The propeller shaft diameter is related to the line shaft diameter and the diameter of the propeller. The ABS propeller shaft diameter formula is of the following form :
T
=
kld
+ -PC
where
T = minimum propeller shaft diameter kl = constant d = required line shaft diameter P = propeller diameter C = constant relating to propeller shaft environment
Although the formula for the tailshaft diameter is simple, it does recognize that bending stresses from propeller weight and off-center thrust exist and relates these important factors directly to the propeller diameter. It has been expressed by some designers that classification rule$ for tailshafts are not adequate because ,the level of bending stress is not controlled [19, 201. A designer should review his own design against the important parameters and against his experience. Classification rules should not be accepted without question, particularly for unusual designs since the formulas are set primarily by past experience. In any case, the propeller shaft is customarily manufactured with a diameter approximately 3 percent larger (10 percent stmnger) than the minimum classification requirements in order to provide an additional margin of safety as well as to make provisions for removing a small amount of surface metal in the event the shafts become superficially damaged. Reference [21] reports on a service life comparison of 15 oversized shafts (74 percent greater section modulus than required by ABS) with 15 shafts of normal size (11.5 percent greater section modulus). The comparison showed that the mean expected service life of the oversized shafts was less than the shafts of normal size. Although the statistical sample was small, the study clearly showed that propeller shaft problems are not necessarily solved by simply making the shaft larger. The approach used to establish the size of naval shafting is considerably different from that used with merchant shafting. The procedure used to determine the size of naval shafting is delineated in reference [22]. As may be noted from the reference, an effort is made to assess all significant shafting loads in each particular case, although some loads are by necessity handled in an approximate manner. For example, in order to allow for the effects of off-center thrust and abnormal loadings due to rough weather and the like, the propeller shaft bending stress due to the static weight of the propeller is multiplied by a factor of 3 for single-screw ships and 2 for multiple-screw ships. An additional difference between merchant and naval procedure is the criteria of acceptance. In naval practice, dual criteria are used. Factors of safety are specified for.al1 shafting and, in addition, a specific bending stress limit is specified for the propeller shaft. The reason for the latter requirement is that fatigue tests run on models of propeller shaft assemblies and crank pins [13] showed that bending stress levels in surface rolled (cold-rolled)
373
shafts in excess of 6000 psi would result in shaft cracks. Therefore, it is not prudent to design with normal operating bending stresses in excess of this stress level. Furthermore, the endurance limit of a propeller shaft assembly can be essentially independent of the fatigue limit of the material in air. If seawater contacts the steel shaft, no endurance limit exists and it is ohly a matter of time before cracks will occur followed by ultimate failure. 4.3 Bearing Locations. In the past, bearing locations have been determined by criteria such as "each shaft span shall have two bearings" or by intuitive judgment. With these criL-?ria.~roblemsdue to unload in^ of bearings, excessive rates bi weardown, shaft whirling, and gearing misalignment were not rare. Problems were frequently related to the system having too many bearings. In order to better understand the optimum locations for bearings, designers began analyzing shafting as a continuous beam. However. the time reauired to analyze one shafting system confibration by hand calculations precluded complete analyses. The development and general dissemination of digital computer programs, such as reference [23], made it feasible to routinely conduct in-depth studies to optimize shafting systems as well as diagnose recurring problem areas. Factors to be considered in determining the number and location of shaft bearings - are: 1 2 3 4 5
Ship's fixed structure and arrangement. Equality of line shaft bearing reactions. Bearing unit loads and LID ratios. Shafting flexibility./ Lateral vibration natural frequencies (shaft whirl).
Ship's fixed structure such as bulkheads and stanchions will usually require compromises in the shafting arrangement. Also maintenance and overhaul must be considered before final bearing locations are set. From a cost and interchangeability standpoint, all line shaft bearines should be identical. Therefore. the bearings should bve spaced such that the bearing reactions are approximately equal. If this is done, the total number of bearings in the run of shafting is set by the total shaft weight, permissible design unit load, and the acceptable LID limits. The number of line shaft bearings required to support a run of shafting can be tentatively determined a s follows:
where
W
total weight of shafting to be supported (note that gear and stern tube bearings may carry some line shaft weight) p = design bearing pressure (maximum permissible pressure less 5 to 10 psi to allow for variations) based on projected area D = shaft diameter in way of journal (normal practice is to increase the shaft diameter %-% in. in way of bearings) =
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS '
MARINE ENGINEERING Table 6
Bearing Bearing No.
Fwd SlowS eed Bear 1
Table 7 Tabulation of Bearing Reactions for the Shafting Arrangement Shown in Fig. 1
Bearing Reaction Influence Numbers for Shafting Arrangement Shown in Fig. 1 (Pounds per Mil of Vertical Displacement) After SlowS eed Bear 2
Fwd Line Shaft 3
Line Shaft 4
Line shaft 5
Line Shaft 6
Aft Line Shaft 7
Stern Tube 8
375
Strut 9
Bearing No. Reaction Line-in-Line Cold (as aligned) Hot (as aligned) Weardown Condition5 Measured (Hot) ~
Fwd SlowS eed 8ear
Aft SlowS eed Bear
Fwd Line Shaft
Line Shaft
Line Shaft
Line Shaft
26900 41700 36300 36300
49500 29700 37300 37100 38500
23200 29600 26500 26700 26700
25500 23700 24900 24300 26400
25200 25700 25400 27200 26300
27200 27000 27100 20100 30300
1
2
3
4
5
Line Shaft '7
6
Stem Tube
Strut
8
9
~
...
.
15900 15900 15900 30800 11200
.
62600 62600 62600 51000
...
.
89900 89900 89900 '92200
'
...
a Assume a strut bearing weardown of 0.200 in. and a stern tube bearing weardown of 0.185 in. (consider weardown proportional to bearing pressure).
Notes: The numbers tabulated above represent the effect of raising a iven bearing one mil; e.g., if the forward line shaft bearing is raised one mil, the forward slow-speed gear bearing reaction increaaea 336 lf~,the after slow-speed bearing reaction decreases 493 lb,the forward line ahaft bearing increases 259 Ib, etc.
LID = bearing length/diameter ratio After tentatively selecting the number of bearings and 'pacing them a of the bearing loads under all normal operating conditions is made. The conditions of primary tance are the cold starbup condition, the hot operating condition, and bearing weardown and misalignment conditions. Involved in this analysis is the influence of the shafting on the reduction gear bearing loads or diesel engine bearing loads resulting from the themla1 change the in the position of these when going cold to the hot operating condition. Criteria for aligning the propulsion unit to the shafting are developed on the basis of this analysis. A parametric study of minimum line shaft bearing 'pacing was carried and in reference [241' The conclusion n3ached was that for shafting arrangements having one or more line shaft bearings the minimum span ratio (i.e., ratio of bearing center distance to shaft diameter) should be 14 for shafts with diameters in the range of 10 to 16 in. and 12 for shaft diameters of l6 30 in' The 'pan ratio be in the range of 20 to 22 but the final determination must be and vibraat the based On strength, shaft tion characteristics. 4*4 shafting Calcu'ation Output. The important output from most shafting calculations includes the following in addition to the required shaft diameters:
(a) Line-in-line reactions. (b) Slope of shafting a t discrete poids. (c) Deflection of shafting a t discrete points. (d) Moments in shafting a t discrete points. (e) Lateral natural frequency of shafting. (f) Bearing reaction influence numbers. The significance of the shafting line-in-line reactions (bearing reactions with all bearings set concentric), shaft slopes in way of bearings, shaft deflections, shaft moments, and lateral natural frequency of the shafting is
apparent; however, the importance of bearing reaction influence numbers is not as readily appreciated. Table 6 ;s a tabulation of the bearing reaction influence numbers for the shafting arrangementshown in Fig. 1. The numbers given ih Table represent the change in the magnitude of the bearing reaction of the various bearings as a result of raising any bearing one mil. Thmugh the application of these influence numbers, which reflect the shafting system flexibility, it is possible to investigate the influenceof shafting misalignment caused by thermal expansion, weardown, and &her such effects.Alignmentrequirements are developed on the basis of the Also, the principles bearing reaction inRuence employed with the hydraulic jack method of checking shaft alignment ,(see Section 4.12) originate with the numbers. bearing reaction 4.5 Gear-to-Shafi Alignment. Particular care must be taken in selecting the aftermost and forwardmost line shaft bearing locations to ensure that adequate shafting flexibility is provided. These bearings are subjected to a varying alignment in service. Weardown of the stern tube bearing alters the load on the aftermost bearing; the thermal expansion of the reduction gear structure changes the load on the foMiardmost bearing. These * effectsmay be noted from the tabulatioh given in Table 7. when the propulsion plant goes from the cold to the operating condition, the slow-speed gear besings may to 30 mils relative to the line shaft bearings. rise his rise can significantly alter the reactions of the slowspeed gear bearings and the forward line shaft bearings. Of particular concern is the fact that the static load on the forward slow-speed gear bearing decreases while that on the after bearing increases,. As can be seen from the typical reduction gear bearing reaction diagram shown in Chapter 9, this causes the slow-speed gear to assume a crossed-axis position relative to the slow-speed pinions which are not similarly affected. As a result, the tooth load will tend to be more heavy on one end of each helix. Reference [25] contains a detailed discussion of the effects
of unequal gear bearing reactions on gear performance and the maximum permissible differences between the gear bearing static loadings. The gear manufacturer should state the maximum difference permitted between the static loadings on the slow-speed gear bearings (see Chapter 9). The alignment in the athwartship direction should be such that no significant forces are imposed on the slowspeed gear bearings in the horizontal plane. The allowable setting error is conveniently used as an index of shafting flexibility in way of the reduction gear. The allowable setting error, ASE, is defined as the allowable difference in the static vertical gear bearing loads divided by the difference between the bearing reaction influence number of the forward slow-speed gear bearing on itself and the after slow-speed gear bearing on itself. Therefore, the ASE is determined as follows:
*ASE =
AR I11
-
I22
(11)
AR = allowable difference between two slow-speed Ill
I22
gear bearing static reactions reaction influence number of forward slowspeed gear bearing on itself = reaction influence number of aft slow-speed gear bearing on itself =
The ASE number represents the total of the error permissible in estimating the thermal rise of the slowspeed gear bearings relative to the line shaft bearings and the error permissible in setting the gear to the line shafting without exceeding the maximum allowable difference in the static slow-speed gear bearing reactions. An absolute minimum acceptable value for the ASE has been recognized to be f.010 in.; see reference [26] for additional discussion of this parameter. If the flexibility of the shafting meets the ASE criterion, the analysis proceeds to an investigation of the gear-to-shaft alignment. Beginning with the line-in-line reactions, that is, the bearing reactions with all bearings concentric, and with the estimated thermal rise of the gear bearings relative to the line shaft bearings when going from the cold to the operating temperature, alignment data are established which will provide approxi-
mately equal slow-speed gear bearing static reactions when in the operating condition. I t must additionally be ascertained that the line shaft bearing reactions are satisfactory under all operating conditions. Since the actual positions of the slow-speed gear bearings are difficult to ascertain relative to the line of shafting, their positions are determined by measuring the drop and gap of the slow-speed gear shaft flange";elative to the line shaft flange. Drop is the vertical distance between the centers of two adjacent flanges; gap is the differencein opening between the top and bottom of the two flanges (nonparallelism of the flange faces). It is, of course,-necessaryto know the position of the line shaft flanges relative to the line shaft bearings. One alignment technique is to support the line shafting sections so that the centers of the flanges are concentric with the bearing centerline and the flange faces are perpendicular to the bearing centerline; this is accomplished by s u p porting the sections at approximately the 2/9 points from each end. When this is done, the drop and gap are measured directly a t the flanges. The positions of the slow-speed gear bearings $elative to the centerline of the line shaft bearings are readily determined from the drop and gap of the slow-speed gear shaft flange relative to the line shaft flange; this is accomplished by means of simple geometry. With these data, the bearing reaction influence numbers can be used to plot the bearing loads for various alignment conditions. Such a plot is shown in Fig. 9. Figure 9 is an informative means of illustrating the effect of thermal rise of thevslow-speed gear bearings and the effect of alignment errors in addition to bearing loads. Table 7 shows the cold alignment bearing reactions with an alignment corresponding to point A on Fi& 9. The hot reactions listed in Table 7 are the bearing reactions estimated after the gear has reached operating temperature; this is point B on Fig. 9. Finally, bearing reaction influence numbers provide a means to study the effects of bearing movements from hull deflections and bearing weardown. An analysis of bearing reactions with the stern tube and strut bearings worn down is given in Table 7. 4.6 Propeller-to-Shaft Interface. Design details of
I'
I I
I
I
I
Fig. 9
I
Gear-to-shaft alignment analysis
i'
the propeller-to-shaft interface are a critical aspect of a shafting system design. DuringWorld War I1 and earlier years, propeller shaft failures in way of the propeller were not rare; and difficulties have been experienced in more recent times [20, 211. However, the advances in design technology (e.g., stress relief grooves a t the forward end of the propeller and the aft end of the liner, shortened-and spooned keyways, slotted keys, and improved sealing methods) have significantly improved the reliability of propeller shafts and increased their service lives. Also, improvements in inspection technology have provided the means to detect incipient cracks and thus have greatly reduced the loss of propellers at sea. Details of the propeller-to-shaft interface required for naval ships are specified by reference [27]. The naval type of propeller-to-shaft interface is consistent with merchant praltice and is reported to have a comparable service history. The propeller keyway is a stress concentration and a weakening factor to the shaft even though the keyway has generous fillet radii and the forward end of the key is slotted to relieve the key load at the forward end. For this reason, propeller nuts have been developed which incorporate annular pistons moved by hydraulic oil or grease. These "hydraulic" nuts provide the means to apply large forces of known magnitude to the propeller, pushing it onto the shaft taper such that no propeller key is required. Keyless propeller designs rely entirely on the friction between the hub and the shaft to withstand the propeller torsional and thrust loads. Conse-
quently, the contact pressure, material stresses, shaft taper, propeller-shaft interface friction coefficient, and the push-on force must be thoroughly engineered. Although some keyless propeller designs use oil pressure to expand the propeller hub while the propeller is forced onto the shaft taper, most designs call for the mating surfaces to be thoroughly degreased before the propeller is fitted to the shaft taper to ensure that the highest coefficient of friction is obtained. However, it is maintained by some that when oil is used to expand the hub for fit-up, shortly after the pressure is relieved the oil layer is squeezed out of the propeller-to-shaft interface, resulting in a friction coefficient equal to that when the propeller is pushed up dry. Most designs use a hydraulic pressure in the propeller-to-shaft interface a t least equal to the calculated radial pressure to expand the hub for removing the propeller. For a dry, greaseless, installation the coefficient of friction may vary from a low of approximately 0.13 to a high of approximately 0.1s. The value of the effective friction coefficient is dependent on the percentage area of the mating faces which are in contact at initial fit-up (prior to forcing the propeller up the shaft) and the ma-, terials in contact. Some keyless propeller designs incorporate a cast-iron sleeve in the propeller hub bore because tests have shown that higher coefficients of friction can be obtained this u-ay. Using an intermediate nodular iron sleeve with an initial surface contact area between the sleeve and shaft taper of 95 percent or more, a coefficient of friction of 0.18 can be achieved. When the
t
1
I
1
I
I I-
initial surface contact area is approximately 50 percent, a coefficient of friction of 0.13 can be expected. When a bronze alloy propeller is fitted to a steel shaft without using an intermediate cast iron sleeve, the maximum coefficient of friction obtainable at the propeller-shaft interface is a~~roximatelv 0.15. The shaft taper used with keyless propellers generally falls within the range of 1:12 to 1:20. The shaft taper should be selected such that the sum of the maximum astern thrust and the axial push-off force due to the taper does not exceed the axial frictional resistance. This consideration will limit the maximum shaft taper. On the other hand, if the shaft taper is too small, the ahead thrust may force the propeller up the taper and overstress the hub. 4.7 Cold-Rolling. The development of fretting fatigue cracks in propeller shafts a t the forward end of the propeller hub and a t the after end of the shaft sleeve is one of the most common modes of propeller shaft failures. Although surface cold-rolling will not eliminate the occurrence of fatigue cracks, cold-rolling of propeller shafts for a distance forward and aft of the forward end of the propeller shaft taper and in way of the ends of the liners has been shown to be an effective means of retarding the propagation of fatigue cracks [28]. 4.8 Protection from Seawater. Except in the case of designs in which all bearings are of the oil-lubricated type, outboard shafting involves the use of sleeves which are shrunk on the shafting in way of bearings, stuffing boxes, and fairings. Shaft sleeves are made of bronze or other materials which are resistant to attack by seawater. Ships having a single short section of outboard shafting employ a single continuous sleeve. Where continuous sleeves are not used, the sections of shafting not covered -by sleeves are protected by applying a rubber [29] or plastic [30] compound directly to the shafting surface. The adequacy of both rubber and plastic protective coverings for outboard shafting has not been unifornily good. The protection offered to outboard couplings by such coverings has been particularly unsatisfactory on occasions. Rotating coupling covers (fairwaters), which clamp onto and rotate with the shaft, thereby eliminating the violent erosive flow of water around coupling bolts, have been used to avoid the deleterious effect of the water. A reliable static sealing arrangement a t the propeller, which prevents seawater from contacting the propeller shaft, is of the utmost importance. A propeller-shaft assembly which allows seawater to contact the shaft will not have an endurance limit and therefore it is only a matter of cycle accumulation before a failure occurs. Details concerning propeller-hub sealing arrangements, which are necessary with systems utilizing water-lubricated stern tube bearings, are contained in reference [31]. 4.9 Shaft Couplings. Except in instances where special considerations preclude their use, shafting sections are connected by means of integrally forged couplings as illustrated on the line shaft section shown in
,,,
TYPICAL BOLT
SHAF
FT
LSPLIT COLLAR Fig. 10
Removable flange coupling
Fig. 10. Although the design of virtually all integral shaft couplings is similar, details of shaft coupling designs can vary considerably. For example, despite individual preferences, no specific number of coupling bolts has been established as optimum and the proportions of flange dimensions may vary from one design to the next. Guidelines for the design of flange couplings for merchant vessels are given in classification society rules such as reference [18] and similar guidelines for naval vessels are given in references [22,27]. An effort to standardize shafting couplings was made in reference [19], and SNAME T&R Bulletin 3-20 (Guide for the Design of Line Shaft Couplings) was subsequently prepared. Couplings with removable flanges are required in some instances; for example, those cases where a liner must be installed on a shaft which requires a bolted flange on each end. Figure 10 illustrates a typical removableflange coupling and shows the means provided to transmit both thrust and torque. Both torque and thrust are normally transmitted by friction between the shrunk-on muff and the shaft. The keys are a backup for the transmission of torque and the split collar is a backup for the transmission of thrust. On occasions shafting arrangements are designed such that it is necessary to remove the forward flange of a stern tube shaft in order to withdraw the stern tube shaft aft; this is not a preferred arrangement because it is most difficult to remove a flange without damaging the flangeshaft interface. A preferred arrangement is one in which the stern tube bushings are made sufficiently large so that they can be removed to permit the stern tube shaft to be withdrawn aft with the forward flange in place. 4.10 Shaft Axial Movements. Axial movement of the shafting relative to structure that is fixed to the hull must be considered to establish proper clearances between the propeller and stern frame structure and the clearances between bearing housings and rotating elements secured to the shaft within bearings (oil slingers, oil disks, etc.) There are four factors which contribute to the movement of the shaft relative to hull structure; these are:
378
MARINE ENGINEERING
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANAtYSlS
effect a static balance. Good practice dictates that adjoining shafting sections be installed such that the residual static unbalance, as determined by a check on the rails after final machining, tends to cancel. Although shafting sections have occasionally been specified to be dynamically balanced (shaft sectiorls SHOWN IN FIG. I rotated in a balancing machine to determine both static and dynamic unbalance), there are conflicting schools of thought regarding the necessity of a dynamic balance. FIRST MEASUREMENT It has been argued that the tolerances customarily im+ SECOND MEASUREMENT posed on the manufacture of shafting sections in conjunction with good shop practice precludes objectionable shafting unbalance. 4.12 Determination of Shaft Alignment. There are basically two ways that the alignment of an installed and complete shafting system can be checked. One, which is akin to the drop-and-gap method of alignment a t initial'installation, is to remove the bolts from a coupling and compare the relative position of the two o 1000 2000 3000 4000 SOW flanges with the calculated value. The second method, JACK PRESSURE, PSIG which is both easier to accom~lishand more meaninnful. , is the so-called hydraulic jack method. F i 11 baring reaction determined by hydraulic lack With the hydraulic jack method of checking shhfting alignment, a calibrated hydraulic jack is used-to determine the actual load Y U D D O bv ~ ~ a~ bearinn ~ and this actual load is then compared with the desired gad. The 1 Thrust bearing clearances. Axial clearances beload is determined by placing a hydraulic actual bearing tween the thrust ccrlla~and shoes permit a corresponding jack as close to the bearing housing as possible (bearing fore-and-aft movement of the entire shafting system. foundations are often designed with an extension to 2 Axial deffections. The propeller thrust results in a provide a jack foundation). A dial indicator is located small axial deflection of the shafting and thrust bearing. immediately above the jack so as to measure vertical 3 Temperature difierences. The shafting can be a t movement of the shaft. Where possible, the anchor a wa& temperature (70-80 1;") relative t o that of the point for the dial indicator should be independent of hull structure (about 30 F). the bearing housing. Before recording any readings, the 4 Hogging and sagging induces bending strains in shaft should be lifted at least once to ensure that the the hull wWe the shafting is not similarly strained. shaft can be lifted 20 to 30 mils without coming into This is conveniently amessed by assuming an extreme contact with the upper half of the bearing; this prefiber hull bending strem and the neutral axis of the hull liminary jacking tends to reduce hysteresis in the shaft in bending; the 8tress, and corresponding hull strain, a t and erratic readings. For short shaft spans, a dial the s h d t centerline ia then determined by interpolation. indicator should also be installed on adiacent bearings The foregoing factors would generally not reach maxi- so that any rise of the shaft in these dearings can be mum values simultaneously, but they are prudently noted. At a later time, this may help t o explain unconsidered to do so. Typical axid movements of the predicted readings. With the dial indicators and jack in place, the shaft is propeller (the point a t which movement is a maximum) d a t i v e t o the hull are 0.5 in., for tankers with very short raised and lowered in increments, noting the jack load shafts, to 2 in., for ships with long shafts. corresponding to each increment of shaft rise. These 4.1 1 Shafting Balance. Solid shafting is inher- data are plotted as shown in Fig. 11. The data points ently bdmced, but hollow shafting requires attention in will conform to two basic slopes. The slope of the liftthb regard. The bdanee of hollow shafting is accom- versus-pressure curve as the load is transferred from the plished during the machining operation by shifting lathe bearing to the jack represents the spring constant of eentem prior to the fina.1 machining cuts. The amount the bearing shell, bearing housing, and the like. Wllerl the of unbalance in a shaft can be determined by either a shaft lifts clear of the bearing, an abrupt change in static OI dynamic balancing technique. the slope of the data points occurs. The second slope After the rough machining cuts have been made, a corresponds to the bearing reaction influence number shaftipg section can be ataticafly balanced by removing for the bearing. Due to friction in the shafting and jack system, the the shaft seetion from the lathe, placing it on rails, noting the equilibrium position of the shaft section, shifting the data points when raising and lowering the shaft do not lathe centers to compensate for the unbalance, and then coincide, the result being the equivalent of a hysteresis taking additional machining cuts on the shaft section t o loop. The deflection-versus-load plot will show a lower
shaft lift at a given jack load for the increasing load curve that the bearing should be raised or lowered should be than for the decreasing load curve. Experience indi- based on the calculated influence numbers rather than cates that the true relationship between the jack load the influence numbers determined by jacking. When jacking bearings that are very close together and and shaft lift is approximately midway between the lines determined when raising and lowering as indicated by in cases where the jack must be located some distance Fig. 11. However, in cases where the increasing and from the bearing, the jack load should be multiplied by decreasing load lines are significantly different, the mean a correction factor to obtain the load a t the bearing. The correction factor is as follows: line should favor the increasing load line. With the mean line representing the true relationship between the jack load and shaft lift established, the load which would be on the jack a t zero shaft lift and with the bearing removed is determin$ by extrapolating the mean line downward to zero shaft lift. Since the jack where and bearing are close together, the load as determined I& = influence of bearing on bearing can also be considered as the load on the bearing if the l i b = influence of jack on bearing jack were removed (or the bearing load being sought). Under favorable jacking conditions (no binding of the These influence numbers are determined by including shaft in the bearing due to athwartship misalignment, both the jack and the bearing being jacked as support interference with stuffing boxes, etc.) experience shows points in the shafting system calculations. To be that the accuracy of the bearing reactions determined is theoretically accurate, this correction factor should be usually within 10 percent. However, the influence num- used for every bearing that is jacked; however, only in bers obtained by jacking may not be as accurate. When the aforementioned two caqes is it a factor of significance. the bearings being jacked are located towards the middle Table 7 contains a tabulation of the measured bearing of the shaft and span lengths are fairly equal, jack influ- reactions for the shafting system in Fig. 1 and illustrates ence numbers are generally within 30 percent of the typical jacking results. The oil in the reduction gear calculated influence numbers. For bearings located near was heated and circulated a t operating temperature; the ends of the shaft, the influence numbers obtained by therefore, the measured reactions should be correlated jacking may disagree with the calculated values by with the hot reactions. 50 percent or more. The hydraulic jack procedure can also be used to Both the load and influence number errors aye due to detect bent shafts in that the bearing reactions can be inaccuracies which are inherent in the jacking procedure; determined with the shaft rotated in 90-deg increments. e.g., the jack not being located at the bearing center, the If the bearing reaction changes significantly with shaft load center in adjacent bearings shifting as the shaft is position, a bent shaft can be suspected. This technique raised, and hysteresis in the shafting system. Conse- is very useful when analyzing a shaft that is suspected quently, when a bearing is to be realigned, the distance of being bent.
CALCULATED
MEASURED
-
379
Sectiwn 5 Bearings 5.1 Introduction. Main propulsion shafting is supported by bearings which maintain the shafting in proper alignment. These propulsion shaft bearings divide themselves naturally into two groups; those bearings inside the watertight boundary of the hull and those bearings which are outside the hull watertight boundary. The requirements imposed upon the design of main shaft bearings are extremely severe. The bearings are required to operate a t speeds ranging from 0.1 rpm, when on jacking gear, to 100 or more rpm in either direction of rotation. And, unlike some applications, the bearing loads do not vary with rpm but are essentially constant at all speeds. Reliability is heavily emphasized'in the design of bearings because there is no redundancy for bearings and a single bearing failure may incapacitate the propulsion system. I n addition to the radial bearings which support the
main shafting, there is located inside the ship a main thrust bearing which transmits the propeller thrust from the shafting to the hull structure. Figures 1 and 2 show the two typical main thrust bearing locations. Often, the main thrust bearing is designed as an integral part of the main engine and is provided by the main engine manufacturer. l'or details concerning main thrust bearings, see Chapters 9 and 20. 5.2 Line Shaft Bearings. Bearings located inside the ship's water-tight boundary are called line shaft bearings, although they are sometimes referred to as steady or spring bearings. Almost without exception, these bearings are ruggedly constructed, conservatively designed, babbitt lined, and oil lubricated. Except in special cases, the bearings are self-lubricated by rings or disks arranged in such a manner that lubrication is effected by the rotation of the shaft. Roller bearings
MARINE ENGINEERING OIL SCRAPER
7
rANTI- ROTATION DOWEL
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS r
WINGED INSPECTION COVER
HALF
\
,-OIL
LUBRICATING RING 7
COOLING COIL CONNECTION
Rg. 12 Self-aligning line shaft bearing with oil disk lubrication
have been used in the smaller shaft sizes, but the advantages of lighter weight and lower friction have in general not been sufficient to offset the higher reliability and lower maintenance costs of the babbitt-lined type. Line shaft bearing housings are made of steel castings or fabricated of steel plates welded together. Completely satisfactory bearing housings are obtained by either method, and manufacturing costs govern the construction method used. Since rigidity is of more concern than strength, low carbon steel is used as the material for bearing housings with the exception of bearings for naval combatant vessels, in which case high-impact shock requirements may necessitate the use of high-strength steel. Bearing housings are split horizontally a t the shaft centerline. The bottom half of the bearing must be very ruggedly designed since it carries the vertical shaft load and any side load that exists. The bearing housing supports a h h v y steel removable shell which is limed with babbitt. The shaft rests on the babbitted surface. The bearing shell can be made with a self-aligning feature by providing a spherical or crowned seat at the interface between the bearing shell and housing. This allows the axis of the bearing shell to align exactly with that of the shaft. Figure 12 is a section through a bearing with a self-aligning feature and Fig. 13 is a section through a bearing that is similar but without a self-aligning capability. The general construction of bearing housings and shells can be observed from Figs. 12 and 13. Except for the aftermost line shaft bearings in mer-
chant applications, it is general practice to babbitt only the bottom half of the bearings since these bearings would never be expected to be loaded in the top. However, the aftermost bearing (the one closest to the stern tube) may become loaded in the top particularly when the stern tube bearing is water-lubricated. Waterlubricated bearings are subject to a large amount of wear which can result in severe misalignment. It is considered good practice to provide the maximum practicable amount of babbitt in the top half of the aftermost line shaft bearings when water-lubricated stern tube bearings are used. With oil-lubricated stern tube bearings, the probability of the after bearing becoming loaded in the top is considerably reduced. . Babbitt that is centrifugally cast onto the bearing shell is considered preferable to that which is statically poured. The former technique dependably provides a more secure bond between the babbitt and the bearing shell. The desire for centrifugally cast babbitt has required some adjustments in bearing shell and housifig design. Babbitt can be of either the lead or tin base type. Tin-base babbitt has greater strength and is generally preferred for shaft bearings; it is specified almost exclusively for centrifugally cast bearings. Lead-base babbitt is preferred where embedding, conforming, and antifriction are primary considerations. Lead-base babbitt has a lower yield point and a slightly better fatigue resistance. Physically the load-carrying length of the bearing
Fig. 13 Nan-self-aligning line shaft bearing with oil ring lubrication
i
1
should not exceed two times the shaft diameter (LID = 2) nor should it be much less than one shaft diameter (Ll D = 1). These limitations are set to assure uniform bearing contact in the case of long bearings and to prevent excessive end leakage of the oil from impairing adequate lubrication for short bearings. The oil reserv.oir must be sized to operate during extreme roll and pitch conditions without leaking oil by the shaft or disabling the bearing lubrication system. Furthermore, the oil quantity and sump surface must be sufficient to dissipate the heat generated. Line shaft bearings are sometimes designed with cooling coils located in the sumps as shown in Fig. 13; however, experience has shown that the cooling coils are rarely, if ever, needed. Line shaft bearings may be lubricated by means of oil rings, an oil disk, or by a supply of oil under pressure (wick-lubricated bearings have fallen into disuse). Ring oil-lubricated bearings contain two or three metal rings with a diameter of 1.25 to 1.5 times that of the shaft (the ratio decreases with larger shaft diameters). The number of rings in a bearing should be selected such that no ring is required to distribute oil for an axial distance g,reater than 7 in. on either side of the ring. The rings rest on top of the shaft and dip into an oil reservoir located beneath the bearing shell. Figure 13 is an example of a ring-lubricated bearing. As the shaft turns, the rings are rotated by the frictional contact with the top of the shaft. Oil which adheres to the ring in way of the oil reservoir is then carried up to the top of the shaft
where, a part of the oil is transferred to the shaft and subsequently carried into the contact region of the bearing. Ring-lubricated bearings have proved to be capable of accommodating large angles of list and trim and have proved to be reliable in service with design bearing unit loads of 45 psi. With regard to the possible adverse effects of trim, tests have been conducted which demonstrated that ring-lubricated bearings can accommodate angles of approximately 10 deg from the horizontal with no sacrifice in performance. Reference [32] discusses the performance of oil rings based on laboratory tests and shows, among other things, the sensitivity of the quantity of oil delivered to the oil viscosity. Disk-lubricated bearings use a metal disk clamped to the shaft a t one end of the bearing shell. The disk may have a flange as illustrated by Fig. 12. As the shaft turns, thelower portion of the disk, which is immersed in , coated with oil. This oil is carried to an oil r e s e ~ o i r is the top where a metal bar scrapes the oil from the disk and guides it into passages where it is admitted to the top of the shaft and then into the contact region of the bearing. Disk-lubricated bearings have been successfully applied with design unit pressures of 75 psi. I n special cases, line shaft bearings may be lubricated by oil supplied by a pump. If the shafting system is very long, sump pumps are required to return the oil from the bearings since a gravity drain is not feasible considering the possible trim and pitch conditions of the ship. While this method of lubrication assures an adequate supply of oil a t all shaft speeds, it has the dis-
382
PROPELLERS, SHARING, AND SHAFitNG SYSCEM VIBRATION ANALYSIS
MARINE ENGINEERING
advantage of the extra pumps and complexity. Furthermore in the event of pump failure, the bearings may be damaged from the lack of oil. The load which can be supported by a babbitted journal bearing is dependent upon the method of lubrication, the bearing length to shaft diameter (LID) ratio, and of course the installation workmanship. I n the past, babbitted journal bearings were restricted to bearing pressures of 20 to 30 psi based on the projected bearing area. This limitation on allowable pressure resulted in bearings with LID ratios as large as 2. Even with high LID ratios, the shafting systems had very closely spaced bearings such that the bearing loads were very sensitive to alignment. The use of higher bearing pressures along with the use of more sophisticated techniques in positioning bearings has resulted in more reliable shafting systems by virtue of the more favorable bearing LID ratios and more flexible shafting systems. The most severe demands on the lubricating system of a line shaft bearing do not correspond to full-power, fullrpm operation, but to the condition when the shafting is rotated by the turning gear a t about 0.1 rpm for extended periods of time to facilitate uniform cooling or heating-of the main turbine rotors. If the lubrication system fails to deliver adequate oil to the journal under this condition, the oil film which separates the bearing journal from the babbitt will not be replaced as it is squeezed out; consequently, metal-to-metal contact and damage to the bearing surface may occur. Lubrication provisions have a strong influence on a bearing's ability to operate satisfactorily in the critical jacking mode of operation; and, consequently, the means of lubrication .strongly influences the extent to which line shaft bearings can be loaded. As a guide, it has been found that as little as 25 drops of oil per minute on the journal surface is adequate to sustain indefinite operation in the jacking mode a t bearing pressures of about 75 psi. With proper attention given to design details, ringlubricated bearings, disk-lubricated bearings, and pressure-lubricated bearings can carry increasingly higher unit loads in that order. Disk-lubricated bearings can carry a higher unit load than ring-lubricated bearings based on the assumption that the oil scraper functions properly. Very close controls must be maintained in the manufacture of oil scrapers because manufacturing flaws which are hardly perceptible can have a large influence on their performance. 5.3 Outboard Bearings. Outboard bearings can be further classified as stern tube or strut bearings. Figures 1 and 2 show the locations of these bearings relative to the ship arrangement. Outboard bearings can either be water lubricated or oil lubricated. Almost without exception in this country, all outboard bearings were water lubricated up till about 1960 when a transition to oil-lubricated bearings began. This transition to oil-lubricated bearings was stimulated by the unduly short service life of many of the water-lubricated bearing assemblies during that period. It is believed that the shortened life of the
water-lubricated bearings was caused by the larger ship sizes which had greater bearing loads and more contaminated water passing through the bearings (larger ships operate with less clearance between the hull and channel bottoms such that more silt, mud, and sand is drawn into the bearing surfaces). Reference [33] reports the experience of ship operators regarding stern tube bearing wear. Minimization of vibration was also influential in the promotion of oil-lubricated bearings. Particularly with larger and fuller ships, variations in the water inflow velocity to the propeller generate large variable bending forces on the shafting; many instances of pounding of the shafting in the forward stern tube bearing and the stern tube stuffing box of single-screw ships have been noted particularly when five-bladed propellers were being used. Oil-lubricated bearings which have close bearing clearances eliminate-the pounding and associated maintenance of propeller shafts and stuffing boxes. Oil-lubricated stern tube bearings also reduce the power losses in the shafting system. For a 22,000-shp ship an efficiency improvement of about 1.5 percent can be expected with oil-lubricated vice water-lubricated outboard bearings. Although oil-lubricated outboard bearings are favored by many, water-lubricated bearings remain in common use. Figure 14 illustrates a typical water-lubricated strut bearing design. A water-lubricated stern tube bearing design is similar ;xcept 'that the bearing bushing is fitted inside the stern tube rather than the strut barrel. Water-lubricated bearings basically consist of a bronze bearing bushing which retains a number of bearing contact elements that may be made of either lignum vitae wood, phenolic composition, or rubber bonded to brass backing strips. A sleeve is installed on the shaft to provide a corrosion-resistant contact surface. Careful consideration must be given to the selection of the liner material in relationship to the bearing material [34]. When brass-backed rubber strip bearings (rubber stave bearings) are used, as is common in naval practice, dovetailed slots are accurately cut in the bushing to accommodate the bearing staves. Sufficient metal is left between each slot to hold the staves securely; the space between staves also provides a cooling water flow ~ a s s d e . As indicated by Fig. 14, bearings employing lignum vitae and phenolic materials are similar to rubber stave bearings. A "V" or "U" shaped groove is cut a t the longitudinal joints of the blocks to provide lubricating and cooling water flow. Brass retaining strips are generally placed at four points around the circumference to secure the contact elements. Lignum vitae and phenolic materials absorb water and consequently tend to swell. Phenolic materials are usually installed when dry and consequently swell significantly when put into service. Swelling must be considered in the design of both lignum vitae and phenolic bearings. Lignum vitae must be kept damp a t all times as it will otherwise become dry and crack. Reference [35]
FhClLtTATE SMAFT WlTHDRAWI)L 1--
\
ATHWARTSHIP V t E I TAKEN AT 'A-A'
F~~RWATER
4
\v
ALTERNATE DESRN FOR LL6UUM VIThL OR PHLWOLtC eOMPOSITIQLIBLOCUS
383
f/
FORE a AFT VIEW TAKEN AT "B-B* RUBBER
i
PEAK BULKHEAD
SEAL SLEEVE
Fie. t 5 TweuI
aif-hbticcrled rtacn hrbe bearing
conhim details pertinent to the application of lignum vitae ta prapelIer shaft bearings. Water-lubricated bearings are designed with L / D ratios of the order of 4 for the bearing adjacent t o the prop&er and 2 for those forward of the prnpdh3r bearing. GIassifieation societies often specify these values as minimum lengths. Unit loadings of the bearings based on prcrf&ed area (shaft diameter times hearing length) are normally under 40 psi; however, great care must be taken in placing importance on the absolute value of bearing eontact pressures which are based on the projected area. Not only does the eccentricity of propeller thrust alter the loading but also the load distribution is both difficult to assess and is subject to radical .change.
Outbwrd baring materials msy w a r 0.2 to 0.5 in. M o r e being replaced. O u t b o d hearings are occasionally aligned to ZL slope corre8pnding ta the static slope of the shaft in way of the bearing in order to obtain mom uniform bearing eantmt when initially placed in service. However, this procedure has not proven entirely satisfactory from a we&down standpoint since only the starting point of the wear process is changed and there may be little influence on the ultimate wear pattern. OiMuhricated beprings, as illustrated by Fig. 15, have been used in stern tubes and bossings more so khan struts. This Is partly because of the fewer bearing problems with water-lubricated strut bearings and partly due t o the
3 84
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
MARINE ENGINEERING
),
T FILL
Fv7
CONNECTION ~
~
~
I 0 FT. A8OVE LOAD WATER LlNE LOW LEVEL ALARM
GRAVITY LlNE
THERMOMETER CONNECTION -DRAIN CONNECTION
Fig. 16 Stern tube lubricating oil diagram
difficulties in adapting the system to struts. Oil-lubricated bearings do not require a liner to be installed on the shaft since contact with seawater does not occur nor is there any significant shaft wear. Also, no bushing is inserted in the stern tube; the bearing shells, which have heavy wall thicknesses, are pressed directly into the stern tube. The L I D ratios of the heavily loaded after
stern tube bearing have ranged widely. Early designs ~ ~ ~ ~ ~ ~ had ratios of 2.5 but a trend toward a value of 1 was subsequently established. Oil-lubricated stern tube bearings are totally submerged in oil, and seals on the after and forward ends of the tube prevent the ingress of seaw3ter and the leakage of oil into the ship, respectively. iThe pressure of the oil in the stern tube is maintained above that of the ambient seawater by means of a head tank which is located about 10 f t above the full-load waterline. Ships which have large draft changes may require two head tanks; one for full-draft operation and one for ballast operation. Figure 16 illustrates a typical lube oil diagram for an oil-lubricated stern tube bearing. A small pump is usually installed M shown to force oil circulation through the stern tube. The oil flow is such that oil is circulated through both bearings. Many variations of this system have been used including the deletion of the pump; owners often specify filters, heaters, coolers, and coalescers to condition the oil as it passes through the circuit. Coolers are rarely used as the temperature leaving most stern tubes does not exceed 120 I?. Although the unit bearing pressure based on the projected area normally falls in the 70-psi range for oillubricated bearings, the actual operating pressure is probably closer to twice this value. An inspection of the bearing contact area after operation reveals that the after bearing is loaded only on the after end for a length of about one shaft diameter; shorter bearings are often advocated for this reason.
Section 6 Propellers 6.1 Introduction. Very early in the preliminary design spiral, the hull resistance and propulsion system must be established. Section 4 of Chapter 1 deals with the considerations involved in developing the main propulsion system. The propulsor, a device which converts engine torque to ship thrust, can be one of the important determinants of the type of propulsion plant employed. Because of the interfaces between the machinery, hull, and propulsor, the design of the propulsor is usually a task undertaken jointly by a naval architect and a marine engineer. The responsibility for the design of the propulsor varies from one organization to the next, but one approach is to assign the naval architect the responsibility of developing the hull lines and the propulsor hydrodynamic design; he is supported by the marine engineer who provides the proper interface between hull and machinery to assure that an optimum overall propulsion system is obtained. This is the general basis upon which the Society's two publications, Principles of Naval Architecture and Marine Engineering, have been written. Accordingly, details regarding the propulsor hydrodynamic design are not covered in this text.
Figure 4 of Chapter 1 compares the optimum efficiency values for a number of different types of propulsors. This information gives guihnce as to the relative merits of one propulsor versus another from an efficiency standpoint. However in the preliminary design stage, more specific information is required in order to make the necessary trade-off studies to support a design selection. Systematic model tests of propulsors provide the necessary information for the trade-off studies, and in many cases the final design. Reference [I] provides the information necessary to carry out in-depth propulsor studies. Section 6.2 provides a description of the mechanical aspects and performance characteristics of the various types of propulsors to aid in defining the circumstances involved with the possible alternatives. Most ship propulsors are of the propeller type; therefore, Section 6.3 provides detailed guidance regarding the preliminary selection of the characteristics of ship propellers. Propeller manufacturing tolerances, which a marine engineer is frequently called upon to control or specify, are discussed in Section 6.4.
~
,
6.2 Propulsor Types. As noted in Chapter 1, the L y type of propulsor to be used must be selected very early in the ship design process as the type of propulsor can have a strong impact on the design of the ship itself. The vast majority of ship propulsors are of the solid fixed-pitch propeller type. Nevertheless, there are a number of other types of propellers which may be more suitable in particular instances. A brief description of the mechanical aspects of the various types of propellers is as follows: Canventional. Fixed-Pitch Propellers. Most propellers are of the 'Conventional fixed-pitch type and are made from a single casting. Conventional fixed-pitch propellers usually have an efficiency, cost, and simplicity advantage over other types of propellers. Detachable-Blade Propellers. Detachable-blade (or built-up) propellers consist of a separately cast hub and blades. The blades are bolted to the hub to form the composite propeller. When operating conditions are such that there is a great probability of propeller blade damage, detachable-blade propellers offer the advantage that individual blades can be replaced. Also some blade attachment designs have elongated bolt holes which offer the advantage that small modifications in pitch can be made, which permits adjustment in the operating rpm. The disadvantages associated with detachable-blade propellers, as compared with propellers made from a single casting, are the greater first cost, greater complexity, and inherently lower propeller efficiency (resulting from the larger hub). Controllable- and Reversz3le-Pitch Propellers. The blade angle (propeller pitch) on a controllable- and reversiblepitch (CRP) propeller can be controlled remotely. This type of propeller is advantageous in any of the followinp: situations:
1 Where the operating conditions vary widely (such tug and trawler applications). 2 Where diesel engines or gas turbines are used for propulsion (the CRP propeller permits adjustment of the engine rpm-power relationship to provide gr~ateroperating flexibility). 3 Where reversing capabilities are not readily obtainable from the main engine (e.g., gas turbines). 4 Where rapid or frequent changes in the direction of thrust is a desirable capability. M
At the propeller design point, the efficiency of CRP propellers approaches the efficiency of fixed-pitch propellers. The larger hub of a CRP propeller prevents its efficiency from exceeding that of a fixed-pitch propeller. Off the design point, the CRF propeller efficiency is less than that of a fixed-pitch propeller designed for that operating condition. This is because all sections of a CRP propeller blade are rotated through the same angle as the pitch is changed; thus, the angles of attack of the various blade sections along the propeller radii are optimum only a t the design point. Pitch changes are controlled remotely. The torque required to turn the blades and hold them in position is
385
obtained by controlling the oil pressure on a piston which is mechanically linked to the propeller blade throws. The hydraulic pistons are located in the propeller hub or in a section of shafting located inside the ship. For a description of a typical CRP installation, see reference [36]. Propellers in Nozzles. There are two types of arrangements which fall into this category; namely, the pump jet and the Kort nozzle. In the pump jet arrangement the propeller is placed in a rather long nozzle with guide vanes either forward, aft, or both places relative to the propeller. The pump jet is normally considered where propeller noise is important. Due to the resistance of the nozzle and guide vanes, the overall efficiency of the pump jet arrangement is strongly dependent on particular circumstances. Kort nozzle propeller arrangements show efficiency advantages in applications where the thrust loading is high; examples of such applications are tugs, trawlers, and large slow-speed ships (see Fig. 4 of Chapter 1). The Kort nozzle arrangement consists of a propeller located in a nozzle of relatively short length (the length/ diameter ratio of the nozzle is in the range of 0.5 to 0.8). Kart nozzles are extensively used in connection with tugboats because the bollard pull and towing pull can be increased 30 to 40 percent as compared with a propeller operating alone without a nozzle. Tandem Propellers. As the horsepower requirements for a ship increase, a single propeller can become inadequate due to restrictions on the propeller diameter, draft limitations, or excessive thrust loading. When this occurs, an increase in the number of propellers is required. Since a single shaft is desirable from an economic viewpoint, there is justification for considering two propellers in tandem on the same shaft. Only small losses in propulsive efficiency (2.2 percent) were reported from model tests in reference [37] for the tandem arrangement over a twin-screw arrangement for a large tanker. The economy of a single propulsion plant, as opposed to two propulsion plants, in addition to the sing%-screw simplicity of the shafting arrangement are the advantages offered by a tandem propeller arrangement. Contrarotating Propellers. Contrarotating propeller arrangements consist of two propellers positioned in tandem on coaxial shafts which rotate in opposite directions. Higher efficiencies can be achieved with this propeller awangement because no rotational energy need be left in the propeller wake. Reference [37] reported a propulsion efficiency improvement of 6.7 percent for a 136,000-ton-displacement tanker with contrarotating propellers as compared with a conventional single-screw arrangement; similar tests for an 18,170-ton-displacement dry cargo ship indicated a 12-percent improvement [38]. Contrarotating propeller arrangements have not been used in connection with commercial ships due to the mechanical complications involved with the coaxial propulsion system arrangement (see Chapter 9). Should the advantages of a contrarotating system become enhanced because of higher horsepower, higher fuel costs,
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
MARINE €NGINEERING
h
or lower first costs of the drive system, contrarotating systems may appear ip merchant ship applications. Some naval installations have been made, but their performance has not been made public. FdEy Cwitating Propellers. The primary chjection to propeller cavitation is the deleterious effect that it has on the propeller blade surfaces. Once the propeller loading conditions become such that cavitation can no longer be avoided, as may be the case with very fast ships, then rather than accept a limited amount of cavitation a more satisfactory choice is to design the propeller such that it cavitates fully. In this event, the cloud of vapor which forms on the suction side of the blades does not collapse until it is clear of the propeller blade, thus having no deleterious effect on the propeller blades. Operation at offdesign conditions may result in severe propeller cavitation erosion; and such operation (accelerating, decelerating, etc.) cannot be entirely avoided in service. For this reason and to withstand the high stresses resulting from the large thrust load, fully mvitating propellers are frequently made of exotic materials. Reference [ I ] gives the expected performance of a theoretical series of 3-bladed supercavitating propellers which can be used for estimating performance. In order to achieve fully cavitrvting performance in a speed range too low for the usual fully cavitating propeller design, but still in the range where conventional .propellers would cavitate excessively, ventilation may be conmdered. Ventilation is the term used to describe the introduction of air into the cavitation areas to produce a fully developed cavity. Experience with ventilated propellers is very limited, but some model testing has been carried out; reference 1397 is a report of one such test. 6.3 Propeller Characteristics. An underatanding of the considerations and trade-offs involved in selecting the design characteristics of a propeller is required when developing a shafting arrangement. Assuming that the ship's power and speed requirements have heen established preliminarily as outlined in Chapter 1, the following propeller characteristics must be settled: Propeller Diameter. In general, higher propeller e%iciencies are associated with larger propeller diameters and lower shaft revolutions. Therefore, it is u m d y desirable to install the largest propeller diameter that can be accommodated by the hull structure. The propeller diameter is limited by the lines d the ship as discussed in Section 2.2. PropeUer rpm. The choice of the propeller rpm involves establishing a balance between propeller efficiency and the weight, cost, and space requirements of the main machinery. This is accompliihed by using standard propeller series data (such as reported in reference [l])to compute a series of curva of propulsive efficiency versus revolutions for various propeller diameters. The envelope of these curves indicates the best efficiency and optimum propeller diameter that can be obtained at each propeller rpm. The point of maximum efficiency on this curve for a propeller diameter is termed the
optimum propeller rpm. This envelope curve is used to assess the sacrifice in &ciency which must, be awepeed for any increase of revolutions. Data fmm this curve combined with the effect of the revolutions on the weight, cost, and space requirements of the main p p d l i n g machinery permit the final selection to be made. It will be noted that at revolutions slightly higher than the optimum propeller rprn for a given propellerdiameter, the propeller efficiency d e e m only slightly. But on the other hand, the effect of relatively small incraws of propeller rpm (with the power remaining the same) on the weight, cost, and spme requirements of the main machinery can be significant. I n the case of higherpowered vessels, it is usual to select a propeller rpm higher than optimum and to w p t some sacrifice of propeller dciency in order to reduce the &e of the propelling machinery. Propeller revolutions higher than the optimum are dm accepted for mamns related to the type of maehery. Direct-drive diesel engines and electric-drive machinery usually operate a t speeds higher than the optimum to permit the use of a smaller engine or motor; %veij&t,cost, and space requirements are factors of major importance with these types of machinery. N a d w of Blsdes. Propellers may have t h e , fod, five, six, seven, or more blades. Over the years, the trend has been to use a larger number of blades; three blades fell into complete diswe for large ships during the 1940's. During the I W s , six- and seven-bladed propellera came into use. The major factor in the =leetion of the number of propeller blades is vibration considerations. Both the hull hydrodynamic pressure forms and the forces transmitted through the shafting system bearings are strongly influenced by the selection of the number of propeller blades. I n general, the,propeller exciting forces decrease rapidly with larger numbers of blades; however, there are exceptions. For more detrtib concerning the relationship between We number of propeller blades and the vibratory f o m generated, see Section 3.2. Prudent selection of the number ofpropeller blades iq an important variable which can be used to avoid the excitation of natural frequencies in the propulsion system. PropeEkr Pit&. The selection of propeller pitch can be made when the power, speed, revolutions, and general hull characteristics have been settled. The pitch ratio may be selected on the basis of standard propeller m d e l series data. However, when a propeller is highly loaded or operates in a non-uniform wake field, it may be desirable to design a propeller with a pitch ratio and pitch distribution tailored to suit the particular operating conditions. This can be accomplished by desiping a propeller in accordance with the circulation theory (see reference [I] for a description of circulation theorgr methods). BE& Skew. A propeller blade is termed skewed when its outline is asymmetrical with respect to a s t d g h t radial reference lime inthe plane of the propeller. Skew is usually introduced by successively &placing the
4
1
,
blade sections awav from the direction of rotation. Propeller blades with skew tend to enter and leave the regions of high wake more gradually, resulting in a reduction of the alternating propeller loading due to wake irregularities. The results of model tests have shown that blade skew is an effective technique for reducing the fluctuating forces and moments acting on a propeller. It is normal practice to skew propeller blades a moderate amount based on past experience, without specific knowledge regarding the benefits achieved. Developed Area. With heavily loaded propellers, which is usual with most modern ships, the developed area must be established with care. Considerations in the selection of the propeller developed area are the penalty in efficiency associated with an excessive develo~edarea and the effects of cavitation due to an inadequate developed area. Effects due to inadequate area can be of greater consequence than those due to an excessive area; therefore, prudent practice dictates that a developed area be provided which is sufficiently large to entail a minimal cavitation hazard. kcor a more detailed discussion of propeller cavitation (and consequently developed area) see reference [I]. Propeller Blade Thickness. Requirements concerning the minimum allowable blade thickness are given in classification society rules such as i~ference[18]. A thorough discussion of the development of the classific* tion society rules is given in reference [40], which in addition provides the basis for making an in-depth analysis of the propeller blade stress. Prweller Hub. The controllintz dimensions for the propeiler hub outside diameter an; length are the stern frame (or strut barrel) and the propeller blade fore-andaft length at the interface with the hub. These parameters onlv establish the lower limit. and thicker hubs may be rkquired to provide adequate strength. Excessively large propeller hubs are disadvantageous in that they increase the expense of the propeller and propeller weight (and consequently propeller shaft stress). Propeller Weight. An estimated propeller weight can be obtained in several ways. The most accurate is to conduct a calculatibn based on detailed drawines. Unfortunately, however, the need for the weight has usually passed by the time detailed drawings are available. There are a number of approaches which may be used to approximake propeller weights; one approach is given in reference [40]. One of the less accurate methods, but one requiring the least information, is given in [41J; this method entails the use of wrves which relate the weight of a propeller to its torque rating. There are other methods such as
I
where
a
W
= propeller weight (manganese bronze),
lb (including hub) K = conhnt, approximately equal to 0.26 D = propeller diameter I
MWR
BTF
= mean width ratio
-
developed area per blade D (blade radius - hub radius)
=
blade thickness fraction
maximum blade thickness extrapolated to shaft axis D Care must be exercised in the use of approximate methods because of considerations such as unusual hub dimensions and allowances for ice strengthening. 6.4 Manufacturing Tolerances. As indicated in Table 8, there is an array of tolerances which control the dimensional accuracy of propellers. Table 8 gives a range of propeller tolerances which have been used and also shows recommended tolerances which are expected to produce satisfactory results. Unless care is exercised in establishing manufacturing tolerances, the tolerances can easily be over-specified; that is, the tolerances may be so tight that increased manufacturing costs may be incurred with no corresponding increase in the value of the finished propeller. Various approaches have been used to check the dimensional accuracy of propellers. The usual technique used with merchant propellers is to measure the blade thickness and pitch at discrete points and to check the remaining blade surfaces for fairness. A flexible steel straight edge is held against the blade contours as a means of detecting irregularities of the propeller surface. The technique used with naval propellers is considerably more rigorous; Cylindrical, edge, and fillet gages (which are sheet metal templates machined to the desired blade contour) are prepared which make it possible to comprehensively check the conformance of the propeller blades, at a number of radii, to the design dimensions [42]. In addition to the tolerances governing the propeller physical dimensions, balance tolerances are also specified. Ship's specifications usually require that propellers be balanced (with static or dynamic equipment) such that the static unbalanced force at rated rpm is no greater than one percent of the propeller weight. The following expression may be used to determine the static unbalance corresponding to an unbalanced force equal to one percent of the propeller weight:
where
U
static unbalance which will generate an alternating force equal to one percent of the propeller weight, in.-lb W = propeller weight, lb N = maximum rated propeller rpm =
Limits are not generally placed on dynamic unbalance because of the large diameter-length ratio of propellers, but good practice dictates that corrections made for static unbalance be accomplished so as to improve the
I i
388
MARINE ENGINEERING
389
PROPELLERS, SHARING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
I
Table 8 Dimension Blade Thicktw.88 Maximum Minimum
Blade Width Maximum Minimum
Propeller Manufacturing Tolerances
-
Naval
Rmge of Experience Merchant -0.00, +% in. or 2% of deaign th~ekness,whichever is greater -0.00, +% in.
Recommended Practice
+%.$a in. or + l % of design -0.00 +>$ in. or 2% of thickness, whichever is greater deaign thickness, whichever is greater - .02 in. % % of thickness) d 4 in. +I % of thickness)
+
+
+g2in. or 1% of width 4
in. +0.25% of width) in. +0.5% width)
+>b in.
iRd2
Bhde Position in Transverse Phne +15 min Maximum +30 min +15 min &15 min Minimum Blade T r d at Tip +1.0 in. Maxlrnum &%6 in. & x 6 in. Minimum +% in. Pro e* Diameter -1.25 in., +0.00 daxlmum &% in. Minimum +% in. -.20 in., +O.OO Deviation of Pitch cf.tany Radius from Designed value Maxlmum +2% +1.5% Minimum % +l.O% Geatest Variation i Any Avetage Blade Pitch to Average Propeller Pitch .Gum +I% Mi imum &0.75% z?5%
dynamic unbalance. Dynamic unbalance is generally not found to be a problem; nevertheless, the dynamic unbalance should be limited such that the alternating force generated at the aftermost bearing is no greater than an alternating force a t the aftermost bearing corresponding to a static unbalance equal to one percent of the propeller weight at the propeller center of gravity. A useful expression for the maximum allowable dynamic unbalance under these conditions is as follows:
+15 min
&% in. &% in.
0
where
D
= dynamic unbalance which will produce same
force at aftermost bearing as a static' unbalance equal to one percent of propeller weight applied a t propeller center of gravity, in.-lb-in. W = propeller weight, lb L1 = distance from propeller center of gravity to aftermost bearing reaction, in. Lz = distance from aftermost bearing reaction to reaction of next bearing forward, in. N = maximum rated propeller rpm E = shaft modulus of elasticity, psi I = shaft rectangular moment of inertia, in.4
Section 7 Torsional Vibration 7.1 General. Severe torsional vibration difficulties experienced with the early reciprocating engine drives and particularly diesel engines moved the importance of torsional vibration as a design consideration to the forefront. Subsequently, torsional vibratian became established as a factor which had to be carefully considered in the design of all types of main propulsion shafting systems, and the design methodology required to conduct a reasonably accurate torsional vibration analysis has been formulated [41,4349]. A summary of the analytical methods used to evaluate the torsional vibration characteristics of geared turbine-
driven main propulsion shafting systems is included in the following pages. The fundamental theory of torsional vibration is well documented in the existing literature; therefore, it will not be reproduced here. A discussion of details peculiar t o torsional vibration of diesel-driven main propulsion shafting systems is similarly not included. 7.2 Modes of Torsional Vibration. The design of most large turbine-driven ships is such that one or more resonant modes of tonional vibration will occur within the operating range. The first mode of torsional vibration is the one in which the node is immediately abaft
the slow-speed gear and the antinode is a t the propeller; in this mode, the inertia of the propeller vibrates against the inertia of the turbines and gears. For vessels with fairly long runs of shafting, the first-mode frequency is excited by blade rate excitation at a very low propeller rpm (at about % of maximum rated rpm). For this reason it is seldom objectionable as the alternating torques developed are not of sufficient magnitude to be deleterious. On the other hand, in the case of ships which have short runs of shafting, the first mode may occur above 50 percent of the maximum propeller speed and warrant a comprehensive analysis. Second and third torsional vibration modes are determined primarily by the characteristics of the prime mover. With geared turbine drives, the turbine-gear system generally cannot be designed such that the second mode of torsional vibration (the one in which the two turbine branches vibrate against each other) is out of the operating range. This being the case, a so-called "nodal drive" is frequently provided [43]. In a nodal drive, the turbine branches are designed to have equal frequencies; this forces the slow-speed gear to be a nodal point. The second mode of torsional vibration will then consist of a motion in which the two turbines vibrate so that their vibratory moments oppose each other with a nodal point a t the gear. This being the case, the turbine branches cannot be excited by the propeller. In the third mode of torsional vibratiqn, the vibratory torques of the propeller and turbines oppose that of the slow-speed gear. The third mode usually occurs considerably above the operating range; consequently, it is rarely of concern. However, very high rpms or a large number of propeller blades may bring it within the operating range. The third mode is difficult to excite because the antinode occurs a t the slow-meed gear, which is not a source of excitation with modern gears, and the node occurs near the propeller, which is a source of excitation but has a very small vibratory amplitude. The mode shapes of the first three modes of torsional vibration are shown in Fig. 17. 7.3 Models for Torsional Vibration Analyses. A typical steam turbine propulsion system is schematically illustrated in Fig. 18(a). From an .inspection of2Fig. 18(a), it is apparent that a comprehensive torsional vibration analysis of such a system would be prohibitively complex if the classical approach were used; cobequently, simplifications must be made to facilitate a practical analysis. The system can be reduced to an equivalent svstem (model) in which all elements are referred to the sime rdtationil speed, thereby greatly simplifying the analysis. Such a procedure may be used to obtain the model shown in Fig. 18(b). Figure 18(b) can be used to evaluate all modes of torsional vibration which would be expected to be of interest in practice. However, if only the first three modes of vibration are of interest, which would generally be the case, the model shown in Fig 18(b) can be further simplified to that shown in Fig. 18(c) without a serious loss of accuracy due to the fact that the equivalent inertia of the turbines and stiff-
( a ) First Mode
( b ) Second Mode
Legend I
2 3 4 5 6
L P Turbine L P High-Speed Reduction HP Turbine H P High-Speed Reduction Slow-Speed Retjuction
Propeller
5
(c) T h i r d Mode fig. 17 Mode shapes of first three modes of tonional vibration of a turbinedriven propulsion system of nodal-drive type
ness of the turbine shafts are very high compared with those of the first reduction gear elements. If only the first mode of torsional vibration is of interest, then it can be approximated in either of two ways. One way would be to directly add the equivalent inertia of the turbine branches [JTL and JTH in Fig. 18(c)] to the slow-speed gear inertia (Ja) and make an analysis based on a two-mass system. A more practicable approach, since the inertias of the turbines and gears are . frequently unavailable, is to assume that the nodal point in the first mode of vibration is four percent of the distance from the slow-speed gear to the propeller aft of the slow-speed gear. With such an assumption, the first-mode natural frequency can be simply determined by considering the system to be a one-degree-of-freedom model as shown in Fig. 18(d). All of the system parameters which are needed in order to evaluate the torsional natural frequencies can be directly determined from the physical properties of the system except for the propeller entrained water. Assessment of the propeller entrained water can be made from the work of Burrill and Robson [49]. I n order to avoid the tedious labor associated with calculating the moment of inertia of the propeller in air, an approximate method such as that described in reference [41] may be used or the propeller radius of gyration may simply be estimated to be between 0.40 and 0.44 of the propeller radius (lower values correspond to larger propeller hubs and smaller numbers of blades). 7.4 Determination of Natural Frequencies. The Holzer method of computing the natural frequencies of lumped spring-mass systems is a convenient procedure for determining the torsional natural frequencies of
MARINE ENGINEERING :ond reduction
PROPELLERS, ,SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS Table 9
rn
LP f i r s t HP f i r s t
Determination of First Natural Mode of Torsional Vibration for a Turbine Driven Propulsion System Modeled as Shown by Fig. 18(b) Number of pro eller blades = 6 Assumed propefler rpm = 22.38 J = in.-lb-set' + 1W k = in.-lb/rad + 10"
reduction r o t i o = n 2 reduction r a t i o = n,
391
o =
oa =
6(22.38) (2a)/BO rad/sec 197.7 rad'/sec"
( a ) S c h e m a t i c i l l u s t r o t i o n of a geored t u r b i n e d r i v e n propulsion system
B L = -0.06477 Bp 011 = -0.06417 Bp 2Q = 1213 B L 321 .O OH 98.8 BP 2Q = -78.57 Bp 20.60 BP 98:s BP
+-
++
2Q = 0; therefore a resonant condltlon
( b ) Equivolent 6 moss system w i t h a l l bronches r e f e r r e d to t h e propeller r p m
(c) Four mass system w i t h o l l bronches r e f e r r e d t o t h e p r o p e l l e r r p m f o r opproximoting t h e f i r s t t h r e e n o t u r o l frequencies o f torsionol vibrotion
Fig. 1 8
( d l Single degree o f freedom system f o r opproximating t h e f i r s t noturol frequency of torsionol vibrotion
details concerning torque excitation, see references [3, 50-521. 7.6 Damping. There are several sources of damping which tend to reduce the maximum attainable amplitude of torsional vibration; one of the most important is the propeller, particularly in the first mode due to its being a t the antinode. Propeller damping can be determined in several different ways [44-46, 531. I n general, the propeller damping coefficient, b, can be expressed as
Equivalent systems for determining natural torsional f requencies of geared turbine-driven propulsionsystems
b = KQ/Q
in.-lb-sec rad
where turbine-driven propulsion systems. I n order to review the computational procedure for a practical example, consider a turbine-driven vessel that is modeled as shown by Fig. 18(b). Typical values of system inertia and spring constants and calculations for the first torsional natural frequency are given in Table 9. The calculation is accomplished by first assuming the resonant frequency of the system, relating this frequency to the propeller rpm, and then calculating the corresponding vibratory torque and torsional amplitude a t the slow-speed gear (inertia JG) in terms of the amplitude a t the terminal end of each branch. For convenience the amplitudes a t the terminal ends of the three branches are initially assumed to be one radian. Since the three branches (propeller, LP turbine, and HP turbine) must have the same amplitude a t the slow-speed gear mass, the amplitudes of the three branches can be expressed as a function of the same unknown amplitudefor instance, the propelle~thereby obtaining the mode shape. The torques imposed on the slow-speed gear are then summed; if the sum is zero, a resonant condition is established. If the sum is not zero, the process is iterated until the sum is zero by assuming a different resonant frequency. The same procedure can be repeated to determine the remaining four torsional natural frequencies but, as
previously mentioned, only the first three modes would generally be of interest. The mode shapes of the first three natural torsional frequencies are shown in Fig. 17. The node in the first mode is seen to be immediately abaft the slow-speed gear. The two turbine branches are tuned in the second mode such that the slow-speed gear is a nodal point. The third mode is the one in which the slow-speed gear is the antinode with the terminal end of the three branches being near nodal points. 7.5 Excitation Factors. There are several possible sources of torsional vibration excitation with turbinedriven ships, but the propeller is the only one of consequence. With the accurate cutting of modern gears, gear-excited criticals are either wholly absent or are of negligible amplitude. Propeller excitation of a frequency higher than blade rate exists but it is normally negligible in magnitude. A number of factors, such as the propeller loading, propeller aperture clearances, number of propeller blades, hull lines, and hull draft, influence the magnitude of the vibratory torque; consequently, generalizations in this area must be used with care. Nevertheless, typical ranges of torque excitation, expressed as a percentage of the mean torque, are presented in Table 2; for additional
K = a constant Q = mean propeller torque, in.-lb Q = rotative speed of propeller, rad/sec
If propeller model test data are available, it may be shown that
dependent and independent variables, moving along a constant pitch line. The energy loss via the propeller per cycle of torsional vibration can be written as
E,
= ~ b w 8 , ~ in.-lb
(201
where w = circular frequency of vibration, rad/sec 8, = amplitude of propeller vibration, radians
Energy is also dissipated as a result of elastic hysteresis in the shafting, sliding fits, etc. Although such internal damping losses in a shafting system would be expected to be small when considered individually, they are frequently estimated to dissipate about 5 percent of the vibratory energy per cycle when totaled. The energy dissipated due to internal damping crtn be expressed as [46]: I
where where s is the propeller slip. If the propeller data are given in the form of J, KQ curves
If given in the form of a Troost diagram (B,, 6) as in reference [I] :
As an approximation for many propellers, K = 3.7 to 4, which may be used in the absence of other data; a value of 4 corresponds to a damping constant which is double the slope of the torque-speed curve. I n all cases the derivatives are computed a t the operating point of the propeller by taking the ratio of small differences in
ar = fraction of energy dissipated; the value of this
quantity is approximately 0.05
J,
= moment of inertia of mass n, lb-in.-sec2
8, = amplitude of vibration of mass n, radians
9
The damping action of the turbines would general be expected to be of secondary importance especial1 in modes where the turbines have small relative amplitudes; however, it may warrant assessment under some circumstances. The energy dissipated due to turbine damping can be expressed as
ET =
z ~ c w 8 , ~ - in.-lb
(22)
where c = turbine damping constant which can be approximated as the ratio of turbine torque to turbine rpm a t the speed corresponding to the point under study, in.-lb-sec/rad
392
MARINE ENGINEERING
PROPELLERS, SHARING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
w = frequency of vibration, rad/sec 8, = amplitude of vibration of turbine rotor, radians
If damping is introduced into the vibration calcolrlr tions, the computational procedure is modified considerably. An external source of damping, such as that at the propeller or turbines, introduces an external moment of -jbw8 on the respective mass concentration; and internal damping, such as shafting hysteresis, between two masses is equivalent to changing a spring constant k to a complex spring constant
evaluation. These attempts have not proven successful due to many variables which must be considered. For example, the propeller design (pitch distribution, skew, number of blades, etc.), propeller aperture clearances, hull lines in the vicinity of the propeller, harmonic content of the wake in way of the propeller in conjunction with the type of reduction gear, system damping, ship operating practices (shallow water, partial propeller emersion, operating point) and similar considerations vary so much from one ship to the next that there are numerous exceptions which can be taken to all, except the most trite, generalizations.
and by substituting the expression 4Q/Q, equation (16), for the propeller damping coefficient and letting w = ZQ, equation (25) can be simplified to
where Z is the number of propeller blades. The maximum amplitude of propeller vibration can be determined from the foregoing expression. In conjunction with the normalized mode shape determined from the system natural frequency calculation, Table 9, where ar is the fraction of the elastic energy absorbed by the propeller amplitude is used to assess the vibratory torque at resonance at any element of the system. The the damper [45]. Calculations which incorporate damping as just alternating torques in the quill shafts between the highindi'cated are somewhat tedious, particularly if the cal- speed gears and low-speed pinions are usually the largest culations must be done by hand; an easier procedure is to from a relative viewpoint; consequently, it is customary compute the effect of damping at resonance only by to analyze these elements when investigating the possibility of torque reversals. equating input energy to dampening energy. As an example, referring to the calculation in Table 9, 7.7 Vibratory Torque Calculations. In many cases, with a propeller excitation equal to 3 percent of the mean the torsional vibration characteristics of a shafting propeller torque, the alternating torque, q, in the lowsystem can be shown to be satisfactory in the design stage with only a computation of the system natural pressure quill shaft in the first mode of torsional vibration frequencies and without predicting vibratory torques would be : q = 1213 8~ X lo6 and amplitudes. Normally this is possible when a q = 1213 (0.06477 8,) X lo6 compar&on is made with a similar system that has q = 78.6 (r/4Z) X lo6 proven satisfactory in service. For designs where the q = 98,300 in.-lb system natural frequencies, vibratory excitation, or anticipated system operation may cause concern, investi- This is the torque in the low-pressure quill shaft referred gation of the magnitude of the vibratory torques and to line shaft speed. With a second reduction ratio of stresses is necessary. 7.5, the actual vibratory torque in the quill shaft will be In order to illustrate the procedures used to assess 98,300/7.5 or 13,100in.-lb. In this particular case at the vibratory torque amplitudes, again consider the calcula- resonant frequency, the low-pressure turbine develops tions shown in Table 9. In the absence of speed-power 55 percent of the total power delivered to the propeller. curves for the ship, the mean operatingtorque correspond- The mean torque in the low-pressure branch a t resonance ing to the resonant frequency can be approximated by is consequently 33,500 in.-lb whereas the alternating determining the rated propeller torque (the torque torque is estimated to be 13,100 in.-lb; therefore, torque corresponding to 22,000 shp at 115 rpm) and assuming reversals in the low-pressure train at the first resonant that the propeller torque varies as the square of the mode of torsional vibration are not expected. The propeller rpm; therefore, the mean operating torque at vibratory stress in the quill shaft, kz, can be calculated the resonant frequency is estimated to be using the alternating torque across this shaft of 13,100 in.-lb. In the more general case, where system damping in addition to that associated with the propeller is of importance, the maximum amplitude of propeller vibraThe exciting torque can be expressed as rQp where tion at resonance can be found by solving the following taken from Table 2, is the alternating torque expressed equation for 8p: as a fraction of the mean torque. With a maximum , vibrating energy propeller exciting torque of r ~ the EE = E, El ET (27) input per cycle of vibration is: 1 nrQ8, = ~ b ~ 8 ; 5 T C U B ~ ~ (28) Y EE = nrQ8, (24) All of the terms, except 8,) are either known or can be In the first mode of vibration, the propeller would be expressed as a function of 8,. Once 8, is established, the expected to be the only significant source of excitation calculation can be continued as shown. and damping; therefore, by equating the expressions for 7.8 Acceptable Limits for Torsional Vibration. the propeller excitation and damping energy, the maxi- Many attempts have been made to standardize the mum propeller amplitude can be determined as follows: procedure to be used for torsional vibration analysis and
+ + +C
+C
393
As a broad rule it may be stated that untuned torsional vibration resonant frequencies should not occur in the range of 60 to 115 percent of rated rpm; however, this rule does not in fact ensure satisfactory torsional vibrs tion characteristics; furthermore, there may be aatisfactory systems in service which it would exclude. Generalizations can serve as a broad guide, but the hazard associated with generaliiations is that their limitations may not be appreciated. There appears to be no satisfactory alternative to conducting an analysis of each particular system and studying each factor individually.
Section 8
I
8.1 Introduction. Severely objectionable longitudinal vibrations were not encountered until the advent of several classes of large naval vessels in early 1941. Reference [54] contains a description of the difficulties experienced with these ships and also presents the most thorough treatment of longitudinal vibration that has been prepared. The works of Panagopulos [47], Rigby [55], Couchman [56],and others have added to the knowledge of the subject; yet the fundamental problem areas encountered today are the same as those identified by Kane and McGoldrick [54]. The low level of shaft axial stress associated with even the most violent instance of longitudinal vibration is not sufficiently large to induce failures in the shafting itself; nevertheless, longitudinal vibration can produce effects which are destructive to engine room equipment. Shafting systems which have longitudinal vibration characteristics that are resonant with propeller blade rate f r e quency forces experience a significant magnification of the exciting forces. Such a force magnification can result in such deleterious effects as:
1 Accelerated wear of gears, flexible couplings, thrust bearings, etc., and destruction of turbine clearances due to the increased relative axial movements. 2 Large vibration amplitudes and stresses in steam piping, condensers, and main and auxiliary machinery which ultimately result in fatigue failure. 3 Cracks in foundation and hull structures. 8.2
Determination of Natural Frequencies. There
are basically three approaches which may be taken in determining the natural frequencies of longitudinal vibration. The first approach would be to use a simplified method for the purpose of quickly assessing a situation. Approximate methods suitable for investigating the firstmode of vibration are given in references [47,,541. A second approach would be to model the shafting system as discrete masses and springs and use the Holzer method to determine the system natural frequencies.
The accuracy obtained with a discrete spring-mass model will depend upon the masses used to represent the shafting system. A third approach is the mechanical impedance method proposed by Kane and McGoldrick [MI. This method is inherently more accurate than the Holzer method since the weight of the shafting is considered to be distributed; however, the impedance method has the slight disadvantage of being somewhat more complex and diicult to grasp. For illustrative purposes, a calculation of the natural modes of longitudinal vibration of the shafting arrangement shown in Fig. 1 will be made using the mechanical impedance method. Figure 19 is a model of the shafting arrangement which is suitable for analysis by the mechanical impedance method. I t may be noted that the difference in the diameter of the inboard and outboard shafting is taken into account; in general,
M p ,= mass o f p r o p e l l e r , 227 lb-sec2/in M g = mgss c f gears, 147 lb-sec2/in M c = mass o f machinery, 7 6 7 lb-sec2/in m l = mass o f larger (outboard) shafting, 301 lb-sec2/in m 2 = mass o f s m a l l e r (inboard) s h a f t i n g , 3 6 6 lb-sec2/in
kt
= spring constant of l a r g e r (outboard) shafting,
k2
2 0 . 3 x 10' Ib/in = spring constont o f smal ler (inboord) shafting, , 5.9 x 10' lb/in
= t h r u s t bearing spring constant, 7 x 10' w i n
kt kf
= t h r u s t bearing foundation spring constant, Ib/in
2
' number of
p r o p @ l l e r blades, 6
Fig. 19 Representation of a geared tutbina propulnion hafting system ( ~ g I. ) for a longitudinalvibration analysis
394
PROPELLERS, SHAFTING, AND SHAFTfNG SYSTEM VIBRATION ANALYSIS
MARINE ENGINEERING
especially with short spans of shafting, this additional degree of sophistication is not yarranted. The majority of the system parameters may be directly calculated from the system scantlings, and therefore no difficulty is experienced in obtaining their value. However, assessment of several of the system parameters can be nebulous. For example, determination of the water entrained with the propeller does not lend itself to an accurate calculation; as a first approximation the entrained water weight may be assumed equal to 60 percent of the propeller weight. The results obtained from the experimental work of Burrill and Robson [49] are widely used in estimating propeller entrained water, and reference [54] suggests other approaches. The behavior of flexible couplings in connection with vibratory movements similarly cannot be stated with certainty. The impact of the behavior of flexible couplings, aside from the effect on the couplings themselves, is not great when the thrust bearing is located well forward. But when the thrust bearing is located aft and there is an appreciable vibratory amplitude a t the slow-speed gear, the impact of flexible coupling behavior can be significant. For a detailed discussion of the behavior of flexible couplings and the complications involved, see [54 and 561. Some machinery liquid and foundation weight participates with the shafting system when vibrating longitudinally as a consequence of being near the main thrust bearing; but assigning a magnitude to these quantities entails numerous uncertainties. Assessment of the "machinery mass," M,, to be included in the mathematics requires judgment which must be based on the specifics of each system. In instances w-here only the first mode is of importance, the machinery mass has a small participation and consequently an accurate assessment of its magnitude is not critical. On the other hand, the machinery mass is expected to have a significant participation in the second mode, in which case care must be taken in its determination. In general, the first reduction gear rotating elements, gear casing, turbines, condenser, foundation structure, or portions thereof may be included as machinery mass. Reference [54] gives some guidance in the assessment of the machinery mass. Reference [57] contains an interesting approach on the treatment of machinery masses in that a portion of the machinery mass is given a leverage ratio relative to the centerline of the shaft. The spring constant of the thrust bearing, k t b , may be considered to consist of three constituents: the spring constants of the thrust bearing housing, the thrust collar, and the thrust elements (or shoes). Aside from the tedious calculations, no diiculty is experienced in calculating the spring constants of the thrust bearing housing and collar inasmuch as the majority of the ' deflections are due to shear and can be estimated on the basis of well-established techniques. But the spring constant of the thrust elements can be difficult to evaluate. In the absence of more specific data, the data
given in [54] may be used for guidance in establishing the stiffness of thrust elements. Determination of the thrust bearing foundation spring corlstant can be a difficult and nebulous undertaking even for an experienced analyst. In general, the thrust bearing foundatiori structure is arranged such that an accurate calculation of its spring constant would be formidably complex. For this reason, longitudinal vibration calculatior~sare frequently conducted such that the natural frequency is expressed in terms of the thrust bearing foundation stiffness. Table 10 contains a calculation for the first and second resonant modes of longitudinal vibration of the shafting system modeled as shown in Fig. 19. Table 10 utilizes the mechanical impedance method described in [54] and is arranged such that the resonant frequencies can be plotted in terms of the thrust bearing fourldation stiffness. Figure 20 is such a plot and indicates the accuracy required of the thrust bearing foundation stiffness calculation. In some instances, as may be the case with a ship having a very short run of shafting, an inspection of the appropriate thrust bearing foundation drawings may be all that is required to provide assurance that the resonant modes of longitudinal vibration will be well clear of the operating range. On the other hand, lengthy and sophisticated thrust bearing foundation stiffness calculations may be required in order to ensure that ships with long runs of shafting have satisfactory longitudinal vibration characteristics. The thrust bearing foundation spring constant may be considered to corlsist of three constituents: the bending deflection of the thrust bearing foundation structure above the inner bottom, the shear deflection of the thrust bearing foundation structure above the inner bottom, and the deflection at the thrust bearing due to inner-bottom deflection. An appreciation for the effects of innerbottom deflection may be obtained from reference [55]. Calculation of foundation deflections above the inner bottom entails the usual obscurities associated with estimating the deflection of complicated structures. In ' order to assess the effects of inner-bottom deflection, it is necessary to make simplifying assumptions concerr~irlg the extent (length and breadth) of inner bottom effectivelj. supporting the thrust bearing and the boundary conditions at its extremeties; each system must be individually studied, in light of the degree of accuracy desired, in order to establish appropriate assumptions and calculation procedures. In cases where design constraints make it impossible to design a shafting system such that it is free of objectionable frequencies of resonant longitudinal vibration, use may be made of a "resonance changer. " Resonance changers are discussed in Chapter 20. Briefly, they are thrust bearings which are modified such that the thrust pads are floating pistons. The volume of oil supporting the thrust pads can then be tuned to alter the thrust bearing spring constant and avoid objectionable resonant frequencies.
Longitudinal Vibration Calculations for Shafting System Model Shown by Fig. 19
Table 1 0
N
o0
rpm
70 80 90 95
US
red/sec 43.98 50.27 56.55 59.69
3
395
x 106
Z,, = osMp ib/in. X 106
0.0193 0.0253 0.0320 0.0356
0.4381 0.5743 0.7264 0.8081
tan a1° = Zb/kl€l
kla tan a1°
"=
180
red 0.1694 0.1936 0.2178 0.2298
= lr € 1
deg 9.706 11.09 12.48 13.17
-
kin lb/in. X 10'
tan (elo el0) sZ. k.~. a
3.439 3.930 4.421 4.665
0.1274 0.1461 0.1643 0.1732
a1°
- r1°
=
7.260 8.312 9.330 9.$26
z* =
red 0.3464 0.3959 0.4454 0.4701
e2O
a
Exciting freqiency, o
tan UP = Z./krrp
3.
zc=
kaea tan as0
2LN
-X number of propeller bladee.
8.3 Excitation Fackrs. Longitudinal vibration of propulsion shafting systems is exclusively excited by the variable thrust developed by the propeller due to the nonuniform wake pattern in which it operates. The predominant periodic component of the thrust developed by a propeller occurs a t blade rate frequency; i.e., the number of propeller blades times the rotational frequency of the shaft. Higher harmonics of blade rate frequency occur but, due to their relatively small magnitude, they are generally not of practical importance. A number of factors influence the magnitude of the vibratory thrust; consequently, generalizations in this area must be used with care. Nevertheless, typical ranges of thrust excitation, expressed as a percentage of the mean thrust, are tabulated in Table 3; for additional details concerning thrust excitation, see references [3, 50-52, 591. 8.4 Damping. There are several sources of damping in the longitudinal vibration system, the most important of which is the propeller. Other sources of damping such as hysteresis and sliding friction are relatively small, difficult to estimate, and difficult to treat in the
NOTES: ESTIMATED THRUST BEARING FOUNDATION SPRING CONSTANT 7 X 10' LB/IN RATED PROPELLER RPM I I S RPM NO.OF PROPELLER BLADES. 6
-
0
Fig. 20
10 20 30 40 50 THRUST BEARING FOUNDATION SPRING CONSTANT, K t , L B I I N X 10'
Effect of thrust bmring Youndotion Mness on bngiidinol resonon( frequencies
.
PROPELLERS, SHARING, AND SHAFTING SYSTEM VIBRATION ANALYSIS
MARINE ENGINEERING
calculations; therefore, their effects are usually considered to be lumped with the propeller damping allowance. A procedure for estimating propeller damping is described in reference [54]. The prpcedure entails plottiug the propeller thrust coefficient CT as a function of the true s l i ~the : rate of change - of thrust with respect to the fore-;id-aft vibratory velocity can then be determined from the slope of the CT versus s curve a t the operating slip point. The propeller damping constant, C,,-, is accordindv - - determined to be 1 C, = - P ~ D ' 12 . as
5
lb-sec/in.
(29)
This can be equated to the damping Per cycle D = nC,wXa2
in.-lb/cycle
(31)
where
Cp
= propeller damping constant, lb-eclin. w = resonant frequency of vibration, rad/sec
to obtain the following expression for the amplitude of vibration at the X, = t/wC,
77
Va -Pn
established to be 2.95 X lo8; therefore, F d = 23,600 lb acceleration of the slow-speed gear should not exceed and the ratio of the alternating thrust to the mean 3 ft/sec2;but this recommended criterion is accompanied thrust is established to be 0.215, an entirely acceptable with the statement: "It is emphasized that each particvalue in view of the low power level a t which it occurs. ular case warrants individual attention and that the 8.6 Acceptable Limits for Longitudinal Vibration. thrust variation levels must be given equal conSeveral attempts have been made to enumerate accep sideration." tance criteria for longitudinal vibration characteristics. There are so many variables which must be considered The most well-known criteria are those specified in when analyzing the vipration characteristics of a system reference [ a ] . Other acceptance criteria have stated that there appears tg be no satisfactory alternative to that reversals of thrust in thrust bearings are not condudting an analysis of each particular system and permitted. One criterion [56] has specified that the studying each system individually.
(32)
Once the amplitude of vibration a t the propeller has been established, the alternating force at other points in the system can be determined by using the mechanical impedance method. The procedure of determining the alternating force on the main thrust bearing may be illustrated by referring to Table 10 and Fig. 20. In addition, the following data are required :
P = propeller pitch, ft n = propeller speed, rps D = propeller diameter, f t CT = propeller thrust coefficient, T/n2P2D2 s = propeller slip = 1
t = maximum amplitude of alternating thrust, lb X, = maximum amplitude of propeller vibration, in.
397
+
Section 9 Whirling Vibration 9.1 Introduction. Whirling vibration can best be kinetic energy is a t its maximum. D u r i n vibration ~ the visualized by considering the motion to be-the resultant total energy in the shaft system remains constant; of two shaft v i b r a t i o ~each in perpendicular planes therefore, the potential energy a t the point of maximum passing through the shaft neutral position. Depending iV&kfifitudeisequ91. to the kineticenergy a t the point of t 6 e maximum d < kinetic and potential . upon the manner in which the vibratory mptions com- ~ ~ ~ ~ l i ' T ..gKliTt;. -dByz bine in the two perpendicular planes, the iesultant enem*-re--~etermined'~~e-. masses by the fEK-~freesEiEiee;;"if-:.of motion may be circular (analogous to the motion of a -.+...-". . TK..-.......,*.-,.. ""L. .... skip-rope), elliptical, or in a single plane (if one of the vibration. ese energies so determined may be equated two combining vibrations is of negligible magnitude). and this equation then solved for the critical speed, N, in Visualization of whirling vibration is further complicated cpm, as follows: by the fact that the whirling frequen~ymay be either at the frequency of shaft rotation -or a multiple of shaft rotation, and the whirling motion can be either in the direction of shaft rotation or opposite to the direction I n this formula d W is the weight of a short section of of shaft rotation. Shaft whirling vibration is discussed shafting whose mass may be con$dered concentrated or in references.[47, 61-63]. the weight of any concentrated mass caxried by the 9.2 Determination of Whirling Natural Frequencies. shaft (such as the propeller) and y is the deflection of the There are basically two approaches which may be taken center of gravity of this mass. The summations in evaluating whirling natural frequencies. One method, include all of the masses in the shaft system. which has been used for many years, is to compute the The static deflection curve can be calculated as natural frequency from the Raleigh approximation. I n described in any standard book on strength of materials. applying this method, an assumption must be made with It should be noted, however, that the loads on the regard to the shape of the shaft centerline when the shafting should be reversed in direction in alternate. amplitude of vibration is a t its maximum. I t is not spans; ,that is, the weights are assumed to act down in necessary that the assumed curve have exactly the same one span and up in the next. This reversal is necessary shape as the actual vibration deflection curve, but it to produce a deflection curve which has the same general should have the same general characteristics. I n form as the vibration curve corresponding to the lowest problems concerning the vibration of beams it has been natural frequency. found satisfactory to use the deflection curve correThis ortlculation is readily accomplished by means of sponding to the static loading condition and this computer programs similar to that described in [23]. assumption is considered sufficiently accurate in connec- Such calculations were made for the shafting system tion with propulsion shafting. shown in Fig. 1 and a plot of the mode shape for the The total energy of vibration a t any instant consists whirling mode of vibration is shown in Fig. 21. It may of two parts, kinetic ener_g;y_.due-Jovthe motion of the be seen from Fig. 21 that the large amplitudes of vibration are confined to the aftermost regions of the shafting shaft-m~?es ~~te1?tk!. stresses caused-, by the sb.scft system. From an inspection of the Raleigh equation it is evident that only the regions of the shafting system maximum .-- amplidge . the_.-nana~stktionary,~~t~eJinetic energy is-_. which have relatively large amplitudes have an important energy stored up ~ n " i ; g i i s a its t maximum. On the impact on the whirling critical frequency. This fact can is other hand, when the a m p l i t u ~ e ~ o " B ~ enoe bendbe exploited to greatly simplify the problem when the ing in the shaft, so&g*gti-dle , but the calculations must be done by hand.
cti6iiiii,=-t;Ke
. *
V, = propeller advance velocity, fps A somewhat different approach for determining system damping was taken by Rigby [55]. Rigby used fullscale data to calculate an equivalent propeller damping coris$ant and concluded that for three-bladed propellers the equivalent damping constant was about 16.5 lb-sec/in. times the propeller developed area in square feet; he further concluded that the propeller damping constant tends to increase with larger numbers of blades and suggested that a damping factor equal to 39 lb-sec/in. per foot of blade edge may give better results based on tests made with four- and five-bladed propellers. In cases where the thrust bearing is located well aft and there is a significant amplitude of vibration a t the slow-speed gear, investigations made and reported in [54] indicate that machinery damping must also be considered; feference [54] suggests a procedure which may be used to include the effects of machinery damping. 8.5 Vibratory Thrust Calculations. A meaningful indication of the importance of a resonant mode of longitudinal vibration can be obtained by assessing and comparing the alternating thrust component to the mean thrust component at the main thrust bearing and the vibrat~ry'am~litude, velocity, and acceleration a t the slow-speed gear. This can readily be qccomplished by assuming that the only source of system damping is the propeller (the restriction need not be quite so severe in that an "equivalent propeller damping" may be used which incorporates other system damping effects) and that the propeller is also the only source of excitation. With this asspmption, the propeller input work per cycle can be expressed as
Mean propeller thrust a t 75 rpm (thrust is assumed to vary as the rpm squared). . . . . . . . . . . . . . . . . . . . Propeller damping constant (based on reference [55]) . . . . . . . . . . . . . . . . . . .
~
110,000 lb 5,800 lb-sec/in.
Ratio of alternating to mean propeller thrust (based on an analysis of the propeller wake). . . . . . . . . . . . . . . . . . 3 percent With these data the exciting force from the propeller is established to be 3,300 lb and from equation (32) the amplitude of vibration a t the propeller is X, = 3,300/[2~(75)(6)/60] [58001 X, = 0.012 in. The procedure outlined in [54] can now be used to establish the amplitude of vibration a t the thrust bearing (point d in Fig. 19). Based on reference [54], the amplitude of vibration a t point b becomes Xb = X, cos (tan-' Zb/klal)/cos (tan-' Za/klal) (33) and the amplitude of vibration a t point c becomes X, = Xb cos (tan-' Z,/k2a2)/cos (tan-' Zb/kz€z)
(34) (See Table 10 for a definition of terms.) A numerical evaluation of the foregoing expression indicates a vibratory amplitude of f0.008 in. a t point d (the same as c). This is the amplitude at the slow-speed gear from which the velocity and acceleration can be calculated knowing the frequency of vibration. The alternating force a t point d is then determined from
, By interpolating the data shown in Table 10, Zd is
,
,'
"-.s-m-
I
I X
,a.,z...-m*
% ,. .,
~
MARINE ENGINEERING
PROPELLERS, SHAFTING, AND SHAFTING SYSTEM VIBRATION ANALYSIS References
1 F. H. Todd, "Resistance and Propulsion," Prin-
C
Fig. 21
Mode shape of shafting arrangement shown in Fig. 1 during whirling mode of vibration
Computation of whirling critical frequencies entails several approximations which are easily criticized. One is the assumed location of the resultant reaction in the bearing just forward of the propeller. There is no way of assessing the load distribution in this bearing; furthermore, in the case of water-lubricated bearings it changes in service. With shafting arrangements. similar to that shown in Fig. 2 (closely spaced stern tube bearings), the forward stern tube bearing can become unloaded or even develop a downward reaction. Since the load condition of the forward stern tube bearing strongly affects the whirling natural frequency, computations are usually made with it loaded and also unloaded, particularly with water-lubiicated bearings which are subject to large amounts of wear. - The effects of entrained water are normally approximated by increasing the weight of the propeller by a percentage (usually 25 percent). Bearing flexibilities are normally neglected in the Raleigh calculations as the ~ri>blemwould otherwise be considerably more corn~licated. However, neglecting bearing flexibility may not be justified if the bearing is softly mounted in rubber to achieve self-aligning capabilities. Propeller gyrosco~iceffects which tend to stiffen the 'ystem are imilarly neglected. Fortunately the inaccuracies associated with considering bearings rigd and neglecting propeller gyroscopic effects tend to cancel. N. H. Jasper developed [61, 62, 631 a calculation procedure which takes bearing flexibility and propeller gy~oscopic effectsinto account. The procedure is relatively simple and is feasibly done by hand calculation due to the fact that only the aftermost regon of the shafting is considered. The results obtained by the procedure developed by Jasper compare favorably with the results obtained by using the Raleigh method; this is , attributed to the opposite effects of the additional factors taken into consideration.
9.3 Acceptable Limits for Whirling Vibration. Important sources of whirling vibration excitation are propeller and shafting unbalance and occur at a frequency corresponding to propeller rpm. Also, where Z is the number of propeller blades, the kZ f 1 harmonics of the propeller wake field produce exciting frequencies; however, these are generally not significant due to the small exciting force and the dampening resulting from the relatively higher frequency. Only the frequency cormpondihg to the propeller rpm is considered to be important by some authorities; however, this point can be debated. I n any case, it must be agreed that the severity of excitation a t blade rate frequencies must be investigated for a specific case before they can be categorically dismissed. I n order for the whirling natural frequency to be coincident with the propeller rpm, conditions considerably different from those on the usual large ship must exist. For example, bearing spans would have to be abnormally long, shaft diameters would have to be abnormally small, a bearing would have to become etc. whirling f ~ unloaded for some corresponding to blade rate frequencies can and do fall within the operating range. Fortunately, however, the exciting forces at blade rate frequencies are generdly unless coupled with other adverse not of great conditions (such as the forward stern tube bearing becoming unloaded). Perhaps the only generalization which Can be made with respect to acceptance criteria for whirling vibration is that shafting arrangements should be d e s i ~ e dsuch that, in the upper operating range, whirling resonant frequencies do not become in close proximity to the propeller rpm or the blade rate frequencies which have strong exciting forces. Blade rate exciting forces can be evaluated as discussed in Section 3.2.
.
399
Standardization of Propulsion Shafting and Bearings, " Afarine Technology, vol. 3, no. 2, April 1966. 20 Archer M. Nickerson, Jr., "Summary Report of the Investigation of Tailshaft Failures," prepared for SNAME under Order 352, January 1966. 21 J. W. Heck and E. Baker, "Marine Propeller Shaft Casualties," Trana. SNAME, vol. 71, 1963. 23 "Propulsion Shafting," Design Data Sheet DDS 9430, Department of the Navy. 23 E. T. Antkowiak, "Calculation of Ship Propulsion Shafting Bearing Reactions on an IBM 650 Computer," Boston Nav J Shipyard, Development Report No. R-1 1, October 1957. 24 W. E. Lehr, Jr. and E. L. Parker, "considerations in the Design of hlarine l'ropulsion Shaft Systems," Tratls. SNAAiE, vol. 69, 1961. 25 H. C. Andersen and J. J. Zrodowski, "Coordinated Alignment of Line Shaft, Propulsion Gear, and Turbines," Trans. SNAAfE, vol. 67, 1959. 26 H. C. Andersen and D. E. Bethune, "Summary of Studies on Recent Ship's Main Propulsion Lineshaft Systems and Recommendations for l'uture Designs," General Electric Co. Report DF1-MSD-200 dated June 12, 1961. 27 "Propulsion Shafting and Components," NavShips Drawing 510-2146000. 28 L. L. Shook, Jr. and C. L. Long, "Surface Cold Rolling of Marine Propeller Shafting," Trans. SNAME', vol. 66, 1958. 29 "Rubber, Shaft Covering Materials (For Marine Propeller Shafts)," Military Specificatio~lMIL-R-15058. 30 "Instruction Manual Glass Reinforced Plastic Coatings for Propeller Shafting, " NAVSHIPS 250-634-4. 31 "I~lvestigationof Propeller-Hub Sealing ArrangeNovember ments," Technical 1959. and Research Bulletin 3-7, SNAME,
ciplesof Naval Architecture, J. P. Comstock, ed., SNAME, 1967. 2 F. M. Lewis, "Propeller-Vibration Forces, " Trans. SNAAfE, vol. 71, 1963. 3 F. H. Todd, Ship Hull I7ibratio1~, Edward Arnold Ltd., London, 1961. 4 T. Iiumai and Y. Sakurada, "On the Measurements of Propeller Surface Force of the Self-Propelled Model of a Tanker," Europea?~SAipbuildi~~y, no. 3, 1966. 5 R. J. Boswell and M. L. Miller, "Unsteady Propeller Loading-Measurement, Correlation with Theory, and Parametric Study" 'XSRDC Report 2626, October 1968. 6 J. H. ~ c ~ a r t "On h ~ the , Calculation of Thrust and Torque Fluctuations of Propellers in Nonuniform Wake Flon-," DTMB Report 1533, October 1961. 7 H. Y. Yeh, "Thrust and Torque Pluctuatio~lsfor APA 249, TMB Model 4414," DTMB Report 1364, June 1960. 8 L. C. Burrill, "Calculation of Marine Propeller Performance Characteristics, " Trans. North East Coast Institute of Etlgineers and Shipbuilders, vol. 60, 1943-44. 9 G. R. Stuntz, Jr., P. C. Pien, W. B. Hinterthan, and N. L. Ficken, "Series 60-The Effect of Variations in Afterbody Shape Upon Resistance, Power, Wake Distribution, and Propeller Excited Vibratory Forces," Trans. SNAME, vol. 68, 1960. 10 J. B. Hadler and H. M. Cheng, "Analysis of Experimental Wake Data in Way of Propeller Plane of Single and Twin-Screw Ship Models," Trans. SNAME, vol. 73, 1965. 11 N. H. Jasper and L. A. Rupp, "An Experimental and Theoretical Investigation of Propeller Shaft Failures," Trans. SNAAlE, vol. 60, 1952. 12 E. P. Panagopulos and A. M. Nickerson, Jr., 32 R. Baudry and L. M. Tichvinsky, "Performance "Propeller Shaft Stresses Under Service Conditionsof Oil Rings," ASME paper, 1936. The SS Chryssi Investigation, " Trans. SNA ME, vol . 62, 33 "Report of Ship Operators' Experience with 1954. Stern Tube Bearing Wear," Technical and Research 13 H. R. Neifert and J. H. Robinson, "Further Bulletin No. 3-12, SNAME, September 1962. Results from the Society's Investigation of Tailshaft 34 W. J. I
-
2-
~
MARJNE ENGINEERING
39 Richard Hecker, "Powering Performance of a Ventilated Propeller," DTMB Report 1487, June 1961. 40 K. E. Schoenherr. "Formulation of Pro~eller * Blade Strength," Trans. SNAME, vol. 71, 1963. 4 1 I. K. Mott and R. Fleeting, "Design Aspects of M , k e Propulsion Shafting Systems," The Institute of Marine Engineers, June 1967. 42 "Manual of Instructions for Design and Application of Propeller Blade Gages," NavShips 387-0547, Philadelphia Naval Shipyard, November 1964. 43 J. H. Smith, "Nodal Arrangements of Geared Drives," Trans. Institute of Naval Architects, 1922. 44 F. M. Lewis, "Torsional Vibration in the Diesel Engine," Trans. SNAME, vol. 33, 1925. 45 F. M. Lewis, "Dynamic Effects," A4arine Engineering, H. L. Seward, ed., SNAME 1944. 46 S. C. Powell and W. V. Bassett, "Practical Aspects of Torsional Vibration in Marine Geared-Turbine Propulsion Units, " Trans. SNAME, vol. 52, 1944. A7 Eugene Panagopulos, "Design-Stage Calculations of Torsional, Axial, and Lateral Vibrations of Marine Shafting, ' Trans. SNAME, vol. 58, 1950. 48 A. D. Andriola, "Torsional Vibration in GearedTurbine Marine Propulsion Plants-An Introduction to the Subject," Trans. SNAME, vol. 58, 1950. 49 L. C. Burrill and W. Robson, 'Virtual Mass and Moment of Inertia of Propellers," North East Coast Institution of Engineers and Shipbuilders, Newcastle upon Tyne, 1962. 50 A. J. Johnson and W. McClimont, "Machinery Induced Vibrations," Trans. The Institzlte of Marine Engineers, 1963. 51 R. .wereldsma, "Propeller Excited Shaft and Hull Vibrations of Single Screw Ships," International Shipbudding Progress, December 1964.
52 S. Archer, "Propeller Excited Vibration; Five Blades or Four?" International Shipbuilding Progress, March 1962. 53 S. Archer, "Torsional Vibration Damping Coeficients for Marine Propellers," Engineering, May 13,1955. 54 J. R. Kane and R. T. McGoldrick, "Longitudinal Vibrations of Marine Propulsion-Shafting Systems," Trans. SNAME, vol. 57, 1949. 55 C. P. Rigby, "Longitudinal Vibration of Marine Propeller Shafting," Institute of Marine Engineers, February 1948. 56 A. A. J. Couchman, "Axial Shaft Vibration in Large Turbine-Powered Merchant Ships," Trans. Institute of Marine Engineers, March 1965. 57 G. P. Antonides, "Longitudinal Vibration of Propulsion System on USS Simon Lake (AS33)," DTMB Report 2147, January 1966. 58 A. A. J. Couchman, "Mechanical Exciter and Axial Vibration Measurements on the Thrust Block and Shafting of a 48,000-Ton DWT Tanker," British Ship Research Association Report No. 8, 1963. 59 R. T. McGoldrick, "Ship Vibration," DTMB Report 1451, December 1960. 60 "Mechanical Vibrations of Shipboard Equipment, " MIL-STD-167 (Ships). 61 N. H. Jasper, "A Theoretical Approach to the Problem of Critical Whirling Speeds of Shaft-Disk Systems," DTMB Report 827, December 1954. 62 N. H. Jasper, "A Design Approach to the Problem of Critical Whirling Speeds of Shaft-Disk Systems," DTMB Report 890, December 1954. 63 N. H. Jasper and L. A. Rupp, "An Experimental and Theoretical Investigation of Propeller Shaft Failures," Trans. SNAME, vol. 60, 1952.
CHAPTER XI1
0. W. Soete
1.
I
Pumos. Forced-Draft Blowers, Compressors, and Ejectors I
'
Section 1 Centrifugal Pumps 1.1
Basic Fundamentals
a. Classification and Types. The basic purpose of a centrifugal pump is to maintain a flow of fluid through a system of piping connecting two reservoirs. If there are differences in fluid pressure or static head between the two reservoirs, the pump must supply the energy ,corresponding to the work done in overcoming these differences, plus the energy loss due to friction in the piping system. An example of an elementary system is shown in Fig. 1. A simple type of radial-flow centrifugal pump is shown in Fig. 2. The caaing is of the volute type and the impeller is double-suction. As there is only one impeller, it is known as a single-stage pump. The vertical position of the shaft axis, the pump base a t the bottom, and the
Fig. 1
Elementary application of a centrifugal pump
driver location above contribute to its being known & a vertical centrifugal pump. A single-suction pump is shown in Fig. 3. The fluid passes from the first-stage impeller to the second-stage impeller; that is, the head is developed by two impellers in series, and it is known as a two-stage pump. This pump is also of vertical construction. Multi-stage pumps of horizontal construction ' are shown in Figs. 4 and 5. Both are of the volute casing type. In each example, the pumped fluid discharges from one impeller and passes to the next-stage impeller through an external crossover passage not indicated in the figure. Pumps are also classified as radial, mixed, ' A d axialflow types according to the direction of the discharged flow. Radial-flow pumps are depicted in Figs. 2 through
OEAERATINQ f EEOWATER HEATER
MARINE ENGINEERING
PUMPS, FORCED-DRAFT BLOWERS; COMPRESSORS, AND EJECTORS
. Fig. 4
Horizontal, axially split casing, two-stage boiler feed pump
PPP
Fig. 2
CASING IMPELLER IST STAGE IMPELLER 2ND STAGE IMPELLER WEARING RING CASING WEARING RING INTERNAL BEARING SHAFT SHAFTSLEEVE JOURNAL SLEEVE BEARING HOUSING MOTOR BRACKET PUMP FOOT
Fig. 3
P
Fig. 5
Horizontal, axially split casing, fou~stageboiler feed pump
1
Vertical, axially split casing, two-stage pump
Vedical, axially split casing, single-stage, double-suction pump
5. A pump of the mixed-flow type is shown in Fig. 0 and an axial-flow type, sometimes known as a propeller type, is shown in Fig. 7. These last two examples are singlesuction, single-stage, vertical pumps. They are also known as end-suction types, whereas the examples of Figs. 2 to 5 are known as side-suction types. A small pump of the radial-flow, end-suction type is depicted in Fig. 8. The essential detail parts of a pump are shown in Figs.
P
CASING IMPELLER STAGE PIECE CASING WEARING RING SHAFT SLEEVES THRUST BEARING LINE BEARING
LEGEND
LEGEND CASING IMPELbER CASING WEARING RINGS IMPELLER WEARING RINGS SHAFT .SHAFT SLEEVE STUFFING BOX PACKING GLANDS THRUST BEARING LINE BEARING BEARING HOUSINGS PUMP BASE MOTOR BRACKET COUPLING
1 2 3 4 5 6 7 8
2 to 8. Wearing rings are fitted in closed-type impellers to prevent discharge or high-pressure fluid from passing to the lower-pressure zone at the suction side of the impeller. They may be fitted to both the impeller and casing, as in Figs. 2 and 3, or may be fitted only to the casing, as in Figs. 4 and 5. They serve to protect the impeller and casing in that, when wear occurs at themclose running clearance, the wear is confined to the relatively inexpensive rings which can be easily replaced. In an open type of impeller, such as in Fig. 7, the ends of the impeller blades form the wearing surfaces, requiring
relatively close clearances. The impeller shaft must be needed"to carry any residual axial thrust as well as the sealed against leakage by a stuffing box. rotor weight. The pumps of Fig. 4 and 5 are inherently Single-suction impellers generally require some means axially balanced in that the two and four impellers, reof reducing the axial unbalanced thrust. In many amall spectively, are positioned ''back to back." pumps the axial unbalance is carried by the pump thrust b. Expression for Total Pump Head. The following bearing. The double-suction pump of Fig. 2 is inherently system of notation is employed with subscripts as indiaxially balanced, but requires a thrust bearing to carry cated: . the weight of the rotor. The two-stage pump of Fig. 3 p = fluid pressure, psi is almost completely axially balanced; this is ach, = friction loss, ft complished by positioning the impellers "eye to eye," by Z = static head, ft proportioning the wearing rings so as to aid in the reducp = weight density, pcf tion of thrust, and by providing a balancing hole at the V = fluid absolute velocity, fps center of the shaft in the first stage. A thrust bearing is H = total pump head, ft
-
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS .
MARINE ENGINEERING
Fig. 6
Mixed-low-type pump
THRUST BEARING SLINGER THRUST BEARING COVER LlNE BEARING SLINGER LlNE BEARING COVER
LEGEND
Referring to Fig. 1, the energy increase in foot-pounds per pound of fluid transferred from point 0 to point 3 is 144bs - po)/p
+
2 8
- zo
(1)
The pump must also supply the total friction loss, hr, in the piping. Thus, the total work done by the pump in foot-pounds per pound of fluid is the sum of these quantities, or H=l44(ps-po)/~$Z8-Zo+hr
M N
(2)
It is desirable to express the total pump head in terms of the energy conditions per pound of fluid at the pump suction and discharge connections, points 1 and 2 of Fig. 1. This is done by application of Bernoulli's equation for the conservation of energy, with an appropriate term for the friction loss. First consider points 0 and 1, and then points 2 and 3, with the following results:
Fig. 8
R = impeller radius, f t D = impeller diameter, f t B = impeller passage width, f t
+
Fig. 7 Axial-low, or pmpelle~typepump
lab2
- PI)/P
(5)
This equation simply expresses the fact that the total pump head is obtained by the difference between the total pressure at the dischaxge and the total pressure at
Radial-low, end-sudon, centrifugal pump
the suction of the pump. When the latter is less than atmospheric pressure, it is generally called the suction lift and is ordinarily expressed in feet, or as a vacuum in inches of mercury. It should be noted that, whereas the aymbol H represents feet of fluid, it also represents footpounds per pound of fluid. c. Velocity Diagrams and Theoretical Pump Head. The principal elements of a radial-flow centrifugal pump, consisting of an impeller with a central inlet and curved vanes and an enclosing casing, are indicated in Fig. 1. The theoretical fluid velocity diagrams for the impeller inlet and outlet are shown in Fig. 9. The following additional notation is used:
Thus, when Z1 = Z2 and VI = V2, as is the case for equal suction and discharge pipe areas and elevations, and further noting that hnl hn8 = hr, it results that the total head is given by H =
SPRING BELLOW DRIVING BAND SEALING RING FLOATING SEAT SEATRING PUSHCOLLAR
impeller angular velocity, rad/sec impeller speed, rpm impeller speed, rps impeller peripheral velocity, fps angle between peripheral and absolute velocities, deg B = angle between peripheral and relative velocities, del3
o = = n = u = a =
N ,
Fig. 9
Velocity diagrams for a radial-low pump
v = fluid relative velocity, fps ,= fluid kinematic viscosity, sq ft/eec TV = weight rate of flow, lb/sec
PUMPS, KIRCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
406
MARlNE ENGINEERING
+
the discharge velocity triangle in Fig. 9, uz" VZ2 ~2~ - 2~2V2cos a 2 . Similarly, at the entrance velocity triangle, q2 = V12 ulZ - 2ulVl cos al. Noting that Vu2 = Vz cos a 2 and Vul = Vl cos all substitution of Vu2 and Vul in the foregoing expressions results in 2ulVu2 = VzZ Uz2 - vz2 and2ul?,,l= V12 u12 - q2. Substitution The expression for theoretical impeller torque is ~ b of these latter expressions in equation leads to the tained by assuming that the flow conditions at the imfollowiog version of Euler,s equation: peller inlet and outlet are perfectly uniform; that is, every particle of fluid hrts the same velocity entering the impeller, and a similar situation exists at the impeller exit. - It is necessary also to assume that the entire flow is steady (that is, independent of time), but other than The first term in equation (11) represents the change in this the conditions of flow within the rotor can be of any kinetic energy of the fluid as it passes through the impeller. In order that this energy change may contribute to sort whatever [I].' the total head, there must also occur corresponding presFollowing the treatment of references [2] and [3], if in Fig. 9 a fluid particle enters the impeller through area da, sure changes elsewhere than in the impeller. Assuming then the moment of momentumof thisfluidparticleismass equal suction and discharge diameters (s and d) so that V8 = Vd, then (Vd2- V12)/2g is the increase of pressure per unit time X velocity X lever arm or energy a t the inlet to the impeller. Also (Vz2 - Vd2)/2g represents the conversion of kinetic to pressure energy in the discharge casing, and the Bum of these two energy Integrating over the impeller inlet area all the total changes is equal to the first term in the foregoing expresmoment of momentum of the entering fluid is sion for H I . The last two terms represent the pressure changes occking across the impeller itself. The term (us2 u12)/2g containing the impeller peripheral velocities is the The foregoing equation applies to incompressible increme of pressure due to centrifugal force, alld the final as P is taken to be constant. In like manner, the moment term (&z - v1~)/2gis pressure change resulting from of momentum of the fluid leaving the impeller is the change of relative velocity of the fluid. In an axialflow pump, this final term is the entire change.of pressure . VzVu2R2h across the impeller, but is not the entire head except in the case where V2 = Vl. Since in most practical cases for pumps, Vu2R2> VulR1, e. The Head-Capacity Curve, Effect of Shape of and the impeller exerts a net torque on the fluid, the Impeller Vanes. In the absence of stationary guide vanes total theoretical torque is at the inlet, a desirable simplification is obtained by assuming that the absolute fluid velocity at the inlet is radial. Thus, there is no peripheral component regardless of the shape of the impeller inlet (that is, Vul is zero). (8) Sincse Vl is perpendicular to ul, it follows that v12 - V12 - u12 equals zero because the three terms are respecIn the foregoing, Vu2R2and VulR1are assumed to be con- tively the hypotenuse and adjacent sides of a right tristant over the two surfaces of integration. angle. The expressions previously given for the theoretiSince Vldal and V2da2are equal and represent the total cal head become rate of flow through the impeller, then Q = volume rate of flow, cu ft/sec q = volume rate of flow, gpm T = theoretical rotor torque, ft-lb H , = theoretical pump head, ft , a = acceleration due to gravity, ft/sec2 -
+
+
+
's"
II
The expressioti for work is obtained by multiplying the impeller torque by the angular velocity o. After dividing through by the weight rate of flow, there is obtained the usual equation foi the theoretical head: 1 (10) 9 d. Physical Interpretation of Theoretical Head. The foregoing ex~ressionfor the theoretical head may be traniforked b y means of the cosine law. ~ e f e r r i i gto
Ht
=
To
--
w
=
- (ulVu2- ulVul)
Numbers in brackets designate References at end of chapter.
LEGEND TUEDRETICAL HEAD
To show the relation of theoretical head to the volume flow, or capacity Q, it is necessary to make another application of the cosine law, whereby 9 The im~ellerexit area Az (in the diredtion of vz) is rD2B2 sin '& in square feet (neglecting the vane tip?, and 02 equals Q/Az, SO that
3.6
ACTUAL HEAD
I- CIRCULATING PUMP, RADIAL-FLOW TYPE (a) NORMAL CHARACTERISTIC,& =2?.s0
3.4
3.2
+ 3.0
j
2.8
5
2.6
P
2-FEED PUMP. RADIAL-FIDW
TYPE
STEADILY RISING CHARACTERISTIC,&~I~~ 3-DREDGE PUMP, RADIAL FLOW T Y P E , ~ ~ = ~ O ~ 4-CIRCULATING PUMP, AXIAL-FLOW TYPE AXIAL-FLOW CHARACTERISTIC,~m=240
2.4
a
K 2.2
t*
2.0
a
1.8 \
2
1.6
FLOW RATIO
=
FLOWIRATED FLOW
Fig. 10 Theoretical and actual charactariatics of various centrifugal pumps
This equation shows that for a constant rotational speed and with a discharge angle & less than 90 degrees (backward-curving vanes), the theoretical head decreases as the capacity is increased, following a straight line of downward slope. This is illustrated by curve l(a) in Fig. 10. In this figure the head-capacity characteristics are shown in ratio form, with all heads, both theoretical and actual, being divided by the actual head at rated capacity and plotted against the corresponding capacity divided by rated capacity. If the discharge angle is 90 degrees (radial vanes), tan & in the foregoing equation is infinite, the second term of the equation is zero, and the theoretical head is a horisontal straight line independent of the volume as shown by curve l(b). Likewise, if the discharge angle is made greater than 90 degrees (forward-curving blades), then tan /32 is negative and the theoretical head increases with capacity as shown on curve 1(c) . The actual head-capacity characteristic is always less than the theoretical, and the various curves of Fig. 10 reveal the wide discrepancy between the theoretical and actual heads. In the region of the designed capacity, the ratio of actual to theoretical head may be anywhere from
0.25 to 0.75 and this ratio is usually smaller when the capacity is greater or less than the design value. This discrepancy cannot be explained by the elementary theory previously outlined, which necessarily assumes an infinite number of guide vanes of zero thickness. However, the '(circulation theory" of hydrodynamics shows that for a finite number of vanes there must exist a circulation or eddy flow within the impeller which has the effect of reducing the mean peripheral component of the absolute exit velocity [2]. This has been confirmed by experimental work. Curveq 2 and 3 depict the characteristics of other radial pump types. The characteristics of the axial-flow type of pump are shown by curve 4. The theoretical headcapacity curve of an axial-flow pump rises steeply, while the actual head-capacity curve has an irregular shape, being relatively flat near design capacity but rising sharply at lower capacity and falling rapidly at higher capacity. The elementary theory for this type of pump is the same as for the radial type, but as shown in Fig. 11 there is a different velocity diagram for every radius at the pump suction and discharge. Thus, in setting up an equation for the theoretical head, it is necessary to use the velocity diagrams at a mean discharge diameter Dm.
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
11
1.
I
CYLINDRICAL SECTION NEAR RIM. R,
approach is one in which actual test results are used in the development of various correlations. The applic* tion of dimensional analysis that follows is a logical derivation of the various centrifugal pump design constants. Primary interest is focused on the volume flow Q, in cubic feet per second, and the actual net head H, in feet of fluid, of a centrifugal pump, but it is also recognized that the most important physical characteristics of a pump are the speed w, in radians per second, and the impeller discharge diameter Dl in feet. Other physical characteristics, such as inlet diameter, width of impeller passages, inlet and discharge pipe diameters, and the corresponding area. or surfaces are described by ratios of these quantities to the basic diameter D or D2,depending on the number of linear dimensions involved. Two physical properties of the fluid may enter into consideration. These are the weight density p, in pounds per cubic foot, and the kinematic viscosity v, in feet squared per second. The acceleration of gravity g, in feet per second per second, must also be taken into account. Thus the problem contains seven distinct physical quantities, all expressed in three fundamental unitslength in feet, force in pounds, and time inxeconds. The basic theorem of dimensional reasoning holds that a general relation between all seven physical quantities may be expressed in seven less three, or four compound quantities, each of zero dimensions. One general expression of this form is
CYLINDRICAL SECTION NEAR ,HUB, %
I a1 "RI
Fig. 1 1
(1) Impeller inlet vane shock losses. (When diiuser vanes are fitted in the discharge casing, there are also similar losses a t the inlet to these vanes.) (2) Impeller and diffuser vane exit losses, due to eddies formed by the edges of the vanes. (3) Friction losses in the inlet section, impeller, diffuser, and diiharge casing; similar to the friction losses in piping. (4) Additional eddy and turbulence losses in the impeller and discharge casing where kinetic energy is converted into pressure energy. I
I
charge casing back to the inlet through the impeller running clearances. (7) Balance-device leakage where the rotor axial thrust is equalized by a balancing drum or disk. (8) Friction losses in bearings and stuffing boxes, including thrust bearings. The foregoing losses can be calculated approximately from appropriate special theory for each type of loss, and the net pump efficiency determined accordingly. However, it is generally more satisfactory to correlate actual test efficiencies of centrifugal pumps from the laws of similarity; this method of approach is developed in the following. The water horsepower, Pw,for a pump is given by and for a brake horsepo.cver, P, the phmp efficiency, El equals Pw/Por
1" the foregoing, Q is the volume flow in Cu ft/~ec. If q represents the volume flow rate in gpm and p'the head in 1. psi, the efficiency becomes
E = - Pq All of the above losses affect the head generated and (17) 1714P therefore contribute greatly (but not exclusively) to lower the value of the actual head. The following losses affect Additional detsils regarding pump design only the power input: tions are contained in references [3] through [a]. 1.2 Laws of Similitude. Purely theoretical reasoning (5) Frictional losses a t the exterior surfaces of the impeller, similar to windage losses of steam turbine wheels. does not at present afford a very satisfactory basis for the (6) Flow losses due to leakage of fluid from the dis- study of centrifugal pump performance. A more direct
equal -to rD. Thus, it is seen that this last group of terms has the usual Reynolds number (R,) form, except that it is inverted. This group will therefore be of importance when considering the effects of frictional forces on the pump performance, but usually it is secondary in importance to the more general characteristics indicated by the first three dimensionless groups. Since the basic groups in equation (19) are dimensionless, they may be combined without a loss of generality. Thus, multiplying the two middle groups gives gH/ n2DW and: denoting the head coefficient by #,
The first group in equation (19) is the specific capacity, q,, divided by r, that is,
This expression for specific capacity states the wellknown relationship that, for a pump of given diameter, Q varies directly as the speed n. Or, for similar pumps a t a constant speed, Q varies directly as the cube of the diameter. To be correct, the head and flow rules as expressed by equations (20) and (21) must be taken together, and, therefore, if the capacity changes in proportion to the speed, then the head changes in proportion to the square of the speed. Likewise, for geometrically similar pumps, if the capacity changes in proportion to the cube of the diameter a t a constant speed, then the head changes in proportion to the square of the diameter. The specific speed for a particular pump may be deThe requirement that each group must have zero fined as the speed required by a pump of similar design, dimensions makes it possible to determine the values of shape, and hydraulic characteristics to develop a head of the exponents a, b, 6, etc., in the foregoing expression. 1 foot when delivering a t a volume rate of 1 cubic foot This is done by substitution of the dimensional formula per second. Thiis, if a set of values D, n, H, and Q are for each physical quantity in the groups. For example, selected for a given pump (usually the rated values), the first group must then have the form then, in order to have similar characteristics, a second pump must be such that its corresponding physical quantities will give numerical values for all the basic dimensionless groups equal to those obtained with the original pump. hence If Dl, nl, HI, and Q1 represent the physical quantities of the second pump, then the following relations satisfy the requirements for equality of all dimensionless groups:
"RZ
Velacity diagrams for an axial-flow pump
It is seen in Fig. 11 that the development of a cylindrical section through the impeller is a series of vanes resembling air-foil sections. This is the basis for the calculation of axial-flowpump performance from airfoil theory, which has given results that are surprisingly close to the actual pump characteristics. f. Pump Losses and Efficiency. Flattening of the discharge angle, as previously described, causes actual flow conditions at the impeller outlet different from those assumed in the foregoing elementary theory, but it should not be inferred that the difference between the actual and ideal theoretical heads represents a corresponding loss in efficiency. The principal losses in centrifugal pumps are as follows:
409
E
Thus, the first group in equation (18) is found to be Q/wD8, and when the same process is carried out for the remaining groups, with n-n substituted for w, the general expression becomes
The last term in equation (19) may also be written v/uD = l/R,, since the impeller peripheral velocity u is
However, Q1 = H1 = 1 by definition. Combining all three equations:
k
f
I.# 1.6
0
g 1.4 D:
5 3
1.2 1.0
W
50 0.8 $
0.6
22 0.4 2 0.2 %
Y
o
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 FLOW RATIO r FLOW / RATED FLOW
Fie. 12
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
41 0
Copacitphead and efficiencycharacteristics of a cedtrifugal pump at various speeds
QIHl -= r8nl8Dl6
1 - QH a8nlaD16 . aanaD5
or
-
multiplying it by the term n/(2g)'/'. Specific speed is the most important single index to the potential performance characteristics of a centrifugal pump. After some limited experience one can examine the shape of a centrifugal pump impeller and make a fair estimate of its specific speed. Then, from curves such as Fig. 14, the efficiency and other performance characteristics of the pump can be estimated. Of course, the brake horsepower , level is also an important index. One great value of the parameter specific speed is that it expresses the requirements for similarity of flow conditions in terms of the three essential pump characteristics-capacity, head, and speed-independently of any physical dimensions ~f the pump. a. Performance Curves for a Single Pump. By means of the dimensionless constants derived in the foregoing, it is ,possible to extrapolate the results of a constant-speed test of a particular centrifugal pump to predict its performance at other speeds and also, within reasonable limits, to predict the performance of any geometrically similar pump of different size. The extrapolation process for a single pump is illustrated in Fig. 12. The solid curve represent, in ratio form, the capacity-head and capacity-efficiency characteristics from actual tests at a constant speed N. To predict the characteristics at a different speed NI, select any point (Q, H) on the test curve and calculate new values (Q1, HI) for the new speed such that the dimensionless constants II. and q, remain the same. Thus for $1 = II.,
-
By combining only the last two equations:
Also, since qel = q,, it follows that
and
It should also be noted that since the head coefficient II. and specific capacity q, have remained unchanged, the specific speed N,, is also the same. The efficiency would also be expected to remain the same, except for the influence of losses which vary according to laws other than the fundamental forms in the foregoing. Test results, such as those shown in Fig. 12, show the efficiency to be essentially constant. The principal results are again summariaed in the familiar rule that if the capacity changes in proportion to the speed, then the head changes in proportion to the square of the speed, the efficiency remaining relatively constant. Also, since the horsepower is equal to capacity times head, it is seen that the bralce horsepower varies directly as the cube of the speed, provided the efficiency is constant. The foregoing rules must be applied cautiously, taking into consideration deviations due to changes in Reynolds number, size effect, and cavitation. Also illustrated in Fig. 12 are curves representing the normal system head and the throttled discharge head. A particular condition of capacity and head is produced at the point of intersection of the pump performance
and finally
Usually the units of rpm and gpm are used. The symbol N,, is sometimes used for specific speed (to distinguish it from n,) and the expression becomes:
The specific speed as defined in the foregoing is not a purely dimensionless number, but this makes little difference since it can be made dimensionless simply by
~
.
~
(
I
I
I
I
I
I
I
I
I
I
~
~
FLOW RATIO= FLOW/ RATED FLOW fig. 13
Characteristic curves for a aeries of geometrically similar pumps
curve with the system head curve. It is customary to specify the rated pump performance at or near the best efficiency point (abbreviated BEP), and it w ill be noted that the pump illustrated in Fig. 12 has a BEP (capacity ra&o = 1.0) almost on the normal system head curve. The capacity-head curves of Fig. 12 tend to approach zero at relatively high rates of flow. It is uncommon to specify pumps for such a broad range of performance due to the deleterious effect of high velocities beyond the BEP, which may cause erosion of the internal parts, and the extreme low head which may cause an upset of the internal axial hydraulic balance. When applying centrifugal pumps to a specific set of performance requirements, an operating flow limit twenty percent beyond the BEP or rated point is commonly accepted. b. Performance Curves for a Series of Geometrically Similar Pumps. The same line of reasoning as described in the foregoing is used to predict the approximate performance of a new pump which is of a different sise but geometrically similar to one that has been tested. I n this case the advantage of plotting the capacity-head values.in ratio form is that such a dimensionless leadcapacity curve then represents approximately the performance of a whole series of geometrically similar pumps, and only the values corresponding to the design point need be calciiated from the similarity relationships. If &, H, N , and D are used as the test values at the designed capacity and speed of the original pump, then the corresponding values for any other similar pump of diameter Dl and speed NI may be determined as shown in the following equations: Q1 = Q
Nl Dl
(z)
and
& =H
(&)'
(26)
It sometimes is useful to plot the characteristics for both the original and the new pumps to the same scale. For example, a design might be laid down for a series of geometrically similar p,umps with a 10 percent range of diameters. Then, since the capacity usually can be varied from about 0.75 to 1.2 times the rated values with not over a 5 percent change in efficiency, this series of pumps could cover a fairly wide useful range of headcapacity performance without any .change in the,pbxnp speed. This is illustrated by the zone outlined with cross hatching in Fig. 13. Referring to the foregoing equations, it is seen that construction of the head-capacity curve for a geometrically similar pump of different size is obtained by varying the head as the square of the diameter and the capacity as the cube of the diameter, if the speedis held constant. This same figure also shows how a new useful range zone can be constructed from the original one by increasing the speed and using a different range of diameters. I n this way it is possible to constmct a series of slightly overlapping zones which would serve as a chart for selecting the required size and speed of pump for any desired combination of capacity and head. c. Factors Affecting Pump Efficiency. The d c i e n c y of various sises and types of pumps may, be plotted as a function of the specific speed N,, with the results shown in Fig. 14. The specific speeds were calculated on the basis of the installation conditions of service and are not necessarily the BEP for the actual pumps. Owing to such considerations,aa reliability, maintainability, first cost, space requirements, or choice of stock pump sires, pumps as selected are generally less efficient than those designed for optimum efficiency. Thus values of maximum efficiency are frequently found in practice that are
*
~
MARINE ENGINEERING LEGEND
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS I
I
I
I
I
I
41 3
I
A -RADIAL-FLOW CIRCULATING PUMPS 8- AXIAL 8 MIXED-FLOW CIRCULATING PUMPS C FEED PUMPS D- MAIN 8 AUXILIARY CONDENSATE PUMPS E CARGO W M P S
-
EFFECT OF VISCOSITY
-
I MAX. EFF. PUMPING WATER AT ROOM TEMP. 2- MAX. EFF. SAME PUMP WlTH OIL AT 6 0 0 SSU 3 MAX. EFF. SAME PUMP WlTH OIL AT 2 0 0 0 SSU
-
e 0.2
5 1x
I
a
u
0
0.2
0.4 RATIO=
0.6 0.8 1.0 ACTUAL FLOW RATED FLOW
1.2
1.4
Fig. IS Hfect of viscosity on single-stage, double-suction pump characteristics
SPECIFIC S@EED,N,
=
RPM fin
I
CEN
Fig. 14
MIXFLO
n'"
PROPELLER
Repmsentahve pump diciencies Venus specitlc speed for actual installations
five to ten points higher than those of Fig. 14, for pumps of low and medium capacities. The zones on Fig. 14 indicate roughly the regions to which the particular pump application is confined by experience and practice. The portion of zone A beyond 6000 specific speed for radial-qow circulating pumps represents a comparatively small number of installations where the puinps were selected for a capacity considerably in excess of the maximum efficiency point. Conversely, the range of zone B below about 6000 for mixedflow and axial-flow pumps represents a few cases where these types were selected for a capacity below the maximum efficiency point. In general the value of 6000 specific speed at maximum efficiency represents the dividing l i e between radial and mixed-flow pumps. Similarly, pumps above specific speeds of 7500 are of the axial-flow propeller type. The c w e s reveal that the efficiency rises to a maximum and then gradually drops as the specific speed is further increased. Also, up to a certain limit, higher
efficiencies are obtained with greater volume flows. Both of these effects are mainly due to the iduence of the frictional losses in the pump. . If the specific speed is maintained constant, then, as the capacity increases, the head loss due t b surface friction becomes a smaller percentage of the actual net head while the shock losses, clearance losses, and rotation losses remain a relatively constant percentage. In the case where the volume flow is maintained constant and the specific speed is varied, the percentage shock loss remains relatively constant while the percentages for clearance and windage losses decrease as the specific speed is raised. The percentage friction loss decreases rapidly at first, reaches a minimum, and then &dually becomes greater as the specific speed is further increased. When using the curves of Fig. 14 for multi-stage pumps, N,, must be computed for a single stage. When considering double-auction pumps, the practice of using one half of total capacity is not consistent and care must be taken to compare vyious pumps on the same basis. d. Viscous Liquids. The effect of viscosity on the characteristics of a singlestage doublbsuction pump is shown in Fig. 15. The head-capacity and efficiency versus capacity curves are plotted as ratios referred to the design conditions for pumping water. Two important effects of increased viscosity are seen. First, the head-capacity curve at a constant speed is lowered so that the speed would have to be increased for the pump to operate at the rated head and capacity. The efficiency also is greatly reduced, although the
I I
I ,
SUCTION LIFT
I "0
0.2
Fig. 16
u
I
I
0.4
I
I
I
I 1.2
I
1
I
I
+ DISCHARGE STATIC HEAD
I
0.6 0.0 1 .O FLOW RATIO = FLOW/:RATED FLOW
I C
I
i
1.4
Condensate pump characteristic curves showing effect of variable suction
epecific speed for a given capacity is higher because of Vl, (c) the impeller inlet peripheral velhity ul, and (d) the reduced head. Most importantly, the horsepower re- the inlet shock angle. For complete breakdown at zero quired is significantly increased, either overloading an shock angle (i.e., at about the designed capacity) the ,--relation is found to be existing driver or requiring a new larger driver. The values corresponding to the peak efficiency for V12 0.085 ulS each curve also have been plotted against specific speed Haw= 1.485 (27) 2g 2s on Fia. 14, where they follow a curve which is not greatly steeper than the effiEiency curves for constant ~ i ~ a c i t - Equation ~ (27) expresses the suction condition limits at low specific speeds. This resemblance supports the strictly as a function of impeller inlet design. For more foregoing conclusion that the reduced pump efficiencies general considerations it is found that the requirements associated with low specific speeds are due primarily to for similarity of impeller inlet conditions can be expressed greater frictional losses. Further information on viscous- by a dimensionless grouping faentical in form with the liquid performance may be found in references [9] and previously described 'specific speed. This parameter is [lo]. known as the suction specific speed S, and is obtained by e. Suction Lift and Cavitation. The maximum abso- replacing the total head H in the usual specific speed lute fluid velocity in the suction part of the system usu- formula by the net positive suction head H,. [ll]. Thus, ally occurs at the impeller inlet, hence this is the zone of in the customary units of rpm and gpm minimum absolute pressure. If at any point in this zone the vapor pressure of the fluid is reached, then a portion of the fluid will evaporate and form vapor pockets in the stream. These cavities disturb the flow stream and then I n a radial-flow pump where the suction and discharge collapse as they are carried into regions of higher pressure, thus producing noise, vibration, and rapid erosion zones of the impeller are clearly separated, the suction of the surrounding metal surfaces. This general behavior specific speed by itself is usually suf5cient to define the is known as cavitation and the necessity to avoid it im- cavitation limits independently of the discharge flow poses definite restrictions on the design and application conditions. However, when the inlet diameter apof centrifugal pumps. C~ndensatepumps are sometimes proaches the discharge diameter, the discharge flow condesigned to operate in the cavitation range as a simple ditions also have an influence on the cavitation condimeans of self regulation in spite of the many disad- tions, so that it becomes necessary to consider both S and vantages already mentioned. The effect of cavitation on the usual specific speed NEW.When considering doublethe pump characteristic curve is shown in Fig. 16. The suction pumps, care myst be taken to use one half of the beginning of cavitation is indicated at point B for ten- total flow rate. The curve's of Fig. 17-"~epr6sent %he upper limits of inch submergence and at point D for eightrinch subfor double-auction pumps handling mergence. Points C and E on the steeply falliig curves specific speed, NEW, clear water at 85 F at sea level, as published by the Hyare in the region of complete breakdown. Conditions for incipient cavitation and also for com- dradic Institute [9]. For other conditions of higher plete breakdown are expressed by correlating: (a) the temperature, for liquids other than water, and for singlenet positive suction head H , , defined as the total suction suction, mixed-flow, and d - f l o w pumps, reference [9] head in feet of liquid, absolute, less the vapor pressure of contains additional data. When considering the cavitathe liquid in feet, (b) the impeller absolute inlet velocity tion pe$ormance of a pump in relation to its application,
+
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
I
SPECIFIC SPEED. Nw
Ill1If
=
RPY M m Hv4
FOR DOUBLE SUCTION PUMPS
Fig. 17 Upper limita of speciflc speeds for double-suction pumps handling clear water at 85 F at sea level
the user must be careful to distinguish between the net positive suction head available (abbreviated NPSHA) and the positive suction head required by ,the pump (abbreviated. NPSHR). For satisfactory performance, NPSHA must exceed NPSHR by a certain margin which depends on the particular application. An appreciation of the extent of analysis and testing that is required in pursuing the solution to cavitation problems may be obtained from references [12-141. '
1.3 Pump Characteristics a. D e s i g n F e a t u r e s . The physical components of a
centrifugal pump consist of one or mor&+pellers, casing, shaft, bearings, stuffing boxes, coupling, minor components associated with the rotating or stationary parts, instrumentation, controls, and supporting auxiliaries. Energy is imparted to the pumped fluid by the impeller, which may be of any of the types described in Section 1.1. All other components play a supporting role in the makeup of a particular pump design. The casing guides the fluid from the suction pipe to the eye of the impeller, collects the fluid exiting from the impeller outlet, guides the fluid to the discharge pipe, and acts as the major pressure boundary in the manner of a pressure vessel. Casing designs, classified as to type of flow, are described in Section 1.1. In addition, casings may be classified as to the type of collector, i.e., volute casings which have the form of a volute or spiral in the direction of flow from the impeller, circular casings which have a constant cross section concentric with the impeller, or diffuser casings which have a multiplicity of vanes to guide the fluid exiting from the impeller and convert the developed velocity head into pressure. .Shafts of centrifugal pumps are designed to transmit the torque needed fox &king the impeller,.to resist the
bending loads that occur due to internal hydraulic forces, external alignment forces, and weight of parts, and to control critical shaft vibratory conditions. Practically all pumps operating at 3500 rpm or less are designed with a first critical speed above the running speed. For speeds above 3500 rpm, such as for boiler feed pumps, the operating speed may be above the first critical speed but suitr ably below the ~econdcrikical speed. Bearings are of the antifriction type (ball or roller bearings) or of the sleeve type. (hydrodynamic type). Ball bearings, such as depicted in Fig. 2, find frequent application. Pumps running above 3500 rpm usually use journal bearings of the sleeve type and Kingsbury thrust bearings, as depicted in Figs. 4 and 5. Figure 3 shows a combination of types, that is, a ball bearing for the thrust and upper journal bearing, and a sleeve bearing at the lower or internal journal. In the latter case, the ball bearing is grease lubricated and the internal sleeve bearing is water lubricated. The ball bearings shown in Figs. 2, 3, and 8 are lubricated by a self-contained system; that is, fittings are provided for injecting fresh lubricant and draining used or excess lubricant. Seals are provided to retain the grease or oil within the housing. Sometimes cooling water is circulated in a surrounding cored pas-sage, as in Fig. 2. The bearings for the pumps in Figs. 4 and 5 are lubricated by a separate pressurized system. The pump in Fig. 7 contains an internal water-lubricated journal bearing immediately above the impeller. However, it has no thrust bearing, the axial hydraulic t h r u ~ t and rotor weight being carried by the thrust bearing in the driver. Water-lubricated bearings are satisfactory only for clean-water service. Packed stuffing boxes prevail in marine centrifugal pumps, although mechanical seals and packless boxes are finding wider usage. The s t a n g boxes for the pump in
'
i
I
Fig. 2 are subjected to positive suction pressure and therefore no special provision is needed to ensure a steady trickle of leakage through the packing rings. When the atuffing box could be subject to lift or vacuum conditions, such as with the pump in Fig. 3, provision must be mqde to inject sealing water under positive pressure at a point ~pproximately midway between several rings. Some sealing water \vill therefore enter the pump, and a small portion nil1 trickle from the box. The packed stuffing boxes in Fig. 5 require water cooling jackets due to the high temperature of the boiler feedwater being pumped (200-350 F), and due to the greater friction heat generated by the higher speeds. The pump in Fig. 4 contains packless boxes. Condensate is injected into closerunning-clearance serrated bushings at both ends of the pump. A small portion of the injection water enters the pump, but the major part leaks out the ends, collects in the drain chambers, and passes out the large drain connections. The pump depicted in Fig. 8 contains a mechanical seal. Many commerical pump designs can be equipped with either conventional packing, mechanical seals, or packless boxes. Couplings connect the pump to its driver. Commonly used types are pin and buffer (Figs. 2 and 3), gear (Fig. 4), rigid or solid (Fig. 7), and hub and spider (Fig. 8). Couplings are flexible to accommodate small misalignments between the pump and driver, except for the single-bearing pump of Fig. 7 which requires a solid (nonflexible) connection. Couplings for small pump shafts and drivers usually have a straight pressed or shrink fit (Fig. 8). Medium size and larger pumps usually have the coupling hubs mounted on a tapered fit (Figs. 2 through 7) for ease of removal. Marine centrifugal pumps are almost universally fitted with packing sleeves (Figs. 2 through 4, and 8) and journal sleeves in water-lubricated bearings (Fig. 3). A small flinger devise is usually fitted on the shaft immediately outside the stuffing box for throwing off water migrating along the shdft (Figs. 3, 4, 5, and 8). Casing wearing rings are universally installed in all types, whereas impeller rings are generally furnished only on medium and larger size pumps up to 3500 rpm (Figs. 2 and 3). High-speed boiler feed pumps (Figs. 4 and 5) generally do not have impeller rings, due to the relatively higher stresses in the impeller, and the difficulty of maintaining a reliable fastening 'at the joint. Centrifugal pumps are usually fitted for a minimum of instrumetation consisting of discharge and suction pressure gages and a lubricant supply indicator. Additional instrumentation may consist of a speed indicator, vibration indicator, pressure gages for seal water, intermediate-stage pressures, leakoffs, and lubrication systems, thermometers or thermocouples for indicating temperatures of the pump fluid, lubricant, bearings, and casing, and' flowmeters to indicate pump flow, recirculation flows, or leakofl flows. Control devices are usually associated with the system (i.e., recirculation control, pressure or temperature control of injection water for seals, remote-operated suction
41 5
and discharge valves, low suction pressure trip) but may be associated with the diiver (i.e., automatic or remote s t a ~arid t stop or shutdown upon loss of lubricant). The wide range of uses and characteristics of centrifugal pumps on shipboard are described in the following paragraphs. b. C o n d e n s e r C i r c u l a t i n g P u m p s . Both the main and auxiliary condensers re@ire a largc+volume flow of circulating water at relatively low heads; therefore, the specific speeds of. pumps for this purpose are high. For the same con9tions of capacity and head, the higher the specific speed, the lower the size and weight of the pump. Thus, circulating-pump designs of the radial-flow type were developed in the direction of increased specific speed until the limit due to cavitation was reached-at about N,, = 6000 at maximum efficimcies for doublesuction pumps. At the same time, mixed- and axial-flow main circulatr ing pumps are used widely for high-speed'qhips fitted with scoop circulation. In such cases the circulating pumps are used only for very low ship speeds and for maneuvering. Since they are operated at full capy5ty for only short periods, the specific speeds are increased beyond the usual cavitation limits. These pumps have slightly lower efficiencies due to the higher specific speeds, but this is unimportant because of their infrequent oper* tion, and is fully justified by the large savings in weight and space. Motor-driven axial- and mixed-flow pumps require a larger motor than indicated by the design conditions, because the steeply rising head'yaracteristic results in a rising horsepower curve as the capacity is decreased. An error on the low side in estimating the system head requirements would result in overloading the motor, if its ratihg were very close to the pump brake horsepower at the design capacity and head. Radial-flow circulating pumps are available with a fairly flat, non-overloading horsepower characteristic, and these units can be used safely with a motor which is rated close to the pump horsepower. A typical circulating pump of the vertical double-suction type, for smaller vessels, is depicted in Fig. 2. Circulating pumps of the vertical mixed and axial-flow types for larger vessels are depicted in Figs. 6 and 7. Seawater circulating pumps for submarines are of the radial-, mixed-, or axial-flow types, but differ in construction $ that their casings must be designed to withstand the ambient pressure due to submergence, shaft seals must be suitable for the ambient pressure, and thrust bearings must be sized for possible high axial thrusts. Circulating p u q ~ ~ r e e c ~ m m ofound n l y to be constant-speed and motor-dr~venin c;mme~cial vessels, although multi-speed motors are sometimes used. Steam turbine drives are often used on naval vessels due to the flexibility of the variabIe-speed driver and the reliability of steam as a power source. Additional details regarding circulating applications may be found in ~eferende r151. . c. C o n d e n s a t e and rain P u m p s . Continuing deA
41 6
MARINE ENGINEERING
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
where p, = atmospheric pressure, psia pop= vapor pressure, psia p = specific weight of liquid, pcf ha = total suction head, ft
SINGLE SUCTION PUMP CAPACITY, GPM
Fig. 18
Capacityspeed limitations for condensate pumps wilh h a f h *rough eye of impeller
velopments in marine power plants have increased the condensate pump head requirements. A typical condensate system consists of an air ejector and gland vapor condenser and a deaerating-type feedwater heater having a shell pressure of 10 to 60 psig. Such installations, in general, require a two-stage condensate pump. The condensate pump (sometimes referred to as the condenser or hotwell pump) is required to develop a total head made up of the sum of: the friction loss in the system due to heat exchangers, piping, fittings, and valves; the difference in elevation head between liquid levels in the hotwell and receiver (usually a deaerating-type feedwater heater); and the difference in pressure levels between the hotwell (usually being under vacuum) and the receiver which is usually at a shell pressure of 10 to 60 psig. Because of the moderate to small capacity of these units, the impellers are generally of the sin&e+uction type. In vertical pumps the submergence at the top inlet eye becomes too low if a double-suction construction is used. With two-stage pumps both suctions are placed on the inner side of the impellers, that is, the impellers are positioned "eye to eye," so that the shaft passes through the casing at the upper part and the single stufFing box is usually under discharge pressure. A twostage vertical condensate pump is depicted in Fig. 3. The specific speed is inherently low because of the low capacity-head ratio, and is limited by the available motor speed, which is 3500 rpm for a-c motors. However, only condensate and drain pumps of relatively low capacity (generally below 100 gprn), such as used with turbogenerator condensers and distilling plants, can operate
at this speed due to the low values of net positive suction head available. The demand for condensate pumps to operate at reduced submergence requires lower speeds, so that the majority of the condensate pumps operate at 1750 rpm or less. Consequently, quite low efficiencies are to be expected with condensate pumps; this is shown by zone D of Fig. 14. Limitations of capacity and speed for condensate pumps are given by Fig. 18 for single-suction pumps; for double-auction pumps, the capacities read from Fig. 18 should be doubled [9]. For the small capacity of 100 gpm for the 3500 rpm pump mentioned earlier, the suggested NPSH is four feet. Computing the suction specific speed, as defined previously in this section, S is found to equal 12,360. For a condensate pump of 330 gpm operating with an NPSH of 2 feet, the suggested speed is 1150 rpm and the computed S is 12,400. In general, condensate pumps perform with a range of suction specific speeds of 12,000 to 18,000, whereas pumps for other applications p e r f o p with suction specific speeds below 12,000. In applying condensate pumps, a distinction must be made between the terms submergence, NPSH, and suction head. Submergence relates the liquid level in the hotwell to the elevation setting of the pump. It is a static dimension, and it is customary to measure it to the centerline of the suction nozzle of vertical pumps, and to the shaft centerline of horizontal pumps. Suction head refers to pressure above atmospheric. NPSH is the total suction pressure above the vapor pressure of the liquid on the absolute scale [9]. Thus, denoting NPSH by Ha,, -
When the pump takes suction from a hotwell where the prevailing pressure equals the vapor pressure corresponding to its temperature, the NPSH is the difference in elevation between the liquid level and the datum (suction nozzle), minus the entrance and friction losses in the suction piping. Actually, condensate pumps operate a large part of the time at values of NPSH below the design value. This is because condensate pumps are generally driven by constantrspeed motors and there is no external response to a change in the amount of steam condensed. Turbinedriven units are controlled only by a speed-limiting governor so that their behavior is similar. The results of a reduction in the amount of steam condensed is shown in Fig. 16. Point A is the maximum-load operating condition corresponding approximately to the pump rating. This point is at the intersection of the pump headcapacity m e with the system m e . Assume there is a sudden reduction in condensation rate to the value E. At first the pump continues to deliver at capacity A, thus reducing the hotwell level and submergence. When the submergence has lowered to the value corresponding to the cavitation curve BC, the pump will be delivering at a capacity corresponding to point C, and finally equilibrium will be reached at the submergence and capacity corresponding to point E. A condensate pump operating on the vertical head curves, such as BC or DE, is said to be "operating in the cavitation break." Thus the part-load operating conditions for condensate pumps are severe from the standpoint of cavitation, and special consideration is given to this in the impeller and overall design. Aside from the possibility of rapid wear, these conditions are also difficult from an operating standpoint. If the cavitation part of the pump curve is very steep, operation is likely to be unstable with large fluctuations of the pump discharge pressure. There is also the possibility that the pump will become completely vapor bound and deliver much less then the required capacity until the submergence has risen considerably above the normal value, thus resulting in intermittent or slug flow in the condensate piping. An alternative method of operation employs an automatic level control. This arrangement causes part of the pump discharge flow to be bypassed back to the hotwell, thereby maintaining a prescribed level of condensate in the,hotwell and ensuring an adequate NPSH at the pump. Pump operating conditions sometimes are aggravated by insistence upon excessive margins in the specified head and capacity of the pumps. Better operation and longer life will result if the capacity and pressure loss require-
417
ments are carefully selecked and the design maxgins are held to moderate values. 'In this connection it should be noted that great care is necessary in the installation to prevent air leaks, as tests show conclusively that very small leaks will cut down the capacity by a marked d e pee. For the same reaaon it is quite important to avoid forming air pockets in the syction line. A vent connection located at or near the suction nozzle and connected to the condenser assists in cleaxing vapor' from the im,.' peller eye. d. Boiler Peed Pumps. The etliciency characteristics of boiler feed pumps are shown in Fig. 14. It is seen that the efficiency zone is slightly higher than for condensate pumps because of the somewhat larger capacity required. For the same capacity and specific speed, aciencies of the two are about the same despite the larger number of stages in feed pumps. The efficiency is about the same for volute and diffuser-type pu,mps, the choice of type being principally a matter of the pump manufacturer's practice or the customer's preference. Advances in design have increased pump speeds and consequently the maximum head per stag6 so that twostage and single4tage feed pumps are available for boiler pressures up to 1200 psig. Centrifugal boiler feed pumps are not widely used for capacities below about 100 gpm because the combination of low capacity and high relative velocity (due to the high head per stage) results in impeller passages which are quite small and therefore are more likely to become clogged or restricted by smqll-particles or boiler water deposits. Figure 4 depicts a horizontal two-stage boiler feed pump and Fig. 5 depicts a horizontal four-stage pump. Single-suction impellers of boiler feed pumps develop a large axial thrust due to the pressure diffeience on the two sides of the impeller unless both back and front wearing rings are fitted. For reasons of reliability and maintainability, however, boiler feed pumps do not have back wearing rings, nor do they have impeller rings at the eye. Where an even number of impellers is used, these axial forces may be balanced by arraaging the impellers to face each other. For more than two stages, this arrangement requires some of the interstage passages to cross over other stages, thus complicating the casing design considerably. A balance drum with labyrinthdthrottling paasages is often used to compensate for the axial thrust of several impellers, but this arrangement will not always balance the thrust under all operating conditions; therefore, an external thrust bearing is also required. One means of automatically balancing the axial thrust unde?5al~~condit,ns of_-operationis the balance disk, which varies the clearance in the throttling element and hence the balancing pressure by permitting a slight axial movement of the pump shaft. Operation of boiler feed pumps under cavitating conditions is never permissible because of the risk that the pump suction might become vapor bound. If this happens, all the water in the pump may be evaporated quickly owing to the large energy losses in the pump
*
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
relative to the volume of water it contains, and serious damage to the running parts probably will result. For this reason most feed pumps are provided with low-suction pressure trips. For the same reason boiler feed pumps cannot be operated safely at conditions approaching shut-off and are provided with a recirculating line which discharges from 5 to 15 percent of the rated capacity back to the deaerating feed tank. The chances of complete vapor binding at rates approaching shut-off are greater than in condensate pumps, not only because of the larger energy losses in the pump, but also because a t higher water temperatures a small rise in temperature results in a greater increase of the vapor pressure. The limiting suction conditions must depend, therefore, not only upon the specific speed and suction specific speed, but also upon the temperature of the water, owing to the greater rate of vapor pressure rise. Figure 19, as presented in reference [9], depicts the recommended minimum NPSH for pumps handling water at temperatures of 212 F and above. Since, for boiler feed pumps, the specific gravity is usually less than 1.0, care must be taken to include this factor when converting actual suction pressures in psi to values of NPSH, and when converting suction and discharge pressures to total head. A boiler feed pump is required to supply water a t a pressure sufficient to overcome the sum of the highest safety. valve setting of the boiler, the pressure due to the static elevation of the boiler above the pump, and the friction head loss in the system of piping, valves, and fittings. It is usual practice to plot the system head curve and superimpose on it the head-capacity curve of the pump as depicted in Fig. 20(a). A turbinedriven boiler feed pump may employ any of the following three types of governors: 1. Constant-speed governor only. 2. Constant-pressure governor to maintain a constant discharge pressure at all capacities. 3. Differential-pressure governor to maintain a specified differential-pressure across the feedwater regulator. These governors are described further by the simplified diagrams of Fig. 20. In the constant-speed governor system, Fig. 20(b), the feedwater regulator throttles all the difference between pump discharge pressure and the required pressure. The throttled pressure differential represents a large waste of power, hence a constant-speed governor is rarely used alone; it is usually provided along with a constant-pressure or differential-pressure governor, in which caqe the speed governor provides a secondary means of control when the pressure governor may be temporarily out of service. In the constant-pressure governor system, Fig. 20fc), the governor controls the turbine and pump speed to provide a specified constant pressure at the pump discharge. The feedwater regulator throttles the variable difference between the constant pressure and the system head curve. In the differential-pressuregovernor system,-Fig. 20(d), the
SYSTEM MAD CURVE WlTH PEEDWATER REGULATOR FULLY
H-o CHARACTERISTIC OF PUMP
500 400
P OPEN
OESIQN -HEAD
300 200
-----
~ O I L E RPRESSURE
I
L C
a o
+.I00
W
r 90
I
I OESIGN
80
0
g
* 5 In
PLUS STATIC ELCVATION
70 60 50
CAMCITY
CAPACITY
(a1 TYPICAL BOILER FEED SYSTEM
40 30
W
2 k
20
SPEED
V)
B
k
I0 I
CAPACITY, GPM Fig. 19
Net positive suction head for single-suction, centrifugal, hot water Pump'
governor oontrols the turbine and pump speed to provide a specified differential pressure across the feedwater regulator. The pump discharge pressure coincides with the system head curve. , The diagram in Fig. 20(e) shows the system head curve for the boiler feed system of a nuclear steam plant. In this case the pump is called upon to operate at its highest speed at the low-flow condition. The operation of the pump is stable, however, since, even though there may be an inflection in its speed curve, there is no inflection in its horsepower curve throughout the range of needed capaoity. When applying boiler feed pumps, the method and amount of recirculation must be coordinated with the method of pressure governing. A widely accepted practice [5] is to hold the temperature rise at low flows to 15 deg F, though rises of 20 or 25 deg I? may be found in use. A boiler feed pump operating under the control of a differential-pressure governor system requires a certain amount of horsepower at the minimum-flow condition, aa indicated in Fig. 20(d). If for some reason the control system is made inoperative and the pump then operates at constant speed, the horsepower to be dissipated is a larger amount, as shown in Fig. 20(b). Thus the reek culation flow must be sized for the \\?orst condition that the system may encounter. In addition, consideration must be given to whether the recirculation flow is permitted only a t or near the minimum-flow condition, or whether it is continuous. The various types of recircualtion devices and systema may be summarized as follows. All systems require a pressure breakdown device which may be a singb plate orXce, a multiple orifice (spool type), or a small-bore tube (friction tubing). The pressure breakdown device
SYSTEM HEAD CURVE wlm FEEDWER RESULATQR
/:
1
---ELEVATION
I'
URE
-*---
BRAKE HORSEPOWER
IMSIGN
CAPACITY
CAPACITY
tb) CONSTANT-SPEED GOVERNOR SYSTEM
H-Q GHARACTERISTIC OF PUMP-CWSTAMT SPEED
WlTH F E E D W E R
is installed in a line connected from the pump discharge back to the deaerating feed tank. For continuous recirculation, no other devices are fitted, except stop valves for isolating the line. For automatic intermittent recirculation, a control device is required. This may be a diaphragm-operated valve, controlled by an air signal from a flow transmitter, which in turn is connected to a flow-measuring orifice in thwpump discharge line. Another type is a special discharge check valve in which the movement of the-internal check operates a pilot valve that opens or $lcises the recirculation line. Most marine boiler feed pumps are driven by steam turbines. Whereas Figs. 4 and 5 depict coupled designs, single- and two-stage types that have pump shafts common with the turbine shaft are frequently found. Smallcapacity pumps for waste-heat boilers or small auxiliary boilers are usually of the end-suction type, Fig. 8, and are motor driven. Some ir~stallationsutilize vertical multi-stage pumps, either motor or turtine driven, and some vessels have the feed pump driven by the main propulsion turbines or by the turbogenerator set. e. Feed Booster Pumps. Feed boosterr-pumps - are needed in connection with deaerating heater type feed systems where deaerating heater space at arestrictions sufficient static prevent elevation locating above the the boiler feed pump suction to provide the required net positive suction head. In the absence of a large static submergence a t the suction, the required cavitation limits are satisfied by the booster pump discharge pressure. With this arrangement a relhxivelv low NPSH condition occurs at thevfeed booster However, these pumps usually are designed for much lower speeds and heads so that the impeller is larger and the suction velocities are lower than in boiler feed pumps. The result is that, for about the same suction specific speed, 8,
PLUS STATIC
I
-
I
CAPACITY
le 1 CONSTANT-PRESSURE GOVERNOR SYSTEM
H -Q CURVES AT SPEEDS ABOVE 100%
H-Q C M M E R I S T I C OF PUMP-160% s m
H-Q CHARACTERISTLC 0 W
REWLAlOR ,
r
V A R I a I L E PRESSU# VARIABLE SPEED 5 STATIC
t
I
I CAPACITY
(dl DIFFERENTIAL-PRESSURE GOVERNOR SYSTEM Fig. 20
CAPACITY
(el NUCLEAR STEAM POWER PLANT
Boiler feed pump and system characteristic curves
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
a lower net positive suction head is permitted for the booster pump. Also, the possibility of complete vaporbation is reduced in the booster pump because the ratio of energy loss to pump volume is so much smaller than for boiler,feed pumps. Feed booster pumps are in many respects similar to condensate pumps. Booster pumps do not operate in the cavitation break, a s mentioned previously for condensate pumps, 'as a positive stable performance is required in order to provide the NPSHR for the boiler feed pump. The capacity of the booster pump is essentially controlled by the boiler feed pump, which in turn is controlled by the boiler feedwater regulator. A booster pump must be provided with a recirculating line, usually piped back to the deaerating feed tank, to protect the booster pump when the feed pump is not in operation. A typical twostage booster pump is similar to the condensate pump depicted in Fig. 3. A singlestage booster pump would be similar to that of Fig. 3 except that there would be no second (upper) stage. Booster pumps frequently use only ball bearings, due to the severe duty imposed upon them by tbe higher water temperatures and danger of flashing w h i ~ bmake water-lubricated bearings less suitr able. f. Cargo Pumps. Representative efficiencies of cargo pumps are shown in Fig. 14. Axially split casing types of pumps, with double-suction impellers, similar to that depicted in Fig. 2, are used in the capacity range of 2000 to 20,000 gpm at total heads of 500 ft. This type of pump can be horisontal or vertical, being driven by a geared steam turbine in the main machinery space, with the driving shaft extending through a bulkhead or deck stuffing box to the pump located in an adjacent pump room; or the pump may be of the deepwell type. The geared turbine drive permits the use of a highly efficient turbine far the large power requirements and permits compatible ,matching of the most desirable pump and driver speeds. It provides variable-speed operation in order to obtain high speeds for producing the highest heads or lower speeds to suit lower values of NPSHA, thus enabling the pump to respond to the wide range of system head characteristics that are encountered under different operating conditions. The horiaontal-shaft arrangement has the disadvantage that misalignment may ensue under different conditions of vessel loading. To overcome this, and to reduce fore-and-aft space requirements, vertical doublesuction pumps are used in many vessels. Here the pumps are driven by a turbine and reduction gear located on a machinery flat extending from the main machinery space, directly over the pump room. Vertical shafting is used to connect the pumps and drivers. Cargo pumps are required to pump a variety of volatile liquids and thus require means to ensure continuous pumping under possible vapor-binding conditions. Most older vessels used reciprocating steam pumps, which were operated during the stripping operation. This type of pump, being positive displacement,was capable of pumping vapor6 as well as liquids. It wm operated coin-
1
I
1
i LEGEND 1 2 3 4 6
8 7 8 B 10 11 12 13 14 15 18
SUCTION BELL FIR-AGE IMPELLER (AXIAL.FLOWTYPEI FIRSTSTAGE DISCHARGE BOWL EWALlZER LINE SPACER PIECE SECONDSTAGE SUCTION BELL SECOND STAGE IMPELLER (MIXED-FLOWTYPE) SECONDSTAGE DISCHARGE BOWL THIRD,STAGE SUCTION BELL THIRD STAGE IMPELLER (MIXED-FLOWTYPE) THIRDSTAGE DISCHARGE BOWL VENTLINE SUPPORT COLUMN WAFT PACKING BOX DISCHARGE HEAD 81DRIVER SUPPORT
Fig. 21
Deepwell-type centrifugal pump
cidently with the centrifugal cargo pump, where it effectively removed the vapor collection at the suction of the centrifugal pump. A variety of systems for removing vapor is found in modern vessels. One such type employs an integral priming or scavenging impeller located at both ends of the regular impeller. The vapor discharged from the priming impellers is piped to a separator tank, from which any liquid present may drain back to the pump suction, and from which vapor is piped to the overboard vent system. Another type makes use of a separate vacuum pump which pumps vapor from a separator tank located in the main pump suction line (see
,
Fig. 21 of Chapter 18). The vapor passes through an interceptor tank, where any liquid carryover is removed and drains back to the main pump suction. The discharge of the vacuum pump is piped to the overboard vent system. Associated instrumentation and con.trols cause the vacuum pump to start and stop as the amount of vapor accumulates in the separator tank. Another type, which is used with the deepwell type of pump, incorporates an automatic priming valve. In addition to steam turbine drives, cargo pumps may be driven by electric or hydraulic motors, by diesel engines, or by gas turbines. A common arrangement is a vertical deepwell-type pump driven by a vertical explosion-proof votor, or by a horizontal diesel or steam turbine driviqg through a rightrangle gear. Such a design, depicted in Fig. 21, is of the single-suction type and is multi-stage. It may contain a specially designed firstr stage impeller to suit the particular suction conditions. For the handling of low-temperature liquefied gases, cargo pumps are almost without exception of the submerged deepwell type, or of the completely submerged canned-motor type. The deepwell type is mounted vertically, with its driving motor located on the deck above. It is also used for chemical cargoes. The canned-motor type is mounted vertically with the motor a t the bottom of the cargo tanks. The motor windings and bearings are completely submerged in the pump fluid. Both types usually contain a special design of suction impeller compatible with the low values of NPSHA. Additional discussi~nregarding cargo pumps is given in references [lo], [16], [17], and [18]. g. Bilge and Ballast Pumps. The general requirements of bilge pumps are similar to those of condensate pumps in that, due to the liftrequired of bilge pumps, the suction pressure is considerably below atmospheric. In comparison, the capacity is larger and the discharge pressure is less. The impeller passages must be much larger in order to pass dirt and bilge debris. As no static submergence is available to fill the pump with water, an auxiliary priming devide is required. This usually ia a positive-displacement air pump either directly attached to the main pump shaft or operated separately. An air float valve is used to isolate the air pump after priming has been accomplished, thus avoiding the chqrning loss of this element during normal operation. Some vessels employ an independent central priming pump for all bilge and ballast services. Bilge and ballast pumps are usually of the axially split type, either horizontal or vertical, similar to the pump depicted in Fig. 2. h. General Service. Centrifugal pumps are used for many minor and supporting services for capacities up to 5000 gpm and for total heads up to about 350 ft. Typical applications are for fire main, flushing, cargo oil tank cleaning, refrigeration condenser cooling, distilling plant feed, condensate, brine overboard, fresh water, and sanitary systems. Pump construction for these services is
42 1
usually that of the axially split casing type, either horizontal or vertical, similar to that depicted in Fig. 2, or of the end-suction type as depicted in Fig. 8. i. Primary Cooling Service. Pressurized water reactor systems require a circulating pump that is of the endsuction type with the impeller mounted directly on the motor shaft. The bearings are lubricated by the pump fluid, and the motor'windkgs are isolated in a separate casing. j. Jet Propulsion. Pumps are finding increased usage as pro dlsion devices, either as the main propulsion device or asPan auxiliary device such as a bow thruster, These pumps are usually of the mixed-flow or axial-flow type. The particular design depends on the exact matching of the pump capacity, total head, speed, and type of driver, which may be an electric motor, internal combustion engine, or gas turbine. Additional discussion of jet propulsion applications of pumps is contained in refer, ences [I91 to [21]. \ k. Materials. Centrifugal pumps for merchant vessels usually are made entirely of nonferrous materials when used for seawater applications. Cashgs and impellers are generally of cast and machined bronze. Shafts and sleeves are generally of nickel-copper alloys, suitably hardened. Sleeves and wearing rings are of bronze or nickel-copper alloys, depending on the severity of service conditions. Occasionally impellers and the internal surfaces of casings are coated with one of several commercial coatings available. Seawater casings subject to submergence pressure aboard submarines are cast of highstrength alloy bronze or of a.8opper-nickel alloy. For freshwater services, cast iron is occasionally used for casings, but bronze is the most common casing material for condensate and other low-pressure applications. Various grades of stainless steel are used for high-pressure boiler feed pump casings. Impellers for low-pressure services are of bronze or nickel-copper alloy and are of stainless steel for boiler feed service. Shafts for lowpressure services are of carbon steel or stainless steel and are of stainless steel for boiler feed service. Sleeves and wearing rings are of bronze, nickel-copper alloys, or stainless steel depending on the severity of the service. Materials for cargo oil pumps must be suitable for seawater, as they are often used for ballasting service. Pumps for low-temperature liquefied gases are usually of stainless steel, the particular grade depending on the low temperature encountered. Wearing rings for the caaing and impeller are always made of dissimilar grades of bronze or other alloys and are of different hardnesses to provide the best wearing properties at the close running cleyances. The same practice is followed for th%rotatiig Eidd s t a h n a r y parts of pressure breakdown labyrinths and balancing disks and drums. Bolting is usually made of bronze or a nickel-copper alloy for low-pressure services and of low-alloy steel for boiler feed pumps.
422
MARINE ENGINEERING
Sectien 9 Reciprecating Steam P~mps 2.1 Ckrssificdion and Types. The use of recipmcatr ing pumps in the marine field has diminished such that they are now primarily used only for specialized purposes. Reciprocating pumps are of the positivedisplacement type; that is, pumps which displace a constant volume of fluid from the suction to discharge port, for each stroke or revolution of the driven shaft. Reciprocating pumps are basically of two types: ( 1 ) the steam-driven directr acting type, where the steam-end and the reciprocating pump are built together as a unit with the motivating Power being provided by the steam-end; and ( 2 ) the Power Pump t.We where the reciprocating action of the Pump is provided by the rotary motion of an external prime mover and is converted to reciprocating motion by crankshafts, eccentrics, and cam plates, piston rods, etc(see Section 3 ) . Regardless of the type, the Pump end is classified the same. A reciprocating steam pump consisting of single steam and liquid cylinders and pistons with one piston rod is known as a single or simplex pump. Figure 22 shows a vertical simplex double-acting type. Where two such pumps of identical size are arranged side by side and the steam valve of one side obtains its motion from the piston rod of the other side, the unit is known as a duplex pump. Figure 23 depicts a vertical duplex double-acting type of pump. Both of the pumps shown in Figs. 22 and 23 are known as vertical pumps because the axial centerline of the cylinders is in a vertical position. Horizontal types are those having the axial centerline of the cylirlders in a horizontal position. The size of a reciprocating steam pump is described by giving first the steam cylinder diameter, then the water cylinder diameter, and finally the pump stroke, all in inches. 2.2 Flow Rate and Piston Speeds. A reciprocating steam pump produces a constant rate of flow throughout most of each stroke; however, the flow is reduced abruptly at the end of each stroke as the piston decelerates and reverses direction. Figure 24 shows, in relative terms, the maximum and mean flows for a simplex pump and duplex pump. It is desirable to fit air chambers a t the discharge to smooth out the flow pulsations to acceptable values. Basic piston speeds, as recommended by the Hydraulic Institute [9], are shorn in Fig. 25. The curve applies to simplex and duplex pumps of conventional design operating on cald water or on liquids possessing a viscosity of 250 SSU or lees. However, for handling viscous liquids, and when pumping hot water, lower piston speeds must be used; for recommended values, consult reference [Q]. Where U is the piston speed in fpm, Q is the volume rate of flow in cfs, D is the liquid piston diameter in feet, and S is the slip ratio, the piston speeds may be calculated by one of the following formulas:
For simplex pumps:
u=-
WQ
(30)
(*/4)D2(1 - S)
pumps:
U
=
'
-
*WQ
(31)
(u/4)D2(1- S )
2.3 ~ ~ l steom ~ + i pump ~ ~ cylinder sizes, The steam piston diameter must be proportioned torrectly in relation to the liquid piston diameter to obtain the desired pump pressure for a given steam pressure. Since the steam cylinder can have no cutoff, due to the negligible inertia of the moving parts, the pressure at the end of the stroke is less than the steam line pressure only because of friction and shock losses. The required ratio of steam piston area to liquid piston area can be calculated from the following expression: Da2 Dm2 &(pa
PW
- p,
--
(32) ~
1
)
where
p, = net liquid pressure, psi p, = steam inlet pressure, psia p, = steam exhaust pressure, psia pl = steam pressure loss entering and leaving steam cylinder, psi D , = steam piston diameter, f t D , = liquid piston diameter, f t Em = mechanical efficiency I n selecting cylinder diameter ratios, it is common practice to add a margin of 10 to 15 percent to the values obtained by calculation. 2.4 Mechanical and Volumetric Efficiency. Average values for the mechanical efficiency of reciprocating steam pumps are shown in Fig. 26. The volumetric efficiency is given as a loss in percentage of the displacement and is then called "slip." The slip a t rated piston speed for new pumps averages from 2 to 5 percent, but this value increases as the pumps wear in service, so that far design purposes it is better to allow for about 10 percent slip. The slip values at different rated pressures are based on pumps with packings designed for these pressures. 2.5 Steam Consumption. The approximate steam consumptio~lof a reciprocating pump at any given rate of fluid flow and total pressure is readily calculated from a consideration of the equilibrium of steam and liquid forces on their respective pistons at the ends of the discharge stroke. At this instant these forces, with due
a
1 2 3 4 6 6 7 8 0
Fig. 22
LEGEND
STEAMCYLINDER STEAM CYLINDER l STEAM ROD LIQUID CYLINDER - LIQUID CYLINDER LINER LIQUID PISTON LIQUID PISTON ROD VALVE SERVICE LIQUID PISTON P ~ K I N G
Vertical simplex pump
allowance for mechanical friction, are equal and the steam cylinder terminal pressure may be determined accordingly. The cubic content of the steam cylinder divided by the specific volume of steam in the cylinder gives the weight of steam contained in the cylinder at this inetant, and this weight of steam multiplied by the double strokes per minute of the pump gives the approximate steam consumption for each end of a double-acting cylinder.
LEGEND STEAM CYLINDER STEAM ROD LIQUID CYLINDER LIQUID CYLINDER LINER LIQUID PISTON ROD LIQUID PISTON LIQUID PISTON PACKING LIQUID VALVE SERVICE AIR CHAMBER
Fig. 23
Vertical duplex pump
The actual specific v o l p e of-the &earn a t the terminal pressure will depend upon the amount of-wndensation occurring during the stroke, and the steam leakage during the stroke also must be considered. It is customary to assume that the steam is in a saturated condition a t the terminal pressure and determine a corresponding overall condensation and leakage factor from actual steam consumption tests. On this basis the steam consumption may be obtained as follows. Using the nota-
MARINE ENGINEERING
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
PUMP STROKE Fig. 26
~ 1 24 ~ .
~
1
the following additional notation
w = pump steam consumption, lb/hr Q = volume rate of liquid flow, cfa
V, = saturated specific volume a t PI, CU fft/lb
1
The steam consumption
for a simplex double-acting pump and twice this for a duplex double-acting Pump. But
-
, '
Mechanical efficiencies and clearance ratios for reciprocating pumps
rate 0 ~ variation in reciprocating pump
tion stated earlier with equation (321, first estimate the steam cylinder terminal pressure, Pr, from
Emplo&
- INCHES
425
40
- WPLEX POWER
c-
TRIPLEX
AND
WMPS
MULTIPLEX
the factor C2 are given in curve form on Fig. 27. It will be noted that the condensation factor depends primarily (38) on the piston speed, but is also influenced by the steam cylinder diameter and the terminal pressure. The choice It is often useful to have this result in terms of piston of these three factors and the wgy in which they are used speed rather than capacity, and since is not accidental, but is based on an analysis of the heat exchange between the steam and the cylinder walls. WQ U = 2.6 Steam-End Valves and Linkages. The simplest (.~/4)D~'(l- S) (39) type of steam valve is the direct-acting D-slide valve showncan onbethe duplex pump of Fig. 23.where Thisthe type of for a simplex pump, substitution of this in the foregoing valve used only on a duplex pump valve steam consumption equation results in motion for one side is obtained from the motion of the W = 0.106 C1CZUDaZpt piston rod of the other side. For a simplex pump, a free(40) moving steam-operated main valve with a meahanically + or twice this for a duplex pump. actuated pilot valve is required since at slow speed there A further useful relation is obtained by dividing the might be only power enough to bring a mechanically Steam comum~tionby the liquid homepower. Thus the actuated main valve to dead center, thus causing the liquid horsepower (any fluid) pump to stop. Even for duplex pumps a slide valve canQP 144 not have lap or lead for the same reason. pw= A When a slide valve, either main or pilot, is used for a 550 vertical pymp, a lateral preseure is required to hold the valve against its seat; for this reason piston-type main valves are more suitable for vertical pumps. However, small-area slide pilot valves are found to be quite satis-
PW= 0.~03427Em(l - S)U(p, - p,)Da2
(41)
So the steam rate per fluid horsepower is
2(r/4)DWzL(1 S)
S.R.
=
w = 30.9 ClCZ -
p,
(42) Em(1 - 8 ) bt - pa) for either simplex or duplex pumps. Average values of the mechanical efficiency, slip ratio, and clearance ratio CI are shown in Fig. 26. Values of p w
Independent inlet and e&aust cyli&r ports are used with the direct-acting D-type slide valve-?n duplex pumps, the exhaust ports being on the inside so that as the piston approaches the end of its stroke the exhaust port is closed, and steam trapped ahead of the piston becomes compressed and serves as a cushion which prevents the piston from striking the cylinder head. For duplex steam cylinders of about 10 in. diameter and over with a direct-acting slide valve, a cushion-
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
release valve is usually fitted between the steam and exhau9t ports which is essentially an orifice the cushioning actionat the end of the stroke. The valve is located in a bypass between the inlet and exhaust ports and is manually adjusted by a handwheel on the outside. If the cushion valve is fully closed, complete cushioning is obtained; if the valve is partly opened, the compression is partly relieved by the amount steam flowing through the omhion valve. B~ manually adjusting the cushion valve a smooth stopat the end of the stroke can be obtained under all ordinary conditions of pressure and ~h~ application of a steamdperated main piston valve with a direct-acting pilot slide valve to a simplex pump is shown in ~ i 22.~ . this case the main slide valve of the fully balanced piston typeis controlled by the pilot valve and has a slide valve which rides above the
Slide valves are limited to sizes of about 25 in' pressures 150 psig' in diameter and steam of They are further to for diameters about 400 F for the larger sizes and 450 of 10 in. and kss. Above these temperatures it is difficult to lubricate the slide properly, and the likely to become warped and cause excessive leakage' Balanced piston valves are used when the size Or arc perature limits exceed those for which D-slide may bo suitable. For a duplex pump the piston must direct-operated, but for simp1ex pumps the consist, as previously explained, of a steam-operated main valve controlled by an auxiliaw of either the 'lido With a steam-operated main or piston type, starting and cushioning at the end of tho stroke are accomplished through the starting and ports. AS the main steam piston approaches the end its stroke, the main port, which is open to exhaust,
'*
In the case of a duplex pump it is evident that the slide valve must not reach dead center until the piston is near the end of its stroke and that this cannot hold for both pistons unless lost motion is provided in the valve linkage. Even with a liberal amount of lost motion some variation in the length of the stroke of a duplex pump is to be expected under different load conditions. 2.7 Liquid-End Valves. Two of the numerous types of liquid-end valves employed appear in the pump sectional illustrations. Stem-guided metal disk valves are ~houmin Figs. 22 and 23. Rubber composition disk valves may be used for moderate pressures handling cold water. For higher pressures \ring-guided valves with a conical facing are used. There are numerous other types of pump valves for various special applications. Among these may be menh n e d the spherical or ball valve and the semi-spherical valve, both desirable for handling viscous liquids because of the clear area through the seat, and the hinge or flap valve which is used for liquids carrying solid matter such HS sewage. It mill be noted that all valves are installed in 11 horizontal position and that the valves are all mounted
427
pumps are not suitable for operation with more than about 50 to 75 deg F of initial superheat, since a higher superheat prevents the formation of sufficient condensate to lubricate the moving parts. 2.10 Materials. For steam-ends, steel is used for greater strength in cylinder castings and heads when the steam pressure exceeds about 300 psig, and is used exclusively for this purpose in naval vessels because of the low shoclr resistance of cast iron. Ductile iron is also used and bas proven to be shock resistant. For pump ends handling fresh water, the cylinder castings may be of iron or steel depending on the pressure. The working parts generally are made of bronze; but piston rods, valve disks, and stems may be made of monel preferably or hewreated .atahless steel for greater strength or improved wearing quality.-'For m&umum corrosiorl resistance, one of several grades of austenitic stainless steel conforming to ASTM specific* tion A296 may be used. For seawater service, nonferrous working parts are used exclusively in the pump ends, and it is better t o have the complete pump end of nonferrous construction. Mineral oils of any character, including petroleum fuel
MARINE ENGINEERING
and lubricating oils, generally are handled by pumps of The pump-end cylinder liner and all-ferrous pistonhaterials are usually made the same as the comesteel is Bponding steam-eIld itemsbut hardened for valve seats, disks, and springs, with the valve seats of a material or hardness different from the Additional discussion of materials for reciprocating steam pumps and their design features are contained in references [9] and [161. 2.1 1 Applicatidns. At the present time Very few reciprocating steam pumps are used for continuous service. They are very useful, however, and continue to be used or emergency units for boiler feed service. as The simplicity of operation and self-priming characteristic of a reciprocating steampump make it well suited to bilge, ballat, general service, fuel oil transfer, and cargo oil stripping.
Vertical Pumps are preferred for thelarger sizes since less floor space is required for a given capacity- 4further advantage of the vertical Pump is that the weight the in less pistons is not carried by the cylinders, piston friction and wear. For the same capacity and Pressure) a simp1ex pump is generally lower in cost and more economical than the duplex type. A simplex pump is also more suitable for severe suction conditions where the pump may lose its supply of liquid temporarily. I n such m5es the simplex pump, which always operates on full stroke, may recover its suction sooner than a duplex pump which may shod stroke under these conditions. The steadier flow characteristics of a duplex pump, as compared to the simplex pump) have been mentioned' This effect may be offset to a considerable extent by the longer pause a t the end of the duplex Pump stroke-
Section 3 Power Pumps Basic Types. The term "power 3.1 Detlnitions pump" is used to describe any type of pump whose action depends on the recipmcating motion of pistons or plungers and whose motive power is from an external
The fixed-stroke power pump is driven through a crankshaft and connecting rods by a driving unit (USUally an electric motor) that is either direct connected, chained, belted, or geared to the crankshaft. Variablestroke power pumps usually employ other means of conVerting rotary motion of the applied power to the reciprocating motion of the pistons or plungers. Power pumps are further classified according to the arrangement of pistons or plungers. A single or simplex pump has one piston or plunger. A duplex pump has two pistons or plungers. In like manner a triplex pump has three, and a pump has more than three pistons or plungers. The reciprocating pistons or plungers may be single-acting or double-acting. 3.2
Fixed-Stroke Power Pumps
General Considerations and Applications. The pump end of a fixed-stroke power pump is identical in function u
|
Power pumps of the triplex construction, which is not practical in directacting steam types, have a smoother discharge flow than single and duplex pumps and were formerly used for boiler feed service or low-capacity Po* feed pumps. Power pumps offer higher mechanical efficiencies than centrifugal Pumps for small to moderate capacities with moderate to high pressures. I n general, fixed-stroke Power Pumps are provided with a discharge relief valve of ample capacity to take care of changes in the flow that are in excess of the range pre vided by the speed regulation and to-prOtect the pump from damage due to the accidental closing of the discharge line. b. Performance Char=teristics- The flow rate ations of single and duplex d0uble-acting and trip1ex single-acting power Pumps are shown in Fig. 24. Owing to the crank action, the piston speed and pump capacity sine wave characterfor a simplex pump have a will be noted that the capacity variation is istic. greatly reduced in duplex pumps and is reduced still more in triplex pumps with equal crank phase angles. Limiting piston speeds of Power Pumps are somewhat higher than for directacting Pumps shown in Fig. 25. The limitation on piston speed is, in general, due to the time required for the neceSSwuY acceleration and retardam of tion of all m0Vi11g elements including the fluid in the suction and discharge pipes rather thanfrom The consideration of stress in the working pa*pump is limited to someurhat lower piston speeds because of the necessity of cushioning or smoothly retarding the steam piston, connecting rod, valves, and linkages which represent a considerable reciprocating mass that is not present in the power Pump. Acceleration and retadation of the fluid columns are often limiting f~Ct0mand
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
I
O FLUID HORSEPOWER
Fig. 28
Mechanical efficiency of fixed-stroke power pumps
are reduced a minimum in the triplex Pump because of less flow variation. The of power Pumps is influenced to a considerable extent by the friction l0SSeS in stuffing boxes, crank and rod bearings, and by the gearing losses. Figure 28 indicates the appfoximate efficiency of singlereduction geared outside-packed power pumps as a funcof the fluid Double-reduction geared units will have a loss from 2 to 4 percent greater. c- ChnetIuction Details. Both the piston-type constmotion with piston packing rings and the outside packed plunger are used for Power pumps. The outside-pacM type has less slip and also has the practiadvantage that any packing ledage will be evident and can be remedied without disassembling the pump. A duplex piston-tme Power pump is shown in Fig. 29- In general, the remarks in Section 2 (on Steam pumps) regarding liquid-end parts and to fixed-stmke Power Pumps when used for similar services. The horizontal duplex type is used for capacities of 1to 500 gpm and mdderate pressures up to PS&. For larger sizes, vertical Pumps are preferred because of their reduced 'pace Vertical Pumps are used for capacities of l5 to 1300 gPm and higher discharge pressures up to 800 psig.
Power ends of horizontal pumps are usually of the enclosed crankcase type with splash lubrication depicted in Fig- 29- Smaller sizes may employ a rotating eccentric instead of a crankshaft. The side thrust of the connect ing rod is carried by a guided crosshead. The piston rod is supported a t one end by the crosshead and by the liquid piston a t the opposite end. The drive is of the double-reduction type, thacis, the crankshaft is driven by herringbone gears from an intermediate shaft which in turn is belt p'i. chain driven by the diving motor. Owing to the inherently low pump speed of 100 to 450 Ppm because of piston speed limitations, motor speds usually are kept ]ow in order to reduce the gearing ratio. A balance must be maintained between the increased cost of slower speed motom and the costof the reduction gearing, chain, or belt drive. F~~smaller sizes this results in motor speeds up to 1750 rpm, but in the larger units 550 to 850 rpm are more frequent, 3.3 Variable-Shoke power pumps. \ There are two fundamentally different types of variablestroke pumps. These are the radial-piston type and the axial-piston type. , a. Radial-Pmn Type. ~ ~ , - ~ i variable~ l - ~ i ~ stroke pumps are espaially suitable for the hydraulic transmission of power where the machinery to be driven is relatively large and 8lOwmoving, &]thoughtheir appfication is by no melimited to this category. ship machinery of this type includes steering gears, windlasses, cranes, and As a direct source of energy.this typeof pump is subject to the limitation that it can be used only wit. fluids which are suitable lubricants for the worbg parts. F~~ hydraulic power transmission, the fluid used generally can be selected to suit the lubrication requirements of the pump. The efficiencyof radial-piston pumps is in the range of 85 to 95 percent. The influence of capacity, speed, presSure, and degree of slip is not entirely consistent, but in
I
1 FRAME 2 BEARINGS 3 CRANKSHAFT 4 CYLINDER 5 LINER
7 VALVE SERVICE 8 PISTON PACKING
Fig. 29
429
Horizontal duplex power pump
MARINE ENGINEERING
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
rn
YOKE \
43 1
, "
i
ADJUSTMENT
OMPENSATOR
PASSAGE "A"
/
YOKE RETURN SPRING
YOKEPISTON AC~UATING
Fig. 3 1
\
PASSAGE "8"
Axial-piston-type pump wickers]
$
general, for a particular capacity, the efficiency is greater for larger slower-speed pumps. The radial-piston type of variablestroke pump is shown in Fig. 30. The rotating cylinder block contains several radial cylinders fitted with pistons. As the cylinder block rotates, the piston sleeves move in and out, and the amount of stroke depends on the eccentricity between the housing and the cylinder block. Control of the stroke may be automatic, that is, of the constantr pressure type, or the control may be of the start and stop type regulated by the rise and fall of pressure in a hydraulic accumulator. Manual stroke control i~ provided usually in addition to automatic control.
b. Axial-Piston Type. Axial-piston variable-stroke pumps are used in the same applications as radial types. They are subject to similar limitations of use with fluids which are suitable lubricants and exhibit the same range of relatively high efficiencies. The axial-piston type is shown in Fig. 31. A cylinder block contains several cylinders; as the swash plate rotates, the pistons are forced to reciprocate within their respective bores. The angle of the yoke determines the piston displacement. Control of the yoke can be automatic or manual, as for the radial-piston type. Additional discussiod%f pow13 p"ump d - e ~ i 6features is contained in reference [9].
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
433
MARINE ENGINEERING
Section 4 4.1 Classification and Types. Rotary pumps, like reciprocating pumps, are known as positive-displacement pumps, that is, mechanisms \vhich displace a constant volume of fluid from the suction to the discharge port for each revolution of the driving shafts. Additionally, however, in the case of rotary pumps, the principal motion of the pumping elements is pure rotation and the pump has no valves except in some instances an internal relief valve. It should be noted that constant displacement does not mean constant delivery, since there is always some slippage of flow through the mechanical clearances.
( 0 ) RlNG GEAR WITH ONE MORE TOOTH THAN PINION
I Fig. 32
Simple geor pump
P,
Fig, 33 Lobe pump (two-lobe typd
= pHQ/550
E = pHQ/550 P
E = pq/1714 P
Fig. 34
( b)
RlNG GEAR WITH TWO MORE TEETH THAN PINION AND WITH STATIONARY CRESCENT
Intemal-gear pumps
the lobe to the discharge side. This is essentially a gear pump with only two teeth per gear. Power transfer between the two shafts must be accomplished by timing geara and the flow pulsations are, of course, greater than in the spur gear type. This type of pump is also built with three and four lobes per rotor and as a combination lobe and spur gear type. Renewable wearing pieces are often fitted a t the tips of the lobes so that the original clearances may be ieadily restored after a period of use. 4.5 Internal Gear Type. Internal gear pumps are of two general types. In one type, Fig. 34(a), the ring gear has one more tooth than the pinion, and the seal is maintained by close clearance between the tips of the gear teeth on the open side. I n the second type, Fig. 34(b), the ring gear has a t least two more teeth than the pinion,
and the seal is maintained by a stationary crescent on the open side. I n either type the fluid may enter the gear spaces axially through ports in the cylinder head or radially through ports in the ring gear. Either the ring gear or the pinion may be the idler. The internal gear pump has an advantage over the spur gear pump due to its larger suction port area and longer arc of admission, together with less change ~f fluid direction entering the pump. This results in less suction turbulence and better vacuum performance at equal gear speeds. However, the internal gear pump cannot be hydraulically balanced,and is, therefore; not as adaptable to high-pressure service as the external gear pump, 4.6 Screw Type. A single-screw pump is shown in Fig. 35(a). The rotor is a single-thread helix of circular cross section rotating and oscillating radially in a doublethreaded helical stator of oval cross section. The pitch of the stator helix is twice the pitch of the rotor helix. At least one complete stator helix is necemary for sealing. The rotor is hollow and is driven by a connecting rod with universal joints on each end. The drive shaft is supported on external bearings. Since _the rotor oscillates, the inertia forces limit the speed of rotation. A twin-screw pump or single-idler screw is shown in Fig. 35(b). The rotors are similar in appearance to square-thread screws, but the helical mating surfaces must be concave. The discontinuous contact between the rotors prohibits power transfer between them, so the shafts must be geared together by auxiliary timing gears. Two sets of pumping elements are usually opposed on the same shafts to balance the end thrusts. Since a lensshaped gap exists between the two concave screw faces
SECTION OF CYLINDER
PU
SECTION OF CYLINDER
SUCTION
(b)
TWIN-SCREW OR SINGLE-IDLER TYPE
( C) Fig. 35
Screw pumps
DOUBLE-IDLER TYPE
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
LEGEND 1 2
3 4 ( 0 ) SLIDING-VANE TYPE
(b)
SWINGING-VANE TYPE
Fig. 36
(c)
5 6
ROLLING -VANE TYPE
Vane pumps
when meshed, a number of pitch lengths are necessary to provide adequate sealing. A two-idler-screw pump is shown in Fig. 35(c). The central rotor is a double-threaded helix with convex surfaces, and the idling rotors are double-threaded with concave surfaces. Power transfer occurs directly between rotors so that auxiliary driving gears are not required. Shaft bearings are not required since adequate bearing area is provided by the outside s u r f w s of the screws in the rotor housing or liner. This type is also built doubleended to balance the thrust loads and increase the capacity of the pumps. Transfer pumps are generally double-ended, but where high pressures are required, such as in boiler fuel service, a single-ended pump is used. The twin-screw and two-idler-screw pumps are particularly adaptable to operation at high rotative speeds since the fluid velocity may be much less than the surface speed of the rotors and a hydrodynamic film is developed at high speeds. 4.7 Vane Type. Vane-type pumps may have'sliding, swinging, rolling, oscillating, or stationary vanes. Typical examples are shown in Fig. 36. All vane pumps involve reciprocating motion and the maximum speed is limited by the inertia forces. The pumps shown in these figures are basically identical except for the form of the vanes. The essential elements are a concentric-running rotor carrying reciprocating vanes which maintain contact with an eccentric or elliptical cylinder. Although some pumps have been built with eccentric circular cylinders, it is preferable to make the portions of the cylinders between the p ~ r t concentric s with the rotor to avoid fluid compression. The portions of the cylinder in way of the ports must provide smooth radial acceleration and deceleration. The bearings may be subjected to a hydraulic load approximately equal to the diameter of the cylinder times the length of the rotor times the pressure differential. However, pumps similar to Fig. 36(a) are built with four or more vanes and sometimes with two suction and two discharge ports disposed alternately at 90 degrees around the cylinder. The hydraulic loads are then completely
balanced and the pump is suitable for high-pressure service. Vane pumps usually depend on centrifugal force and springs to maintain the vane tips in contact with the cylinder. Thus, for any condition of vane mass and fluid viscosity, there ia a minimum speed below which the pump will be ineffective. T h i type is at a disadvantage when handling high-viscosity liquids, because the speed and vane weight required for sealing produce high inertia loads on the bearings. The performance is not affected by vane tip wear up to the point where the vanes fail to seat properly in the rotor recesses, or until the weight is reduced below the amount required for centrifugal sealing. 4.8 Rotary Liquid-Piston Type. A rotary liquid-piston pump is shown in Fig. 37. This unit is similar in condruction to the vane pump of Fig. 36(a) in that it consists of a vane-type rotor revolving in an e1lip;tical casing. The operating principle is, however, entirely different. The "vanes" have a different function, since these vanes, in combination with their side shrouds, really constitute a series of chambers which function very much as do the cylinders in a reciprocating air compressor. This is primarily an air removal pump, and depends for its oper* tion on a water (or other liquid) seal which is maintained a t a constant level from an external source of supply. When the rotor is turning a t operating speed, the water, revolving with it, is thrown to the outer periphery of the elliptical casing by centrifugal force. Air, or other gas, is admitted to the rotor chambers by suitable ports, and is entrained between the walls of the chambers and the water. As any particular rotor chamber rotates from the narrow part of the elliptical casing to the wide part, the water, following the contour of the casing, leaves a gap at the inner portion of the chamber. Into this gap air, entering the chamber through slots in the rotor at the bottom of each chamber, is drawn from an inlet port in the fixed central cone in the rotor hub. As this particular &amber passes the wide part of the casing and again approaches the narrow portion of the casing at the opposite side, the entrapped air is forced out through a Cia-
ROTOR ROTATING LIQUID CHAMBER NARROW PART OF CASING ' WIDE PART OF CASING ELLIPTICAL CASING ROTOR VANES
Fig. 37
.
Rotary liquid-piston pump
charge port also located in the central cone. This cycle represents a half rotation, and occurs twice in a full revolution, producing a fully balanced condition in the pump. The air removal capacity of single-stage pumps of this type is relatively constant up to a vacuum of about 22 inches of mercury, dropping off somewhat a t higher vacuums. Thia pump is used for vacuum priming service for centrifugal pumps and has been used in lieu of air ejectors as main vacuum pumps for the condensate system. 4.9 Regenerative Type. The regenerative type of pump is shown in Fig. 38; it does not come within the definition previously given for rotary pumps, since it is not a positive-displacement pump. This pump is definitely not a centrifugal type; the nearest physical counterpart is the helical-flow steam turbine, so that it might well be described as a peripheral or helical-flow type pump. For many years, however, this type of pump has been given the inappropriate designation of "turbine" pump. It is included here because its range of usefulness and performance characteristics are more closely akin to the positive-displacement types. The radial cross section in Fig. 38 illustrates the peripheral or regenerative principle of this pump. The suction passages connect with the casing channel surrounding the impeller vanes at a point near the casing discharge port. Liquid is carried once around the casing , from suction to discharge but, as shown in the radial section, the liquid is recirculated through several impeller passages in its course. The liquid is picked up in one passage between impeller vanes and then thrown out radially into the casing channel by centrifugal force, and
@
-6
8
LEGEND ,
Fig. 38
Regenerative pump
k-
- -
"1 2' 3 4 5 6
7 9 8 10 11
.
LINERS IMP-EUER CASING &COVER O.RING SEALS IMPELLER BUSHING
IMPELLER NUT IMPELLER NUT WAS1 SHAFT PACKING SLEEVE PACKING GLAND SEAL CAGE
436
MARINE ENGINEERING
later picked up by another impeller passage and the process repeated several times between the suction and d i s charge ports. Thus the same liquid particles have energy imparted to them by the rotor not merely once as in the centrifugal pump, but several times. For this reason the head developed by a regenerative pump will be several times that obtained from a centrifugal pump of the same rotor diameter and speed. The pressure in the pump casing channel increases continuously around the periphery from the suction to the discharge port, and, therefore, the pump capacity becomes less as the total head increases due to increased back flow through the casing channel. The regenerative pump characteristics are intermediate to the centrifugal and rotary pump, and in the region of maximum efficiency are more closely related to those of the rotary-type pump. 4.10 Applications. The major field of application for rotary pumps is for high viscosity, small to moderate volume flows (fraction of a gpm to 5000 gpm), and moderate pressures (25 to 500 psig) to high pressures (3000 psig). They are used also outside of this range for both small and large volumes and for high pressures where their self-priming or constant-delivery characteristics are desirable. They cannot compete in efficiency with centrifugal pumps for large-volume, low-pressure, low-viscosity applications. In small sizes, the rotary pump is more efficient than the centrifugal pump, even for low pressures and viscosities. As compared to the reciprocating type, the rotary pump is valveless, it can be run at higher speeds, and is usually smaller, lighter, and cheaper. The smaller sizes of rotary pumps can be direct-connected to standard-speed electric motors. A rotary pump can be designed to pump any material that will flow into the suction port, including gas, vapor,
liquid, slurry, or powder. Rotary pumps are used for circulating lubricating oil to the bearings of engines, turbines, .and other machinery; supplying fuel oil to oil burners and diesel engines; pumping water for marine engine cooling; and for transfer services. Spur gear pumps are used for very small to medium capacities and for high pressures. To reduce the slip and shaft length and deflection, they are provided with ball or roller bearings. Double- and triple-screw pumps are used for large capacities because higher rotative speeds are permissible. Internal gear pumps are limited to lowpressure services. Vane pumps are used for small to medium capacities, for vacuum and low-pressure services, and for low- and medium-viscosity fluids. Rotary liquid-piston pumps are commonly applied as primers for centrifugal pumps on bilge, ballast, and other services requiring a suction lift. They may be incorporated on the same shaft as the main impeller, or they may be used as separate pumps, either belt driven from the main pump shaft or as independent units. Rotary liquid pumps have also been used as vacuum pumps in condensate systems. Regenerative pumps are well suited to small capacities up to 70 gpm at moderately high heads up to 750 feet. They may be used for distilling plant, freshwater, and sanitary services where there is little variation in flow and wide variation in head. Unlike centrifugal pumps, as the flow is reduced the head rises steeply as does the power. Hence the pump driver must be sized for the maximum head that will be encountered. Additional discussion of rotary pump design features and applications are contained in references [9], [15], and [22]. Regenerative pumps are described further in reference [7].
Section 5 Forced-Draft Blowers 5.1 Classification and Types. Forced-draft blowers are of two types, centrifugal and axial-flow. Forceddraft blowers (sometimes referred to as "fans") are used to develop a relatively low pressure, as contrasted with compressors. The term "blowers" normally refers to those equipments which are used for a compression of air from inlet to outlet resulting in an increase of density of up to seven percent. For practical purposes, the air is considered to be incompressible under these conditions. Centrifugal blowers consist of a runner or impeller having blades, whose sides generally are parallel to the shaft axis; the impeller is enclosed in a casing. The blades ,may be radial, backward curved, forward curved, or double curved. Centrifugal blowers are further classified as single-inlet (where the air enters the impeller from one end of the casing) and double-inlet (where the air enters at both ends of the casing).
Axial-flow blowers are usually of the multi-stage type (generally two or three stages) consisting of alternate rows of stationary and rotating blades having an airfoil shape. The casing and rotor are arranged for single inlet and through-flow in the axial direction. The required rotative speeds are relatively high, making them adaptable to direct-coupled high-speed prime movers. 5.2 Basic Fundamentals. The following notation is used in the ensuing discussion:
Q
=
volume rate of flow, cfs
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS TYPE
A
.LEGE% 8!Pate
~ A I C H T [RMVLI eLAMS
IS0
437 LEGEND
Y-lTf RATIO
TIP SPEED . FPS
la0
B c u u v e~ L~n a s
--
Fig. 39
-_.
D
DOUBLE BACKWAR0 CURVED BLADES
180
0.65
108 300 300
0.72 0.60 0.50
RATIO TO CAPACITY A1,MAXIMUM EFFICIENCY
RATIO TO CAPACITY AT MAXIMUM EFFICIENCY
I
-0-
VELOCITY RATIO
Characteristic curves for centrifugal blowen with straight
Fig. 40
Characteristic curves for centrifugal blowers with backward-curved
and forward-curved blades
blader
H, = total head, actual, f t air H', = total head, theoretical, ft air P = brake horsepower, hp Pa = a u horsepower, hp V = air absolute velocity, fps , v = air relative velocity, fps u = impeller peripheral velocity,fps p = weight density of water at 70 F (62.4 pcf) pa = weight density of air, pcf w = impeller angular velocity, rad/sec The net energy output of a blower moving a column of air is in ft-lb/sec, resulting in the following expression for air horsepower: (43)
It follows from the Bernoulli equation that the total head developed by a blower is equal to the sum of the absolute static head and the velocity head at its d i s charge less the sum of the absolute static head and the velocity head at its suction. When the suction is open and at atmospheric pressure, then the total head is as meastlred at the blower discharge. The velocity diagrams in Fig. 9 for radial-type centrifugalgumps apply equally to centrifugal blowers. The develo~mentof the ex~ressionsfor the total theoretical head fiom the ~ u l e r i kor momentum theory are identical for the two cases. The theoretical head is conveniently expressed in the form
The total head in feet of air, H,, may be expressed in the customary units of inches of water, h,, as follows,
q = volume rate of flow, cfm
N re h, h, h',
=
blower speed, rps
= blower speed, rpm = = =
static head, actual, in. water total head, actual, in. water total head, theoretical, in. water
Substitution of this result in the previous equation leads to the following popular expression for air horse power:
h'' = 12 where
0.075 (=)
UZV,~ uzVUz -= g
2230
(47)
Vuzis the tangential component of the absolute
11111' 1
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
438
Fig. 4 2
Centrifugal forced-draft blower
RATIO TO CAPACITY AT MAXIMUM EFFICIENCY Fig. 4 1
Characteristic curves for an axial-flow blower
discharge velocity Ve, and the weight density of standard air is taken as 0.075 pcf. Centrifugal blowers are more responsive to changes in vane outlet angles than are centrifugal pumps. Figures 39,40, and 41 depict the shape of the characteristic pressure-capacity and horsepower-capacity curves as affected by vane design. Static pressure, efficiency, and horsepower characteristics for radial, partially backwardcurved, and forward-curved vanes are shown in Fig. 39. The same characteristics for backward and doublecurved vanes are shown in Fig. 40, and for an axial-flow blower in Fig. 41. The vane system of an axial-flow pump depicted in Fig. 11is equally applicable to an axialflow blower. It is customary to assume that the flow proceeds through the vanes in cylindrical layers of radius R and thickness AR. Then, for each cylindrical layer, the Euler equation is
5.3 Laws of Similitude. The similarity equations and design coefficients for blowers can be derived in exactly the same manner as was done for centrifugal pumps in Section 1. The general sinlilarity laws which state that the capacity varies directly with the speed and the head varies as the square of the speed follow from the fundamental analysis and are equally applicable to centrifugal blowers. The efficiency of a blower is the ratio of the air horsepower (output) to the brake horsepower (input). Using the air horsepower as expressed in equation (43), for a brake horsepower, P, the blower efficiency is expressed as
5.4 Performance. The volume flow requirements of forced-draft blowers depend primarily on the air necessary for efficient combustion plus a margin for leakage losses and special service conditions. Allowances must be made for deviations in air inlet temperature, pressure, and relative humidity and the composition and heating The resulting head quantities at various radial sections value of the fuel. Air leakage losses may occur in ductare then combined into a weighted average head for the work, boiler casings, s-nd air heaters, and at soot blower impeller. The results by this method, when proper al- penetrations. The blower pressure must be sufficient to overcome the lowances are made for friction and shock losses, are in good agreement with test results for axial-flow blowers various elements of resistance in the system, such as the inlet duct, air and gas ducts, steam or regenerative air and pumps. A complete treatment of the theory of centrifugal and heater, boiler wind box, fuel oil burner registers, boiler axial-flow blowers is given in references [3] and [6], and superheater and main tube banks, economizer, and stack. In machinery spaces, the interaction of the ventilation selected topics are discussed in reference [23].
1 2 3 4
Caring inlet Impdlen Guide vanes Labyrinth seals Fig. 4 3
'5 6 7 8
Bearing Bearing housing Support bracket Shaft
Axial-flow blower f':s=
supply and exhaust fans must be considered, as the blower inlet pressure has a significant effect on the blower capacity. The effect of natural draft due to the temperature of the stack gases is relatively small. In older vessels where the pressure requirements were
-
-
-
moderate, the radial or forward-curved blade blower was suitable; their characteristics are shown in Fig. 39. More recent installations employ double-cased boilers with positive furnace pressures requiring higher blower pressures. The backward-curved blade fan meets the
440
requirements for moderate speeds up to 1750 rpm and higher pressures up to 25 inches of water. Figure 40 depicts the characteristics of blowers having backwardcurved blades. For naval installations, the importance of weight and space results in higher blower speeds, and pressures in excess of 45 inches of water. The axial-flow blower with a direct turbine drive fits these requirements. This type is operated in the range of 5000 to 10,000 rpm and gives a performance typically as depicted in Fig. 41: 5.5 Speed and Volume Control. With direct-current motors, 30 to 50 percent speed adjustment below rated speed is generally provided by rheostat control. A further adjustment of capacity may be obtained by the use of outlet dampers. When alternating-current motors are used, they are sometimes of the multi-speed type. Speed combinations of 1800/1200/900/600 rpm are common. The rated speed, in this case, would be 1200 rpm, that is, 2/3 of maximum speed. Thus a maximum speed 50 percent greater than rated would be available; this is an advantage in vessels where a significant boiler overload is a requirement. Inlet vane control is generally used with blowers driven by either single-speed or multi-speed a-c motors. Vanes a t the blower inlet are positioned in such a manner as to vary the rotational velocity of the air entering the impeller, thus modifying the flow characteristics of the impeller to suit the air requirements of the system. When the blower flow is controlled by means of inlet vanes, as opposed to the alternative of throttling the discharged air with an outlet damper, the horsepower requirements are reduced. In lieu of using multi-speed motors and inlet vane control to achieve capacity variation with a-c motor-driven blowers, a variable-speed fluid drive has also been introduced which can provide a flexibility comparable to turbine drives. For the blowers in naval vessels, a turbine drive is used to provide the necessary flexibility of speed control to meet efficiently and rapidly the wide range of boiler operating conditions and maneuvering demands of the vessel. Frequently two blowers are installed with a single boiler; one is used for air up to rated power and two are
put in operation for the boiler overload condition. In general, any type of blower All parallel satisfactorily provided the system head curve for parallel operation does not intersect the blower characteristic a t a point considerably to the left of the blower's best efficiency point. Paralleling of a-c motor-driven blowers is readily accomplished as they are constant-speed machines. For turbine-driven blowers, the nozzle control valves of paralleled units must have matched characteristics. This permits a single control handwheel to be mechanically connected to the nozzle control valves of both units, resulting in both units running at nearly the same speed. Another method is that of using a single steam throttle valve for regulating the steam flow to the two units. Control of volume can also be accomplished by using inlet vanes that can be moved in such a manner aa to vary the rotational velocity of the air entering the impeller, thus affecting the impeller's characteristics. 5.6 Design Features. A centrifugal forced-draft blower is depicted in Fig. 42. Blowers of this type are always mounted with their shaft axis horizontal. The shaft bearings may be of the journal type with ring or forced-feed lubrication, or the bearings may be of the L ball or roller type. The figure depicts a double-inlet blower without inlet ducting; that is, the blower takes suction directly from the machinery space. Partially backward-curved blades are used. The shaft is connected to the driver by a suitable flexible coupling. Such a unit is mounted where ,it can be accessible for operation and servicing in addition to being near the boiler. With the low speeds and moderate pressures found on merchant vessels, noise is seldom a problem, thus permitting the exposed blower inlets. High-speed axial-flow blowers may be horizontal or vertical; a horizontal one is depicted in Fig. 43. Lubrication of the shaft bearings is by pressurized oil furnished by a lube system common to the turbine driver and blower. This type is often enclosed in a separate blower room due to its high noise level. A detailed discussion of design considerations for boiler forced-draft systems is given in reference [24]. A special instance where combustion air is provided under supercharged conditions is described in reference [25].
Section 6 Compressors Compressors may be of either the reciprocating, rotary, or centrifugal type. Compressors for medium and high pressure are of the reciprocating type, and are almost always multi-staged. They are usually single-acting and are further classified as h~rizontalor vertical depending on whether the orientation of the reciprocating action is in a horizontal or 6.1
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
Classification and Types.
vertical plane. Compressors for marine services are usually vertical. When the reciprocating action is in the same vertical plane as the horizontal crankshaft, it is termed "in-lime construction.'' When the vertical plane of the reciprocating action is at 90 degrees to the horizontal crankshaft, the cylinders are disposed about the shaft in such a manner as to resemble letters. Thus a
two-cylinder machine is commonly termed "V" type, a three-cylinder machine is termed "W" type, and a fourcylinder machine is termed "X" type. Compressors for low and medium pressures may be of the reciprocating, rotary, or centrifugal type. R ~ t a r y compressors are further classified as screw, lobe, fixed vane, or sliding vane. Centrifugal compressors resemble a centrifugal pump in construction, and are typified by an impeller with a relatively large number of vanes discharging to a vaneless or multi-vaned diffuser. They may be single or multi-staged. Compressors are further classified as to being air cooled or water cooled. 6.2 Basic Fundamentals. The following notation is used in the ensuing discussion:
HP, .= theoretical horsepower for compression, hp pl = initial pressure, psia p~ = discharge pressure, psia r = compression ratio = p2/p1 VI = actual volume flow at initial pressure, cfm k = exponent for adiabatic compressioii (1.4 for air) The work of adiabatic compression may be exprkssed as
,HP -
( ) [
k 144 33,000 k -
k-l
I
-1
(50)
For air initially a t atmospheric pressure, this equation reduces to I
HP,
= 0.225 Vl(r".286- 1)
(51) and is, of course, applicable to any type of compressor, either single- or multi-stage, in which the compression is substantially adiabatic with no appreciable external cooling during or between compression stages. For very low pressure ratios approaching unity this equation can be simplified to the following approximation
which, with proper regard for units, will be found to be identical to equation (43), which is the expression for the, air horsepower of blowers. In a reciprocating compressor the actual compression curve follows a polytropic curve with the same type of formula, but with an exponent m = 1.3 to 1.35 for air in place of the adiabatic exponent k. The horsepower for isothermal compression is 144
HP c - 33,000 PIVI log. r
(53)
For very low pressure ratios this equation also approaches the air horsepower expression given for blowers. Intercooling between stages in a multi-stage compressor materially reduces the work of compression by reducing the volume flow to the succeeding stage or stages. For minimum work of compression the work should be
.a
44 1
divided equally between the various stages. For two stages this requires pt := (plpz)'" where pi is the intermediate compression pressure. Perfect intercooling requires a reduction of the interstage compression temperature to the inlet air temperature, 80 that the net compressive effectfor any stage and its aftercooler is isothermal. This means that the pv product after each intercooler equals the initid pv value. Application of these considerations leads to the following expression for minimum work of adiabatic compression in a mullti-stage compressor with perfect intercooling : 144 HP, = -
where n represents the number of compression stages. Other important advantages of intercooling are (1) a reduction in the temperature in the succeeding stages, which greatly simplifies the lubrication of the pistons; (2) a certain amount of the original water vapor content is condensed and can be drained off; (3) volumetric efficiency is higher, due to reduction of the clearance expansion loss; and (4) maximum piston loads are reduced. 6.3 Performance. The volumetric efficiency of a compressor is the ratio of the actual intake air volume flow to the compressor displacement. Volumetric losses are due principally to re-expansion of the air in the clearance volume, heating of the air during admission to the cylinder, and leakage past valves and pistons. The first item can be determined from the compressor indicator card as shown on Fig. 44, but the overall volumetric efficiency can be determined only by actual measurement of the air received by the compressor. For multi-stage compressors, the piston displacement of the first stage only is commonly stated as that of the entire machine. Figure 44(a) depicts an idealized indicator card for a single-stage compressor. At point 1, the piston is at the top end of its stroke and a certain amount of gas a t discharge is trapped in the clearance volume. At the top end of the piston stroke, the clearance volume is that which exists between the piston and head and includes the small amount around the piston down to the first ring and under the valves. As the piston starts its return stroke, the clearance gas expands until, at point 2, the inlet valve opens and a charge of air starts to enter the cylinder. The cylinder continues to fill until the piston reaches point 3. Just after leaving point 3, the inlet valve closes, permitting the compression to take place. The piston then returns to point 4, compressing-the gas along curve 3 4 . At point 4, the pressure in tiie cylinder exceekthat in the discharge line and the discharge valve opens, permitting the piston to discharge the compressed gas into the system along line 4-1. Figure 44(b) represents an idealized indicator card for a two-stage compressor. The compression begins a t point 3 and would ordinarily continue to point 4 in a single-stage machine. In a two-stage compressor, how-
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
MARINE ENGINEERING
442
-
(a) SINGLE STAGE COMPRESSOR
-
\6
1 2 3 4
5 6
LEGEND 1 2 3 4
5 6 7
Fig. 45
VOLUME (b) TWO-STAGE
Fig. 44
COMPRESSOR
Theoretical indicator cards for reciprocating compreuon
ever, the compression in the first stage continues to point 3a, at which point the gas is withdrawn from the cylinder and cooled, reducing its volume to that represented by point 3b. Compression then continues in the second stage along curve 3b4a. The work of compression is considerably reduced by a lower inlet air temperature. On vessels where the installdinn --- - - nf - - an . .- inlet air duct is economically unjustified, the compressor should be located in a space where the ternperature normally does not run significantly above the outside air temperature, such as an auxiliary machinery space or pump room. The mechanical efficiency of a compressor is defined as the ratio of the air indicated horsepower from the compressor indicator cards to the brake horsepower input. The mechartical efficiency accounts for all mechanical friction losses and ranges from 88 to 93 percent.
INTERCOOLER ' FAN &BELT PISTON &PIN CRANKSHAFT MAIN BEARING DRIVESHAFT OIL PUMP
Two-stage reciprocofhg compressor
Compression efficiency is the ratio of the adiabatic work of compression to the indicated horsepower from compressor indicator cards. Compression efficiency accounts for all thermodynamic and fluid friction losses. For multi-stage compressors the minimum work of diahatic com~re88ionwith perfect intercooling is some---times used inAplaceof the adiabatic compression work without intercooling. The actual basis used should always be stated in any particular case. The compression efficiency ranges from 85 to 93 percent. In the case of motor-driven compressors, the overall efficiency of the compressor is defined as the ratio of the work of isothermal compression to the brake horsepower of the motor. The overall efficiency ranges from 75 to 85 percent. It should be noted that the compression work equations and corresponding efficiencies are based on actual intake air and are, therefore, not influenced by the clearance volume. The clearance volume affects only the required cylinder sizes for a given duty and does not influence the horsepower required. Air, of course, enters the compressor at the prevailing relative humidity. Owing to the rise of temperature, the
Fig. 46
COMPRESSOR INCREASER GEAR DRIVER INTERCOOLER CONTROLCABINET AUXILIARY LUBE PUMP
.,
Multi-stage centrifugal compressor
relative humidity of the compressed air is much lower when discharged. However, after the air is discharged, it is cooled eventually to a condition approaching its original temperature. At a constant temperature the relative humidity varies almost directly with the pressure so that condensation eventually will occur. I n the case of air discharged to hydraulic-pneumatic accumulator systems, using oil as a working fluid, it is important to remove as much condensed moisture as possible to avoid contamination of the hydraulic oil and corrosion of parts. Moisture causes excessive wear of valves, cylinders, and rings. For this reason moisture separators often are installed at air inlet connectiops. Dust entering compressors from the inlet air causes rapid wear of working parts so that inlet air filters usually are fitted. These are often of a silencer type which also helps to reduce air compressor noise. 6.4 Reciprocating Compressors. A cross section of a two-stage, single-acting, motor-driven reciprocating air compressor for general ship's service is shown in Fig. 45. It is driven by a motor through multiple V-belts. The intercooler is of the air-cooled radiator type with the fan belt driven from the main shaft. The crankshaft is s u p ported by tapered roller bearings. Sleeve bearings are used for connecting rod ends. All bearings are pressure lubricated through holes drilled in the crankshaft; rods are pressure lubricated from a gear-type oil pump on one end of the crankshaft. Cylinders and pistons are splash lubricated. Automatic stopping and starting of this type of compressor is controlled by an air-operated switch which stops the motor when the accumulator pressure has reached the rated value and restarts the motor when the
pressure has dropped a certain amount. In order to start the motor without load, a special unloading device is necessary which forces open the suction valves for both stages and releases air in the discharge line up to the discharge check valve. A del@ feature is incorporated in this unhader which permits the motor to reach full speed before the compressor is loaded. Compressors of this type are used for many shipboard air services such as diesel engine starting, ship's service air, combustion and other control air, tube cleaning, air tools, and sand blasting. They range from 50 to 250 psig and from 15 to 600 cfm. Air or water may be used for cylinder cooling and intercooling. They may be belt driven, direct c o ~ e c t e dthrough a coupling, or closecoupled, and are usually limited to 1150 rpm. Multi-stage compressors are also found with their cylinders in a horizontal plane, disposed radially about a vertical shaft. For charging high-pressure flasks, O2N2 plants, and submarine services, this type ranges from 500 to 5000 psig, from 5 to 100 cfm, and usually has 4 or 5 stages. 'Specific types for refrigeration, LPG, and other cryogenic services require the incorporation of special design features in the seals, bearings, and lube systems to ensure no entrance of lubricant into the gas so as to maintain the integrity of thet>w side, .6.5 Rotary Compressors. A two-lobe-type of positive-displacement rotary air compressor resembles the pump shown in Fig. 33. The speed of this type of compressor makes it well suited for attachment to highspeed diesel engines to supply scavenging air. The principle of operation is identical with the lobe-type rotary pump described in Section 4. Other rotary pump types which are used successfully
444
MARINE ENGINEERING
as air compressors include the sliding-vane type and the liquid-sealed elliptical-casing fixed-vane type, similar to the vacuum priming pump described in Section 4. Rotary types generally are suitable for pressures up to 100 psig, but are predominantly used for lower pressures. Intermediate pressures up to 100 psig are usually furnished by reducing valves fitted in the 100psig (or higher pressure) system being served by a multi-stage reciprocating compressor. 6.6 Centrifugal Compressors. A motor-driven multistage centrifugal compressor of the tjrpe used for general service air supply is shown in Fig. 46. A noise-attenuating enclosure is usually provided, although such an enclo~ure is not shown in Fig. 46. Details of construetion, including the shape of impeller vanes, are quite similar to those of centrifugal pumps. The inherently high speeds of centrifugal compressors require them to be driven through a speed-increasing gear connected to the driving motor or engine, or they
may be direct connected to a turbine. Speeds in the range of 5000 to 40,000 rpm are common. They are used for diesel engine scavenging services and refrigeration and air conditioning services, as well as general ship air service. A high-speed turbine drive is ideal for special services requiring variable speed. Centrifugal compressors may be either single-stage or multi-stage, and they range up to 125 psig and 2500 cfm. A special advantage is their ability to deliver oil-free, clean air. They are essentially free of vibration, and noise can be attenuated to acceptable levels. The theory for centrifugal compressors is similar to, and is as fully developed as, the theory for centrifugal pumps described in Section 1. However, the compressibility of the gas is an additional factor which must be considered. A complete discussion of centrifugal compressor theory is given in references [3] and [6]. Special compressor applications are described in references (171, [18], and [25].
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS LIVE STEAM INLET
1
EXPANDING
*
SUCTION, INLET
CONVERGING
ABSOLUTE PRESSURE AT SUCTION
I
t
Section 7 Ejectors The steam jet ejector is employed principally to remove air and noncondens able gases fromvacuum equipment. The ejector is a simple type of compressor in which high-pressure steam is passed through a nozzle where it is expanded to a pres sure corresponding to the desired vacuum. The steam exits from the nozzle at a high velocity, and air and other noncondensable gases surrounding the issuing jet are entrained in the high-velocity jet and carried into a diffuser. The passage of the mixture through thii divergent tube effects conversion of the kinetic energy into pressure energy, thereby elevating the pressure a predetermined amount above the suction inlet pressure. The ejector, in schematic form, is illustrated in Fig. 47. Ejectors are classified as single or multi-stage depending on whether the compression is accomplished in one unit or several units in series. They are classified as single or multiple element depending on whether one or more units are installed in parallel. They are further classified as to being condensing or noncondensing, that is, with regard to whether or not the discharged steam is condensed. 7.2 Performance. The amount of pressure elevation, commo,nly referred to as compression ratio, that can be accomplished satisfactorily in a single stage of compree sion is governed by the initial steam pressure available for operating the ejector, the vacuum to be maintained, the importance of efficiency measured in terms of steam consumption, and the requirements for stability. Ejector stability under all normal operating conditions is a matter of prime importance. Unstable operation results when the steam jet momentarily and at irregular inter7.1
Classification and Types.
vals breaks down thereby allowing a temporary reversal of flow to take place through the combining tube; this can result in a partial or temporary loss of vacuum. The experience of the ejector designer dictates the maximum allowable ratio of compression per ejector stage that will ensure stability, after giving due consideration to the governing factors. For conditions usually encountered with marine installations, it is customary to employ single-stage ejectors for vacuum requirements up to 26+ in. of mercury, thereby establishing a maximum compression ratio of 8.57 to 1, and two-stage ejectors where the requirements exceed thii figure. Threestage ejectors can be designed having a better operating efficiency than two-stage ejectors for installations operating a t 29 in. vacuum or more, but they are seldom used in marine service by reason of their more. complicated structure, higher initial cost, greater weight and space requirements, and certain operating deficiencies, all of which combine to offset the initial advantage of better economy. An ejector must be designed for a predetermined minimum motive steam pressure. If the steam pressure a t the inlet to the ejector nozzles is less than the design pressure, the ejector will not operate satisfactorily and will not maintain a vacuum. An ejector will work satis factorily with a reasonably higher steam pressure than that for which it was designed, with a resulting increase in ejector steam consumption proportionate to the increase in absolute steam pressure above the design value. It is normal to operate ejectors up to 10 psig above rating to allow for steam line pressure fluctuations. Partial clogging of the nozzle has the same effect as reduced
445
Fig. 48
OISCHARBE Fig. 47
- INCHES
MERCURY
,
Capacity of two-stage ejector with inter- and after-condenser for various air sudion pressures and temperatures
Steam jet air ejector
steam pressure. Selection of the operating pressure is of great importance. Very low pressures are uneconomical, whereas extremely high pressures necessitate very small nozzle bores with consequent fouling difficulties. Operating pressures from 80 to 300 psig constitute a preferable range, with the upper limit being 150 psig for smallcapacity ejectors. Superheat is of no economic benefit; however, a nominal degree of superheat is desirable since it provides added insurance for a supply of dry steam to the ejector nozzles. Wet steam seriously interferes with the operation to an ejector, with the smaller sizes being affected more than the larger sizes. Wet steam also causes rapid erosion of the ejector nozzles and should be avoided by the use of suitable steam-line separators or other drainage devices located close to the ejector steam inlet connection. The air leakage into a system under vacuum is subject to considerable variation and does not lend itself to a precise assessment. Air leakage design values selected must be influenced by experience. The air removed from a steam condenser is saturated with water vapor. It is, therefore, necessary to determine the amount of water vapor contained in the air-vapor mixture to be removed from the condenser by the ejector. The water vapor component is dependent on the absolute pressure and temperature existing at the exit from the air-cooling section of the condenser or external air cooler. To provide ample air ejector capacity, the temperature of the air-vapor mixture at the air-vapor outlet of a welldesigned surface condenser equipped with either an internal or external air cooler is considered to be 73 deg F less than the saturated steam temperature corresponding
to the absolute pressure required at the steam inlet to the condenser. The amount of water vapor with which the air will be saturated can be calculated by the following expression: #
i-
where
W = pounds of water vapor per pound of air leakage pw = steam saturation pressure, inches of mercury absolute, corresponding to the assumed temperature of the air-vapor mixture pa = absolute pressure, inches of mercury, at the airvapor outlet of the condenser 0.62 = 18/29 = ratio of molecular weights of water vapor and air It is customary to provide sufficient ejector capacity to permit a condenser vacuum of 29 in. to be developed when thq cooling water temperature and loading conditions permit, even though a lower degree of vacuum is specified for the base design condition. For this reason the estimated ejector capacity is usually specified at an absolute pressure of 1 in. of mercury when intended for turbines. Figure 48 i&&-rates fypical performance curves for a two-stage ejector having intesYand aftercondensers for various suction inlet temperatures and pressures when handling a saturated air-vapor mixture. Figure 49 illustrates a similar set of c w e s applicable to a single-stage ejector. The steam consumption of air ejectors for a specified condition is dependent on design refinements and other highly important operating features; namely, the ability
446
0
Fig. 49
MARINE ENGINEERING
I 2 3 ABSOLUTE PRESSURE AT SUCTION
4
PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
5
- INCHES MERCURY
Capacity of single-stage ejector forvarious air suction pressures and temperatures
to handle an overload without an excessive loss of vacuum. r a ~ i devacuating ca~acitvfor starting UD and quick pick^* in case the"sysiem & accidentally flioded with air, and the ability to operate satisfactorily under conditions with verv low water auantities being: circulated through the inter- and after-condenser with due regard for pressure drop limitations at full power. For estimating approximately the steam consumption of air ejectors employed for marine service, a value of 4.5 lb of steam required per pound of air-vapor mixture removed at 1 in. of mercury absolute and 71.5 F temperature can be used for turbine service. For other lowvacuum services a value of 6.5 lb of steam required per poqnd of air-vapor mixture at 3.5 in. of mercury absolute and 113 F temperature may be used. 7.3 Applications. Air ejectors are used aboard ship for the following systems: Main propulsion engine condensing plant. Turbogenerator condensing plant. Low-pressure-type distilling plant condensers. Heating system drain condensers. Auxiliary condensers for condensing auxiliary steam. Air ejectors usually are installed in duplicate, one set being sufficient for normal operating requirements and the other set used for standby service and abnormal requirements. This arrangement is commonly termed a "twin unit" and is designed so that either one of the ejector elements may be inspected and cleaned without interfering with the operation of the other element. This may be accomplished by the use of interstage isolating valves or by compartmentation of the inter- and aftercondensers.
Fig. 50
TWO-stagedual-element air ejector
Figure 50 illustrates a twin two-stage air ejector unit with surface-type inter- and after-condensers arranged in a common shell employing interstage isolating valves. This type of unit is used extensively for condensers serving main propulsion turbines and for condensers of turbogenerator sets. The flow of steam and air-vapor mixture is depicted in Fig. 51. Live steam enters at (1) through strainer,,, (2) to second-stage nozzle (3) and also firstrstage nozzle (4). After expansion through the nozzles the mixture of entrained noncondensables and steam from the firststage nozzle enters the first-stage diffuser (9) wherein the first stage of compression to inter-condenser vacuum occurs. The compressed mixture then enters the intercondenser through interstage valve (5), wherein condensation of steam takes place as the mixture passes upward in successive vapor passes formed by longitudinal baffles (6). The remaining noncondensables exit through manifold valve (7) to the suction of the second-stage element (8) and are entrained by the jet of steam issuing from second-stage nozzle (3) and finally compressed to atmospheric pressure after passing through the secondstage diffuser tube (10). The discharge from the secondstage diffuser enters the lower section of the after-condenser through opening ( l l ) , and again passes upward in four successive vapor passes formed by horizontal baffles (12). Saturated noncondensables are expelled through the after-condenser vent opening (13) to the atmosphere.
.
I i 3
448
MARINE EN( PUMPS, FORCED-DRAFT BLOWERS, COMPRESSORS, AND EJECTORS
Condensate discharged by the condenser condensate pump is circulated first through the inter-condenser and then through the after-condenser in a closed circuit; the condensate flows inside the tubes, thereby preventing absorption of noncondensables by the condensing medium while effecting full heat recovery through condensation of the ejector exhaust steam. The drainage from the inter-condenser, which operates at a vacuum somewhat lower than is maintained in the condenser being served by the ejector, is effected through opening (14), by means of a loop seal drain pipe connected to the condenser. The amount of loop seal required depends on the maximum vacuum differential. A minimum loop seal of 7 ft is usually sufficient for most ejector installations. The drainage from the after-condenser dhrough opening (16) is usually led by gravity to a freshwater collecting tank located in the engine room. Although air ejectors can be designed t,o operate without condensers, all marine ejectors are condensing in order to conserve feedwater and recover heat. The use of an inter-condenser between stages of a two-stage ejector effects a substantial saving in overall steam consumption, since the motive steam of the first stage plus some of the entrained vaDor are condensed in the inter-condenser. thereby greatly reducing the required second-stage capacity. A turbine gland steam condensing section is frequently provided in the air ejector after-condenser, thereby effecting additional heat recovery (by the condensate being used as circulating water) through condensation of the turbine gland leakoff steam. When such an arrangement is employed, a gland leakoff exhauster is connected to the after-condenser vent for maintaining a slight suction effect in the after-condenser. The slight vacuum must be sufficient to induce flow of the gland leakoff steam through suitable pipes leading from the turbine glands to the air ejector condenser. A suction effect of 5 to 10 in. of water usually is sufficient for this purpose. In order to cool effectively the noncondensables prior to their exit from the after-condenser vent opening, it is preferable to limit t,he condensate circulating water outlet temperature from the air ejector condenser to a maximum of 140 F. When insufficient condensate is available to maintain this limiting temperature, due to light loading or while maneuvering, recirculation is employed which may be manually or thermostatically controlled. Air ejectors for main engines and turbogenerators always use condensate as the cooling medium. Ejectors for distiIling plants use the saltwater feed as a cooling medium, thereby providing a stage of feedwater heating. The ejector principle is used also to pump water and other fluids against a low pressure by means of a highpressure jet. They are used aboard ship for laundry and plumbing system drains and are commonly referred to as "eductors." They offer a convenient means of evacuating remote compartments that are not connected to a drainage system, but which have a seawater fire main nearby that can furnish the motive water. They may- be motivated by air pressure and often are combined with centrifuga1 pumps to afford a means of priming.
Further discussion of design and operating features of ejectors is contained in references [4], [5], and [26]. References
1 S. A. Moss, C. A. Smith, and W. R. Foote, "Energy Transfer Between a Fluid and a Rotor for Pump and Turbine Machinery," Trans. ASME, 1942. 2 A. Stodola, Steam and Gas Turbines, translated by L. C. Loewenstein, Peter Smith, New York, 1945. 3 George F. Wislicenus, Fluid dfechanics of Turbomachinery, Dover Publications, Inc., New York, 1965. 4 I. J. Karassik and R. Carter, Centrzfugal Pumps, F. W. Dodge Corp., New York, 1960. 5 I. J. Karassik, Engineers' Guide to Centrifugal Pumps, .McGraw-Hill Book Company, Inc., New York, 1964. 6 G. T. Csanady, Theory of Turbomchines, McGraw-Hill Book Company, Inc., New York, 1964. 7 S. Lazarkiewicz and A. T. Troskolanski, Impeller Pumps, Pergamon Press, New York, 1965. 8 F. J. Wiesner, "A Review of Slip Factors for Centrifugal Impellers," Trans. ASME, 1967. 9 Hydraulic Institute Standards, Hydraulic Institute, New York. 10 A. W. Feck and J. 0 . Sommerholder, "Cargo Pumping in Modern Tankers and Bulk Carriers," Marine Technology, July 1967. 11 G. F. Wislicenus, R. M. Watson, and I. J. I
Fluid Power," ilfechanical Engineering, September 1968. 23 A. J. Stepanoff, Pumps and Blowers, John Wiley and Sons, New York, 1965. 24 R. P. Giblon, K. M. Shauer, and I. H. Rolih, "Design Considerations for Boiler Forced-Draft Systern," Marine Technology, October 1969.
449
25 R. P. Hoover and G. J. Rascher, "Design of Simplified Combustion Control System for Supercharged Steam Generators," ASME paper 62-WA-288, November 1962. 26 Heat Exchange Institute Standards, Heat Exchange Institute, Wew York.
M A I N A N D AUXILIARY CONMNSERS
CHAPTER Xlll
FsebaM
I Main and Auxiliary Condensers
Section 1 General Characteristics 1.1 Introduction. Rankine's modification of the Carnot cycle provides the practical basis for the design of the modern marine steam power plant. With the passage of time, the Rankine cycle has undergone a number of changes and modifications but still remains fundamental to the more efficient and refined regenerative and regenerativereheat cycles in common use today. Small steam plants, and more frequently than not the auxiliary marine steam plants, are classic examples of the elementary or basic Rankine cycle. Rudolf Clausius, a theoretical physicist, in 1850 first enunciated the second law of thermodynamics in the following form: "Heat cannot of itself pass from a colder to a hotter bodv." In ~racticalterms this means that in order to constrht an e&ne which will operate in a cycle and produce work, heat must be rejected from the cycle to a heat sink. The smaller the amount of heat rejected to the heat sink in relation to the amount of heat supplied, the greater is the amount of heat energy available to produce useful work. Consequently, the lower the temperature at which the heat energy can be removed from the system, the higher will be the thermal efficiency of the cycle. For the shipboard steam power plant, the ocean, or any body of water in which the ship operates, provides an ideal heat sink. It is a heat absorbing medium of practically unlimited capacity, although its ambient temperature limits, in effect, the extent to which heat energy is made available for producing work by the cycle. The steam surface condenser provides the means for the efficient utilization of the body of water in which the ship floats as a heat sink for the steam power pIant. 1.2 Steam Surface Condensers. The steam condenser controls the exhaust pressure or back pressure at which the turbine operates by condensing the turbine exhaust steam at a pressure corresponding to its saturation temperature. Typical examples of main condensers are shown in Figs. 1, 2, and 3; auxiliary condensers are shown in Figs. 4 and 5. The condensing process is accomplished by the transfer of heat from the exhaust steam to the cooling water and is the basic function of the steam condenser. A heat transfer system must have a heat source and a heat receiver, and if there is to be a significant flow of heat from the source to the receiver, there must be an appreciable
-
Fig. 1
Main condenser for turbines with axial-tlow exhaust (single-plane machinery) [I ]
temperature difference between the two. I n the marine steam power plant, the turbine exhaust steam is the higher temperature heat source and the circulated cooling water is the lower temperature heat receiver. In addition to the requirement for a temperature difference there must also be an area or surface thrbugh which the heat can pass. I n a surface condenser this area is composed of thin-walled metal tubes, enclosed within a shell or pressure container. It provides a means for the heat to flow from the steam to the circulating water, and also is the means for keeping the two fluids separated. I n a surface condenser steam is condensed; that is, it changes from its vapor state to its liquid state at substantially constant pressure, and at the saturation temperature corresponding to that pressure. During the condensing process the steam gives up its enthalpy, less the enthalpy of its liquid (condensate) to the condenser 1
Numbers in brackets designate References at end of chapter.
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
TUBE LENGTH, % Fig. 6
Condensing and circulating water temperature dirtributions typical surface-type steam condenser
throu~houtthe condenser steam space. The T; shin r Fig. 4
Typical auxiliary condenser [2]
cooling water. As the cooling water absorbs this heat, its temperature increases during its passage through the heat transfer tubes. A typical temperature distribution of the condensing steam and the heat-absorbing cooling water is shown in Fig. 6. For convenience, the steam condensing temperature is shown as constant, although this is not completely accurate. The condensing temperature shown in Fig. 6 may be considered as an average temperature; whereas in the actual condenser the steam temperature and preasure vary in a saturation temperaturepressure relation-
/,
S T W INLET AND EXPANSION JOINT
$
-
tion, whii; small, reflects internal pressure losses and must be considered in the design of the condenser. T---h e -smaller the tem~eraturedifference between the heat source and the heit receiver, the lower will be the condensing temperature for a given flow of cooling water and the lower will be.the amount of heat energy rejected from the cycle. If the condensing temperature could be reduced to that point at which it approaches the cooling water inlet temperature, a maximum availability of heat energy for producing useful work would result, and the highest cycle efficiency might be obtained. An approach to such a condition can be made by consideration of a number of factors-among these are increased heat transfer rate, increased cooling water flow, and increased
FLEXIBLE MOUNTING BRACKET
Fig. 5
OUTLET
1 Condensate is of high purity, containing only extremely small quantities of dissolved solids. 2 Condensate is neutral or slightly alkaline in character, in terms of hydrogen ion concentration, with very small amounts of various gases (oxygen, carbon dioxide, nitrogen, or ammonia) in solution. 3 Condensate is collected a t a temperature corresponding to condensing pressure or slightly higher, in the order of magnitude of 1 to 3 deg F, providing the means for effecting condition 2 as described above, ,and enhancing the thermal efficiency of the cycle. $
INLET AND EXPAWION JOINT AUXILIARY EXHAUST DUMP /RIGID MOUNTING
WATER BOX VENT
\CONDENSATE
condensing surface-all of which reduce the temperature difference between the condensing steam and the cooling water. The net effect is a reduction in back pressure on the turbine and improved cycle efficiency. Available space on board ship, the increased volume of steam at low pressures, the high moistlure content of lowpressure exhaust steam, and mechanical limitations associated with turbine design generally determine practical condensing pressures in relation to cooling water temperatures. However, improvement in heat transfer rates achieved through the design of the condenser, selection of tube materials, and tube geometry are practical means for reducing the size, weight, and cost of steam condensers for a given set of design conditions. Therefore, designing for the highest heat transfer rate possible is of primary importance to both the designer and user. Because of its effect on the economy of the steam plant, optimization in condenser design and application is considerably more important with main condensers than with auxiliary condensers. 1.3 Feedwater Recovery. One of the important functions of the surface condenser is the recovery of feedwater. In the process of condensing the exhaust steam from the main or auxiliary turbines, steam condensate is recovered to be used'as feedwater. In a properly designed and maintained feedwater system the condensate has the following three important characteristics:
\-CONDENSATE OUTLET
Auxiliary condenwr arranged for overhead deck mounting and longitudinal expadon wppofl [2]
Current steam plant design pressures with their associated boilers or steam generators require extremely pure feedwater, and it is no longer practical, because of the condensate loss, to construct noncondensing steam plants or steam plants using jet or contact condensers. The small, simple, low-pressure marine steam plant generally has given way to internal combustion engines. The auxiliary condenser, like the main condenser, will also produce condensate of high purity. However, because of size and geometric configuration it does not lend itself to appreciable condensate reheating. 1.4 Condenser Shell Struclure. The condenser shell structure may involve many combinations of geometry; some to obtain an efficient tube bundle arrangement, some reflecting economies in manufacturing costs, some reflecting minimum space requirements, and others a combination of these factors. Further variations occur in relation to the means used to achieve the pressure integrity of the various geometries such as ribbed supports, diaphragm supports, compression stays, self-
supporting shell structures, and various combinations of these. While the pressure differentid between atmospheric pressure and the condenser operating pressure is rel* tively small, approaching 14.7 psi as a limit, the conditions under which this pressure must be supported require relatively heavy or well-brayed structures. A loading of 14.7 psi is small when compared with the operating pressure of boilers or high-pressure heaters, but it is substantial with respect to vessels, such as condensers, subjected to Ahexternal pressure. There are several re* sons for this. First, condensers are relatively large structures, especially the main condensers. Second, structural failure from external loading is usually in buckling with cylindrical shapes, and in bending with flat or nearflat surfaces. This is in contrast with failures, generally in tension, for the cylindrical or spherical shapes used for boilers and pressure vessels subjected to internal pressure. As an example, a 72-in-dia cylindrical vessel designed for an external pressure of 14.7 psi would be suitable for an internal pressure of approximately 130 psi. Even though t'here are many variations in condenser shell geometry, only two basic configurations are generally used. For small main condensers and for auxiliary condensers, condenser shells of circular cross section are most frequent. This type of condenser shell is usually designed with sufficient thickness to be self-supporting under pressure without the use of internal or external bracing. The supports are, &tached so that external mechanical loads are properly distributed through the use of saddles, ribs, doubling plates, and, on occasion, by increased thickness of shell plate over that required for pressure alone. For main condensers, the condenser shell is usually designed as a flat-sided, box-like structure. It may be ribbed, compression stayed, diaphragm-supported, or supported withvarious combinations of bracing to achieve adequate strength and minimum weight.. Flat-sided steam condensers serving engines are never designed as self-supporting box structures because of the excessive plate thickness needed and the resulting excessive weight. When carefully designed as braced structures, they compare reasonably well in terms of weight with large unsupported cylindrical shell condensers. Their major advantages are campactness and their inherent structural rigidity which makes them suitable for supporting the turbine or auxiliary machinery with little additional bracing needed. 1.S Condenser Suppolting Arrangements. Most steam turbines used for m ~ i npropulsion exhaust downwardly into a steam surface condenser located bebw the turbine (see Figs. 2 and 3). However, some turbines have been designed with an axial-flow exhaust, and discharge into steam condensers located forward or alongside of the turbine, with the centerline of the steam inlet of the condensers on, or close to, the same level as the centerline of the turbine (see Fig. 1). Condensers servicing the more conventional downward-exhausting tur-
,
454
MARINE ENGINEERING
bines may be mounted in a number of different ways. Three of the most common arrangements are with t,he condenser supporting the turbines, the turbine support7 ing the condenser, and wit,h the turbine and condenser mounted separately. Condensers used to support main propulsion turbines offer the advantage of compactness. When they are used in this manner, condensers are built with supports as an integral part bf the shell structure. The supports are located high on the shell structure, near the steam inlet. With this arrangement, they hold the weight of the turbine and a minor portion of the condenser (usually the exhaust transition section) on top of the s u p port, +th a major portion of the condenser (shell, tube sheets, tubes and internal baffles, support plates, and the waterboxes) hanging below the support. In principle, the upper portion of the condenser structure can be likened to a number of beams upon which the turbine rests and from which the condenser hangs. The beam structure itself is supported from foundations built into the ship. A typical condenser design for supporting the turbine is shown in Fig. 2. In addition to the integral supporting structure holding the condenser and turbine in position, sway braces and collision chocks at the lower section or bottom of the condenser are also provided. Misalignment resulting from temperature changes during operation can be minimized by locating the condenser supporting structure as close to the turbine centerline as possible. Expansion and contraction movements are especially undesirable if the turbine shaft is attached to separately mounted gears or to electric generators, which are not subject to the temperatuie changes characteristic of the condenser and turbine. Small vertical movement , can become relatively unimportant if the condenser, as a supporting structure, can be used for supporting the turbine and its associated reduction gear or electric generator. In recent years, two of the main objections to using the condenser for supporting the turbine have been largely eliminated. One objection, that of shell structures generally too light to support the turbines, was eliminated by mounting the supporting structure high. The other, that of significant dimensional changes associated with temperature changes during operation, has also been minimized by locating the supports high on the condenser shell as described previously. Using the above configurations, condensers supporting turbines are not limited to small plants. With, separate turbine foundations it is frequently possible to suspend the condenser from the turbine flange. In some cases, depending upon the weight of the condenser and the strength of the turbine casing as a structure, the entire weight of the condenser can be hung from the turbine flange. In those cases where the loading is too great, a portion of the condenser weight may be supported on springs designed to carry the weight of the cooling water in the condenser tubes and waterboxes, with a portion of the load from thermal expansion leaving a downpull on the tybine flange a t all times. Whether
the condenser is wholly supported from the turbine or only partially so, sway braces, to react with normalinertia forces, and collision chocks are also required. A typical example of a main condenser of this general type is shown in Fig.3. Separately mounted condensers and turbines are usually associated with auxiliary condenser installations, but not necessarily so. Common pract,ice is to arrange auxiliary condensers with top supporting structures for mounting on the underside of the deck. In most instances, the auxiliary turbine is mounted on the topside of the deck and the turbine and condenser are connected with an expansion joint between. A typical design is shown in Fig. 5. The principle of a separately mounted turbine and condenser can also be applied to the main propulsion unit,. In this case the condenser is mounted rigidly on foundations near or at the bottom of the shell. The turbine is mounted on separate foundations above, and the condenser and turbine exhaust openings are connected together using an expansion joint or a flexible element. Such installations are rugged, but the requirement of separate foundations for both the condenser and turbine results in more weight and more cost. 1.6 Condensing Pressure and Performance Characteristics. Steam surface condensers for marine power
plants, and in particular those used for the main propulsion system, are required to work at absolute pressures which are somewhat lower than those associated with the more recently built stationary power plants because of the trend to use cooling towers as the heat sink for stationary power plants. For the main condenser, on merchant ships the operating pressure at full power is selected to correspond to a temperature approximately 16 to 20 deg F above the cooling water temperature. The operating pressure at full power for main condensers on naval ships is selected to correspond to a temperature that is 30 to 60 deg F above the cooling water temperature. Large steam condensers in stationary service are usually selected for operating pressures with corresponding temperatures 25 to 40 deg F above the cooling water temperature. An extremely low absolute pressure imposes the need for great care in design in order to meet performance requirements. Pressure losses within the condenser are most critical, and large-volume steam flows are difficult to accommodate without excessive pressure losses, in compactly designed units. For example, a 0.10-in. Hg pressure loss at 1 in. Hg absolute represents a temperature loss of approximately 3.19 deg F ; a t 2 in. Hg absolute, the same pressure loss represents a temperature loss of 1.71 deg F. The respective steam volumes ass~ciat~ed with these pressures are 652.3 and 339.2 cu ft/lb. When considering the physical properties of steam in a condensing environment it becomes apparent that these characteristics must be skillfully accommodated in the condenser design if satisfactory performance is to be obtained. As a consequence, low-pressure condensers must have liberal flow passages on the steam side for
MAIN AND AUXILIARY CONDENSERS Table 1
Recommended Design Conditions, Pressures, Temperatures, Cleanliness Factors, and Water Velocities
Large turbines, high efficien low exhaust%& Intermediate or large turbines, moderate efficiency, moderate exhaust loss Small or intermediate turbines,
low efficiency, high exhaust loss Very small
turbines,
low efficiency, high exhaust loss
Vac. temp., F Abs. press., in. Hg Cleanliness factor, % Water vel., fps Vac. temp., F Abs. press., in. Hg Cleanliness factor, % Water vel., fps Vac. temp., F Abs. ress., in. Hg c l e d n e s a factor, % Water vel., fps Vac. temp., F Abs. press., in. Hg Cleanliness factor, % Water vel., fps
78-83 0 . @7-1.14 85 6-7 93-98 1.56-1.82 85 6-7 100-107 1.94-2.38 85 5.5-6.5 120-125 3.45-3.95 85 5.5-6.5
steam distribution around the perimeter of the tube bundles, as well as longitudinally for their entire length. Obstructions to the steam distribution flow path should be held to a minimum consistent with safe structural design. The selection of the condenser design conditions is extremely important to the economics of the turbine and to the steam condensing plant. While an extremely low back pressure can improve the cycle efficiency, it may also increase the cost of the turbine, the condenser, the condenser auxiliaries, and their installation disproportionately. The most economical backpressure can also be a function of machine size. The cost of small mechanicaldrive turbines requiring low back pressures to attain high thermal efficiency can seldom be justified. As a result, most condensers serving small auxiliary turbines are designed for higher back pressures.than those serving the main propulsion turbines or those driving larger auxiliary equipment, such as the main generators. Table 1 lists recommended condenser design points, giving absolute pressures, corresponding condensing temperatures, cleanliness factors, and water velocities that represent good engineering practice. 1.7 Air Inleakage and Removal. Without exception, steam surface condensers used in marine service are required to condense impure vapors; that is, vapors which contain air or other noncondensable gases. Because most condensers operate at atmospheric or subatmospheric pressures, air leaking into the condenser or into the condensing system is the most common impurity in the condensing vapor. However, some of the noncondensables present are not the result of direct air inleakage, but enter the boiler feed system from the makeup feedwater evaporators or by condenser tube leakage, and may even result from boiler feedwater treatment chemicals. The noncondensable gases most commonly found to originate from these sources are carbon dioxide and ammonia. The presence of a very small amount of air in the condensing vapor has little or no effect on heat transfer.
88-93 1.34-1.56 85 6-7 98-103 1.82-2.12 85 6-7 107-112 a 2.38-2.75 85 5.5-6.5 12&130 3.95-4.52 85 5.5-6.5
'
99-104 1.88-2.18 85 6-7 10k110 2.25-2.60 85 6-7 112-117 2.75-3.20 85 5.5-6.5 130-135 4.52-5.18 85 5.5-6.5
102-107 2.05-2.38 85 6-7 108-113 2.45-2.83 85 6-7 112-117 2.75-3.20 85 5.5-6.5 130-135 4.52-5.18 85 5.5-6.5
10&110 2.25-2.60 85 6-7 111-116 2.67-3.10 85 6-7 117-122 3.20-3.70 85 5.5-6.5 135-140 5.18-5.85 85 5.5-6.5
In fact, some observers have been able to show that under controlled conditions, noncondensables may promote dropwise condensation and actually improve heat transfer. Similar experiments with larger amounts of air present, and with heavy liquid films on the condensing surfaces, have shown substantial reductions in the measured heat transfer. Equally important, however, is that air or other noncondensables may accumulate in the condenser to such a degree that a significant increase in total pressure in the condenser will result. Such an increase in total pressure almost always results in a reduction in cycle efficiency. Should a significant loss in performance occur in a welldesigned condenser, it is usually an indication of excessive amounts of air leaking into the system, or fouled heat transfer tubes. Either or both can cause overloaded air-removal equipment, and the condenser performance will be limited accordingly. Good ~erformanceunder various conditions of air inleakage ;equires adequate and effective cooling and dehumidification of the noncondensable gas-vapor mixtures so that both their saturated weight and volume are reduced before being withdrawn by vacuum pumps or other air-removal equipment. A vapor flow path of the displacement type, free from zones of stagnation and short circuiting, is essential for achieving maximum condenser p'erformance. Adequate noncondensable cooling permits the application of minimum-sized vacuum pumps, thus effecting a saving in weight, space, and cost. The Heat Exchange Institute suggests in their standards that air-removal ep-ipment he sized proportionately to steam flow as in' icated in Table 'L, Paragraph 5-29 of the HE1 Steam Surface Condenser Standards also suggests values for the venting equipment suction temperature. 1.8 Condensate Removal. Surface condensers which operate below atmospheric pressure require pumps for removing the collecting condensate. The volute type of centrifugal pump, designed to operate with low submergence heads, is commonly used for this purpose.
456 Table 2
Recommended Air Pump Capacities [3]
Steam Condensed, lb/hr
CFM of Dry Air at 70 F
Special multistage vertical volute pumps have been built which are capable of handling condensate from a condenser hotwell with as little as 12 in. of water submergence. More frequently, submergence heads are higher and may be as much as 60 in., especially with auxiliary condensers. Main condensers, because of their location in the ship, require minimum-submergence condensate pumps to permit the condensers to be installed low in the ship. Most marine condensate pumps are designed to operate with 18 to 30 in. of water submergence. Condenser hotwells used for collecting the condensate are usually equipped with strainers and antivortex devices at their condensate outlet. They are designed to
457
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
provide optimum hydraulic conditions for the condensate pump. The water level in the hotwell may be controlled by the pump itself (i.e., the pump is permitted to cavitate when there is insufficient condensate to maintain rated flow of the pump) or by level controllers and condensate recirculation systems. In general, most marine installations are arranged with devices to control the condensate level in the condenser hotwell so as to minimize cavitation erosion of the condensate pump impeller. 1.9 Condensate Storage. The steam surface condenser can also serve as a means for collecting and storing feedwater. This function is of considerable importance with respect to the main propulsion system, as the storage volume in the condenser provides some additional surge capacity for the boilers. The storage capacity provided in most hotwells for main condensers corresponds to about 1 to 1.5 minutes of full steam flow. With some installations, condensate levels in the hotwell are used to maintain an adequate suction head for the condenate pumps by using float-operated mechanisms to regulate the condensate pump discharge flow. The deaerating feed tank is used as the principal control point for the boiler feedwater system and provides the necessary surge volume for control.
I
COND~NSATE OUTLETS
Fig. 7
Large single-pau condenser with central air cooler [ I ]
-.
STEAM INLET
/
EXPANSION JOINT -AIR
VAPOR OUTLET
FIXED MOUNTING BRACKET
*.
Section 2 Condenser Design 2.1 Steam Condensing Space. Most marine steam surface condensers are designed so that steam condenses on the outside of heat transfer tubes, with cooling water flowing inside the tubes, and with the steam condensing space contained by the shell. Good design practice requires that the steam entering the condenser have free access to all areas defined by the perimeter of the tube bundles with minimum pressure losses. In order to achieve this requirement, the condenser shell must be sufficiently large in cross section to allow the steam a relatively free passage around the tube bundles and also to provide for longitudinal distribution of the steam to both ends of the condenser. The cross section of flow areas must not be excessively blocked by braces, baffle plates, ribs, diaphragms, or other obstructions. For main condensers, especially the larger ones, the steam inlets cover a substantial portion of the condenser shell, and lateral distribution of the vapor flow within the condenser can be readilv achieved without excessive pressure losses. Should the available space for steam flow be unduly limited, its effect on condenser performance is likely to be serious; therefore, pressure loss d e terminations should be made and evaluated. Steam velocities at the entrance to main steam lanes between tube bundles, or between tube bundles and the shell, should be held below accepted maximum values. Recommended velocities, which vary inversely with
design pressure, are given in Table 3. Longitudinal flow velocities should not exceed 50 percent of the values given in Table 3 at the steam entrance section of the condenser. The main condenser cross section shown in Fig. 7 illustrates the good steam distribution characteristics inherent with condensers of larger sizes. Steam d i e tribution around the tube bundle is achieved by providing space above and below the tube bundle to permit effective lateral distribution of the steam. The auxiliary condenser cross section shown in Fig. 8 illustrates steam distribution provisions for accommodating an off-center steam inlet. Baffles are provided to reduce steam impact effects on the tubes and to direct steam both laterally and longitudinally within the shell.
Table 3
.
Recommended Steam Design Velocities
Condenser Design Pressure, in. Hg
Recommended Maximum Main Steam Lane Entrance Velocity, fpsa
1 2 3 4 5
500 400 300 250 200
Assumed as average velocity with section area under consideration.
flow at the cross-
!
l 6 A G E GLASS
LSTRAINER
PLATE
CONDENSATE OUTLET
Fig. 8
Two-pass auxiliary condenser with centrally located air cooler [2]
The baffles are perforated to allow a predetermined por1 The depth of tube banks (number of tube rows tion of the total steam flow to feed the tube bundle sec- deep) can be controlled by : tion directly beneath them. changing the geometric shape by using substan2.2 Arrangement of Condensing Surface. In recent or trapezoidal tially circular, squaf9- rectangular, years, almost all condenser designers have adopted ar-- -. cross section configurations, rays of heat transfer tubes which are best described as controlling the effective tube bank depth through bundles. Essentially, each bundle is an operating conthe use of variable tube pitch, denser. Small condensers need only one such tube selecting the optimum number of tube bundles per bundle, but the larger main condensers may need two, as needed to control the tube bank depth condenser although seldom more. and steam flow distribution. The principles which apply to the design of a unit condenser tube bundle are quite fundamental. Five of 2 Steam fed to the tube bundle throughout its Ihe more important ones are listed as follows: whole perimeter decreases the average steam entrance
MARINE ENGINEERING
458
MAIN AND AUXILIARY CONDENSERS
459
I
I I
Ii
i I
i Fig. 9
Typical full-load h a m flow distribution at a turbine exhaust flange [4, 51
velocity to the tubes and thereby mitigates the tube bank pressure loss. 3 Steam fed from all directions, including the bottom areas of the tube,bundle perimeter, aids in condensate deaeration. 4 Location of tube bundles as separate modules in the condenser shell promotes steam distribution to the tube bundles and condenser surface areas. 5 Means for direct steam flow to the condenser hotwell provides for condensate reheating. The tube arrangement in Fig. 7 is an example of a single-tube-bundle condenser with no auxiliary steam inlet lanes in the tube bank, but with main feeder lanes around its perimeter. The tube arrangement in Fig. 8 is an example of a single-tube-bundle condenser with no auxiliary steam inlet lanes in the tube bank, but with main feeder lanes around its perimeter. The main steam lanes reduce in cross section at the bottom of the condenser shell, but still provide adequate flow area for condensate reheating and deaeration. The single tube bundle is circular in configuration. 2.3 Steam Flow Profiles and Distribution. Most auxiliary condensershave relatively small steam inlets in relation to their shell size, and should be provided with impingement and distribution b a e s to redirect the incoming steam. The steam flow to small condensers displays little or no stratification. I n some cases, especially with long steam supply lines, its velocity contour may approach the ideal "flat" front flow. Generally, there is no major steam flow distribution problem which is directly caused by the turbine exhaust steam flow pattern. The main condenser, with its large steam inlet, is usually subjected to highly stratified steam flows. Its design geometry, its installation in the ship, or both, can significantly affect the manner in which the undesirable characteristics of stratified flow can be lessened. The flow pattern characteristic of the downward exhaust single-flow turbine is markedly stratified, but the pattern is reasonably symmetrical and usually can be
effectively accommodated. A steam flow distrib~t~ion map for a single-flow, downward exhausting turbine at a high flow is shown in Fig. 9. The numbers shown on the flow contours are the factors by which the average veloc- . ity is multiplied to obtain the local velocity of the steam, and it may be seen that there are areas where the steam velocity reaches 2.9 times the average velocity. With average velocities in the order of 500 fps, such a turbine exhaust pattern can easily have flow areas which sustain velocities in the sonic range. In general, the flow distribution shown in Fig. 9 will not interfere with the performance of condensers with single- or multi-tube-bundle designs, provided the turbine shaft and the condenser longitudinal centerline are substantially parallel. When the turbine shaft is perpendicular to the condenser centerline, the condenser tube bundle design can partially compensate for the nonuniform flow characteristics of the turbine. However, in some instances, offsetting the tube bundles from the condenser shell centerline may be required in order to widen one of the main steam distribution lanes so that steam may be fed into the tube banks at high flow rates from both the bottom and top of the tube bundles. On occasions, b d e s may be needed to redirect some of the steam to other sections of the condenser. The use of baffles for this purpoee will result in some pressure loss in the distribution of steam, although if steam flow conditions are severe, a greater loss might result if baffling were not used. Each case must be resolved with full consideration of the specific conditions which exist. 2.4 Control of Tube Vibration [6]. The failure of heat transfer tubes in steam condensers through the mechanism of tube vibration has often created problems. Fortunately, high incidence of failure from this cause has been periodic rather than general, and for the most part more annoying than catastrophic. If the interior of the condenser is accessible, mechanical means such as installing wood, fibre, or metal slats between the tubes may be used to alleviate the vibration failure incidence. As an interim repair, this is a reasonably effective modification for reducing tube failures. With small condensers, the addition of slats is seldom
I
rr possibility. However, in some instances, slats can be installed between the tubes by cutting holes in the condenser shell. If this cannot' be done, the condenser must he rebuilt internally, and additional support plates and now tubes will be required to correct the problem. With the present technology there is little or no need to experience severe condenser tube vibration failures, r~lthoughan occasional tube failure from vibration may occur, especially when a severe corrosive environment is present. In such cases the tube failures normally show ovidence of corrosion fatigue. The primary cause of tube vibration is from the dynamic effects associated with the flow of steam. A secondary cause may be transmitted mechanical vibration from attached machinery, from nearby machinery, or from other vibration sources, with energy transmitted through the foundation, piping, or other solid connections. Transmitted mechanical vibration, as a general cause, is seldom of sufficient magnitude to be trouble-some. Transmitted vibration from the turbine has commonly been suspected as a major cause of tube vibration but is rarely a significant contributor. The allowable vibration amplitudes for large turbines and the resulting transmitted energy is geaerally too small to develop significant tube vibration, even in those cases where the rpm of the turbine shaft corresponds to the natural frequency of the tubes. Field and laboratory tests have shown tube vibration amplitudes at their natural frequency to be within the limits of 0.011 in. and 0.030 in. when excited at the same frequency from their supporting structure, with the energy input to the supporting structure in the laboratory equal to that measured on a condenser in the field. Vibration amplitudes of these magnitudes do not result in tube vibration failures. The primary cause of condenser tube vibration is steam velocity and its lift and drag effect on the tubes. For condensers it is not usually related to the von Karmen vortex effect of critical velocity. Laboratory tests show that with condensers operating at a low absolute pressure, the velocity energy level at which the von Kannen vortex effect occurs is much too low to have a significant effect on tube vibration. For example, for a a/rin.-~dtube it is about 32 fps. Obviously, the flow energy of steam at these velocities, with the low density associated with condenser pressure, is negligible. However, with auxiliary condensers, some of which may operate at atmosphere pressure where vapor densities are appreciable, the critical velocity effect may need to be examined. It can be computed from the following:
-
Fig. 10
Drag coefficient for tubes in cross tlow [6]
V is the critical velocity when f equals the natural frequency of the tube. S ranges in valut? from 0.10 to 0.30 for Reynolds numbers varying between 100 and 630,000. A Strouhal number of about 0.20 is a typical value for steam entrance conditions to a tube bank, whereas 0.10 is a reasonable value for the innermost sections of the tube bank. For high-pressure closed heaters, the evaluation of the critical velocity is of fundamental importance. In the condensing environment, the velocity effects of steam surrounding the tubes initiate the significant vibrating forces, and the tube4 vibrate at their natural frequency. The steam floh causes a loading to be imposed on the tubes from both drag and l i t effects, and results in tube deflection as a function of this loading. Experience has shown that it is reasonable to assume that the velocity loading effects of lift and drag will deflect the tubes according to the theory of beams under static loading and that the tubes will vibrate with that amplitude without significant amplification. The equations which apply follow:
where
W a= load on tube in direction of fluid flow, Ib Cd = +&agcoefficient A = tube projected area, sq ft p = fluid density, pcf V = fluid velocity, fps g = gravitational acceleration, ft/sec2 +' '< - Pv2
w, = CIA-
I
where
V = critical velocity of fluid stream, fps S = Strouhal number, dimensionless I
f = tube natural frequency, cps d = diameter of tube, ft
.
.-
MACH NUMBER
-2
--
2s
7
'
(3)
where
W L= load on tube 90 deg to direction of fluid flow, lb CL = lift coefficient The value of the drag coefficient Cs in equation (2) may be obtained from Fig.10. It is important to note
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
that Cd reaches a maximum value of 2.1 at Mach 0.95. Data on the lift coefficient CL, or coefficient of lateral thrust, is not as well documented as the data on the drag coefficient. Values which have been suggested range from 0.20 to 1.00, and 0.65 to 1.71. In general, for a cylinder, the lift coefficient is less than the drag coefficient, Cd,and the effect of drag controls the amplitude of vibration. The value of Wd from equation (2), when used in conjunction with applicable beam deflection equations, can be used to determine the minimum distance between support plates or between support plates and tube sheets for steam condensers operating at a low absolute pressure. The design steam velocity should be determined from data similar to that shown in Fig. 9 using full-load flow for determining the average velocity at the turbine exhaust flange. If the product of the average velocity and the local velocity multiplying factor approach the sonic velocity, the design should be based on the sonic velocity. The design should also be evaluated at the lowest absolute pressure expected in the condenser at full load to determine if those conditions are more severe than the design operating condition. The classic equations for beams under uniform loading can be used to compute tube deflections. Because of the relatively high span length to tube diameter ratio normally used, the deflection equation for a simply supported beam may be used for the center spans, and the deflection equation for a beam simply supported on one end and fixed at the other may be used for the end spans. The equations are as follows: For center spans
For end spans
where
Y = tube maximum deflection, in. Wd = load on tube (sum of drag, weight, and contents) between supports, lb L = length of tubes between supports, in. E = modulus of elasticity, psi I = tube moment of inertia, in.4 The mid-span or maximum tube deflection determined from equations (4) or (5) must be less than one half of the ligament between tube holes in the tube sheet or the tubes will contact one another. Good practice is t o limit the deflection to one fourth of the ligament, or one half of the ligament less one sixteenth of an inch, whicheyer is greater. The tube spans or support plate spacing should be finalized only after determination that the bending stresses in the tubes are of an acceptable magnitude, and that premature failure will not occur from fatigue. The
5
1
I 30
40 50 60 70 CIRCULATING WATER INLET TEMPERATURE, F
------Fig. 1 1
80
I
LEGEND IDEAL CONDENSER PERFORMANCE AIR PUMP PERFORMANCE WITH NORMAL AIR LEAKAGE AIR PUMP PERFORMANCE WITH EXCESSIVE AIR LEAKAGE
Effect of air leakage on condenser pressure
maximum bending stresses in the tubes may be computed by using classical beam equations. The fatigue life should not be less than the expected life of the tubes with respect to anticipated waterside and steamside corrosion and erosion effects. 2.5 Air Cooler Design. The purpose of the air cooler is to reduce the water vapor content of the noncondensable gases to be removed from the condenser. The Heat Exchange Institute recommends that the airvapor mixture be cooled 7.5 deg F below the saturation temperature correspond,ing to design pressure or 25 percent of the difference between the saturation temper* t u e corresponding to the condenser pressure and the design inlet circulating water temperature, whichever is greater. In some instances it has been found practical to cool to lower temperatures, and some condensers are designed to cool to within 5 deg F of the cooling water inlet This increased cooling results in a decreased load on the air pump (or other venting equipment) for a given air leakage, and as a result a smaller air pump may be used. Also, im~rovedcondenser performance can result by applying standard-size air pumps to condensers provided with more effective coolig sections than those designed for 7.5 deg F cooling. An example of the effect of air leakage on condenser pressure under conditions of normal operation and under conditions of excessive air leakage or an undersized vacuum pump is shown in Fig. 11. The line 1-2-2'-3 shows the typical performance of a condenser, in terms of absolute pressure versus cooling water inlet temperature, avsuming that air leakage has no effect. The line 4-2-5 shows the pressure which may be
1
46 1
the main tube banks before entering the air cooler section. At the same time, the air cooler should not be such a large portion of the total condensing surface that it significantly detracts from the high heat flux surfaces in the tube banks of the condenser. Air coolers containing 5 to 7 percent of the total condensing surface can be designed to meet normal perfgrmance requirements effectively. The displacement flow path is extremely important to the correct fundtioning of an air cooler. If possible, it should be incbrporated into the design by providing large entrance areas into the air cooler with the flow cross section decreasing gradually to the air-vapor outlet. It is just as important, however, to use a geometry where the length of the flow path through the air cooler is not less than the effective depth of the main tube banks. One of the most important requirements for condenser air coolers is40 provide means for positive flow, in and out of each section, between support plates. This requirement is fundamental to good condenser performance because it eliminates short circuiting and a buildup of large zones of stagnation with the attendant accumulation of noncondensables that result in a loss in performance. It prevents the air pump or venting equipment from being overloaded by causing all air cooler sections to work equally and effectively, and the dehumidification of the noncondensable gas-vapor mixture becomes where optimum. P6= total pressure in condenser The determination of the heat transfer surface required P, = vapor pressure of condensing steam for an air cooler to meet spepific requirements is a comPo = partial pressure of nondondensable gases plex task. The heat transfer rates at.the . - entrance . . - - t,n - - the - -- As the vapor-gas mixture flows from the steam inlet to cooler, where the noncondensable gas content of the the vapor-gas mixture outlet, the ratio of P, to P, mixture is low, are close to the average rates for the condecreases, while P6also decreases, thereby causing flow. denser. The velocity of the air-vapor mixture at the Bince the condehsing temperature also decreases along entrance to the air cooler is usually greater than at any the flow path as P, decreases, the temperature of the other point in the air cooler. As dehumidification tii-vapor mixture also decreases. As this occurs, the becomes appreciable, the heat transfer rate decreases and quantity of vapor in the vapor-gas mixture also decreases. the velocity of the mixture decreases. For the heat The degree to which this is accomplished is related to the transfer system involved, the opposite situation for vapor flow path, which also affects the conductance of velocity is desirable and would aid in sustaining high I;he condensing boundary of the heat transfer system. transfer rates throughout the flow path. In practical As condensation takes place, the vapor-gas mixture designs this is not always possible to achieve and addibecomes more impure, and if optimum heat transfer is tional surface is needed to compensate for the effect of t,o be attained, the concentrating noncondensables must decreasing velocity. be displaced toward the end of the flow path in the conDesign methods for determining air cooler size are, for denser by the incoming purer vapor. This displacement the most part, approximations and are largely based on flow path minimizes the effect of diffusion of noncon- empiricism. The Colburn and Hougen basic method is densables throughout the condenser, and results in Po complex. Reasonably good approximations have also being at minimum values in the major portion of the been made by Mickley. Simplified methods by Colburn unit. With suph a flow path, noncondensable gases can and Hougen, and by Smith, compare well with limited oasily be vented from the condenser without requiring test results. Silver al~&~eniploye*-an ap-proximate excessively large air pumps. method for interpreting data from condensers where low, The ability to maintain such a flow path is, in con- noncondensable gas velocities were encountered, with niderable measure, dependent upon the air cooler and its reasonable success. The fundamentals of these works rolation to the remainder of the condenser. It must be may be found in reference [7]. located so that it draws the air-vapor mixture proporThe air coolers shown in Figs. 7 and 8 are single lionately from all sections of the tube banks. I t must (vapor) pass types. Each air cooler section between be large enough to maintain adequate steam or air-vapor support plates discharges into a collection header through mixture flow velocities at the point where the fluid leaves proportioned orifices, assuring that each section between
produced by the air pump, assuming normal air leakage rind subcooling of the air-vapor mixture. The portion 4-2 of this line represents conditions wherein the air pump cannot remove the air-vapor mixture at as low a pressure rm that which the condenser can produce. The condenser operating presstre will then correspond to line 4-2, and (,he air pump is said to limit the back pressure. The portion of the line 2-5 represents a condition where the Llir pump can remove the noncondensables at a lower pressure than the condenser can produce; under these oonditions, the condenser is said to limit the back presnure, and the condenser pressure then corresponds to line 2-2'-3. The l i e 4'-2'-5' represents vacuum pump performance with a condition of excessive air leakage (the same effect can be caused by too small a vacuum pump or an inefficient air cooler). It should be noted that in this case, the vacuum pump controls the condenser pressure from points 4' to 2'. Above point 2' the vacuum pump is adequate and the condenser will perform along the l i e 2 ' 3 . The total pressure in a condenser at any point is the ' sum of the vapor pressure at the lobal saturation temperature and the noncondensable gas pressures. This may be expressed as:
-
-
462
MAIN AND AUXllLlARY CONDENSERS
MARINE ENGINEERING
-020
. t J
0 0
i
,015
Y+ X
o .010
W
3
z
V)
6 ,005
o 0
Fig. 12
20
40 -60 PERCENT LOAD
80
Performance of typical mechanical deaerating sections with partial air preuure negligible
the support plates receives its share of the load and is properly vented. Liberal air cooling surfaces in the designs shown in Figs. 7 and 8 have the potential of cooling very close to the cooling water ,inlet temperature. These designs effectively use the last rows of tubes in the air cooler which have a very low temperature rise, usually less than 3 deg F, because of the extremely low heat flux characteristic of air-vapor mixtures which have a low percentage of vapor. 2.6 Condensate Reheating arid Deaeration. The principles associated with effective deaeration and reheating of condensate zire not unlike those applicable to power plant deaerators of the tray type. As the heat passes through the films of condensed steam adhering to the condenser tubes in its path to the cooling water, the condensate on these tube surfaces reduces in temperature to slightly below the saturation temperature at condenser pressure. The temperature gradient across these condensate films, which must occur in order to produce a flow of heat, requires that the mean temperature of this film be less than the condensing steam temperature, but greater than the temperature of the contacting tube wall. Unless provisions are made to reheat this cohdensate before it is removed from the condenser hotwell, a measurable amount of heat energy is lost from the cycle. Ih addition, the subcooled condensate absorbs air an! other gases, such as carbon dioxide and ammonia, from the condensing steam. If these gases are allowed to accumdate in the oondensed steam, corrosion of the boiler feed system and the boiler may result. The heat loss from condensate subcooling may be eliminated by directing a portion of the incoming steam to the condenser bottom and hotwell area in such a manner that a substantial amount of its velocity energy is converted to pressure. I n this way, the local static pressure in the hotwell area and under the tube bundles
may actually exceed the static pressure at the condenser steam inlet. If provisions are made, the condensate falling from the tube bundles through this zone of increased static pressure can be heated to a temperature above that which corresponds to the saturated steam pressure at the condenser steam inlet, and thereby effect a thermal gain in the cycle. I n the process of being reheated, the condensate also loses its tendency to retain noncondensable gases in solution. This tendency for the condensate to become deaerated may be augmented by providing deaerating trays or bafhes underneath the tube bundles for receiving the condensate and prolonging its time of contact with the heating steam. The degree of deaeration which may be effected in a steam surface condenser, while principally dependent on the amount of direct-contact heating surface provided, is also dependent upon a number of other factors. I n a steam condenser, deaeration occurs under vacuum and can be affected by excessive air inleakage, especially if the leakaee is below the water level in the hotwell. Other factors which can affect deaeration are the introduction of makeup feedwater, heater drains, and other condensate returns to the condenser. The capacity of the air-removal equipment in operation limits the condenser vacuum (Fig. 11) and consequently affects deaeration. The effect is greatest a t very light loads and with cold cooling water. To install enough vacuum pump capacity to overcome the effects of this characteristic is usually economically unsound. Four factors which influence deaeration are: (1) the noncondensable gas pressure in the deaeration area, (2) the amount and effectiveness of reheat and deaeration surface, (3) the vacuum pump characteristics in relation to air inleakage, and (4) the effectiveness of converting inlet steam velocity energy to pressure in the hotwell area. A further explanation of these factors is as follows : (1) The partial noncondensable gas pressure in the deaeration zone may be minimized with good steam distribution and by eliinating noncondensable buildup from stagnation. Tube bundle arrangements with centraliued air coolers are most effective in this regard. (2) The amount and effectiveness of deaeration and reheat surface below the tube bundles is of extreme importance. Figure 12 shows the deaeration effect of two arrangements of surface working in an environment of very low partial air pressure. With one arrangement, direct surface-producing bafiles made of perforated trays are provided. With the other arrangement, the condensate falls without interruption from the bottom tubes of the tube bundles to the water level in the bottom of the shell or hotwell. (3) The effect of partial air pressure, resulting from vacuum pump capacity limitations, is illustrated in Fig. 13. The curve shown is for a given cooling water inlet temperature. A family of curves would result if a number of d i e r e n t cooling water inlet temperatures were shown.
PERCENT LOAD Plg. 13
Typical effect of partial air preuure resulting from vacuum pump limitations PERCENT LOAD
The combined effect of (2), deaeration trays, and (3), vacuum pump limitations, is shown in the performance curves in Fig. 14. Poorest performance occurs a t very light loads with cold water. Best performance begins a t about 45 percent load, but may be extended to higher loads by providing more reheating and deaeration surface or a larger air pump. (4) The effectiveness of converting velocity energy to pressure energy a t the condenser hotwell zone is usually a function of the directness of the main steam distribution lanes in relation to the inlet steam flow and to the bottom of the condenser. Single-bundle arrangements do not perform quite as well as condensers with two bundles. The latter appears to be an ideal geometry. The prediction of deaeration performance which may be expected from a given condenser design and its associated special deaeration provisions is not easily made. First, such predictions are related to the projected area of the tube bundle as a measure of unit liquid (condensate) flow rate as it falls froq the tubes to the water level below. Secondly, deaeration is related to the height of fall or to the time of contact of the condensate to be deaerated with the reheating and scrubbing steam. Thirdly, if perforated or other types of trays are used, their effectiveness is not only a function of the number of vertically disposed horizontal rows, but also of the area they cover (the projected bundle area or less). In addition, the liquid-vapor flow paths within the deaeration b d e section have a marked effect on deaerator performance. Preferred arrangements have a crosscounterflow design with a positive vapor demand, through the baffled section, to a portion of the main condenser tube bundle directly above. As an approximation, with a tube bundle construction, liberal steam flow around the bundle, and with the effective feeding of steam around its perimeter, a dissolved oxygen content of approximately 0.03 cc/liter may be expected with a 12 in. free fall of condensate. This arrangement is also usually adequate for condensate reheating.
Fig. 14
Performance of a typical deaeraing hotwell
,-
The use of b d e s or trays to augment condensate deaeration in auxiliary condensers is seldom, if ever, practiced; nor are they often used with main condensers. In general, dissolved oxygen levels not exceeding 0.03 cc/liter are satisfactory since most conventional marine power plants include deaerating feed tanks for further deaeration. The deaerating feed tanks reduce dissolved oxygen levels to 0.01 cc/liter asd less before the feedwater reaches the boilers. 2.7 Condensate Returns and Feedwater Makeup. Condensate drains a t pressures above deaerator pressure are usually drained into the deaerator. Most of the condensate drains which are discharged into the condenser are generated in the feed system at pressures less than the operating pressure of the deaerating feed tank. The lower pressure drains are usually from the gland leakoff condenser, the gland leakoff ejector condenser, the air ejector intercondenser, the shell-and-tube feedwater heaters below the deaerator in the cycle, the drain cooler, and the distilling plant. (Since there are many variations in feed system design, the components listed may not apply in their entirety.) Some of these drains in passing downward in the system are cascaded in order to impar$ heat to the feedwater, and thereby reduce thermal losses from the cycle. Not infrequently, and especially in systems using cascading, drains are mixed and are collectively discharged to the condenser. The condensate drains are frequently at pressures below atmospheric and bg a s o q e of air infiltration to the condenser. The steam air ejector inter- sad aftercondenser drains are also a source of noncondensable gas infiltration. In order to compensate for these possible sources of air inleakage, provisions must be made for their deaeration to a degree sufficient to maintain the desired purity of the feedwater. Most condensate drains are returned to the condenser at temperatures substantially above the temperature q
4 64 DISTRIBUTION HEADER
HYDRAULIC POCKET
LIWID INLET
v,=o , n v2
I
,
Vp= 0.75 Vs Vn=VEU)CITY LENING ORIFICE
Fig. 15
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING Table 4
Maximum Recommended Energy Level Coefficients for Flashed Returns
Coefficient, Service Ca Cmtinuous--Sprayed against shell 79 Continuous-Sprayed into hotwell condensate 159 Intermittent-Sprayed on tube bundle8 79 ZntermittentSprayed against shell 159 Intermittent-Sprayed into hotwell condensate or 238 on erosion-resistant target areas a Values of C are for guidance and are b&ed on orifice diameters not exceeding Q4 in. b Except for the deaeration of makeup, this practice should be avoided if possible. I
Design details of makeup, condensate return, and bypass steam distribution system
pressure differential across the orifices; therefore, the velocity must be maintained at sufficientlylow values so corresponding to condenser pressure. In general, a tem- that excessive erosion of impingement surfaces will not occur, Acceptable velocities for the flashing condensate perature of 5 to 10 deg F above the condenser tempera- mixture discharging from the orifices are given by the ture provides sufficient energy to deaerate the condensate equation: returns to less than 0.03 cc/liter. However, the design of the system within the condenser can have a significant effect on its ability to provide adequate deaeration. Most frequently the design of the condensate return system inside the condenser incorporates a "perforated" pipe header to distribute the condensate over a wide where area. These perforations are usually drilled holes which V = acceptable velocity of mixture, fps act as orifices. Their size and number are predetermined. p, = density of mixture, pcf The smaller the diameter of the holes and the greater C = energy level coefficient (see Table 4) their number, the less concentrated will be the jet energy The density of the mixture, p,, is calculated immedieffect from the fluid flow streams upon emerging and ately downstream of the orifices. The velocity of the discharging into the condenser; in addition, a greater amount of direct-contact surface will be provided for mixture, V, is calculated by using the discharge area of heating with condenser steam should the drain temper* the orifice in conjunction with a suitable coefficient. tures fall below the condenser temperature, or for di- For orifices made by drilling holes in a pipe header or sipating high-energy flashed vapor should the temper* plate, a discharge coefficient of 0.80 to 0.85 may be used. The design details of a typical spray header distribution ture become higher. The return distribution headers should be located in system are shown in Fig. 15. The relationship of the condenser steam distribution space and above the velocities given are for relatively short lines. It is maximum expected water level in the hotwell. The important to adhere to the velocity sequence shown if sprayed returns may be arranged to impinge directly uniform distribution is to be achieved. For long spray against the condenser shell or into the condensate in the headers, the friction losses should be determined. Headhotwell on the bottom of the condenser. In general, as ers should be sized so that the friction loss from any the temperature of the condensate returns approaches or section to the end of the header is approximately equal becomes lower than the temperature corresponding to to the velocity head at that section. The preceding design requirements are applicable in condenser pressure, the sprays should discharge into the condenser at a higher position to allow for deaeration principle to both main and auxiliary condensers. 2.8 Bypass Systems. Nuclear-powered steam power and/or reheating. High-temperature, high-energy replants require heat removal from the reactors even when turns are preferably sprayed directly into the collected condensate with their velocities and distributions de- the propulsion turbines are out of service. This heat signed to avoid splashing condensate on the tube bundle. energy, in the form of generated steam, is arranged to Makeup, although frequently admitted to the hotwell, bypass the turbine and discharge into the condenser for requires deaeration. Very effective deaeration of makeup energy absorption. The steam generated for reactor may be achieved by spraying the makeup high onto cooling may be as much as 10 percent of the reactor the tube bundle, preferably on the top tubes from above. capacity and must be cooled, condensed, and returned However, the design of the sprays must provide a good to the steam generator. Since the available energy from this steam is not being distribution, low admission velocities, and minimum hole used to produce power, its total enthalpy less the enthalpy diameters in order to avoid damage to the tubes. Considerable velocity may develop as a result of the of its condensate at condensing temperature is absorbed
by the condenser. The main condenser when supplied with rated cooling water flow is more than adequate to absorb this energy. The important consideration is that of providing means within the condenser to dissipate the velocity energy of the steam in addition to absorbing its available heat. The most elementary and yet most effective system for handling bypass steam employs two pressure breakdown elements. The first is the friction in the piping between Ihe condenser and the steam source; the second is the pressure loss through the orifices provided in the steam bypass distribution system imide the condenser. The greatest portion of the total pressure loss occurs within Ihe condenser, and the steam bypass distribution system must be capable of dissipating this energy so that it can be absorbed as heat without causing mechanical damage. Three of the most important conditions which must be satisfied are: (1) the steam must be desuperheated within the condenser rapidly and effectively; (2) shock wave effects from sonic discharge impinging on tubes or other metal surfaces must be avoided; and (3) the massvelocity energy effects must be distributed over a wide urea. In designing a steam bypass distribution system to meet these requirements, condition (1) can be satisfied by introducing the bypass steam into the condenser directly below the tube bundles. In this way it is discharged into the "rain" of its own condensate, effectively desuperheating the steam without additional desuperheating sprays or other automatic devices for temper* ture control. Condition (2) can be met by selecting appropriate diameters for the holes in the distribution headers as discussed in the previous section. In this case, however, the vapor leaving the breakdown orifices in the distribution headers will be substantially superheated and will be at velocities of Mach 1 or greater. Good practice is to discharge the steam above the condensate level in the hotwell such that it is directed at the aondensate surface in the bottom of the condenser; this avoids damaging metal surfaces by erosion, temperature, shock wave effects, and vibration. When bypassed steam is discharged into the exhaust trunk, care should be taken to avoid its impingement on the condenser tubes. Further mitigation of shock wave and highvelocity erosion effects may be achieved by using many small-diameter orifices in the distribution system within the condenser. The shock wave from sonic flow through an orifice dissipates in a distance of about 30 orifice diameters; therefore, it ib good practice to exceed this distance by a substantial amount before allowing the stream flow to impinge on metal surfaces, with the exception of specially designed impingement baf£les. Condition (3), that of spreading the velocity energy effectsover a wide area, is extremely important and will influence the requirements under (2) above. In the design of headers for the purpose of flow distribution it is important that the maximum fluid flow velocity in the header does not exceed 75 percent of the fluid velocity leaving the orifices in the header. If it does, uneven
465
distribution will result and the effectiveness of the system will be impaired. 2.9
Cooling Water circuit
a. Condenser Cooling Water Passes. Condensers in which all of the cooling water flows through the condenser tubes in one direction, from cooling water inlet to cooling water outlet, a d Irnott-11as single-pass condensers. Figures 3 and 7 are examples of typical singlepass condenserg, Configurations where all of the cooling water flows from one end of the condenser to the other through one half of the tubes and then returns in the opposite direction, through the remaining half of the tubes, to be discharged are classed as two-pass condensers. Figures 2, 5, and 8 are typical examples of condensers of this type. The tubes of multi-pass condensers are arranged in groups of substantially the same number, so that the cooling water which flows through each group successively from water inlet to water'outlet is at the same velocity. Each group of tubes represents a water pass, and the number of such groups in a given condenser gives rise to the terminology of one-, two-, or three-pass condensers. The air cooler tubes are preferably located in the first water pass so that the coldest water is made available to cool the noncondensable gases. The use of more than one water pass in the design of a condenser increases, in effect, the tube length or the distance the cooling water must travel from the water inlet to the water outlet. The product of the tube lengtk and the number of water passes is the effective tube length of the condenser, and as such may be directly compared with the actual tube length of single-pass condensers. For a given heat load, the primary variables which determine the condenser tube length are: (1) (2) (3) (4)
the temperature rise of the cooling water, the water velocity through the tubes, the tube diameter and gauge, and the condenser surface and space conditions or requirements.
With respect to (I), the greater the temperature rise, the longer the effective tube length required for a given velocity. With respect to (2), the higher the water velocity, the longer the effective tube length required for a given temperature rise. Regarding (3), the larger the internal tube diameter, the longer the effective tube length required for a given velocity and temperature rise. With respect to (4)) the condenser surface and the available space conditions, the effects of (I), (2)) and (3) will contribute to the fipl selectiq of effective tube length and, consequently, the required number of water passes in the condenser. One of the important factors which must be considered in condenser design and which directly afTects the number of water passes required is the ratio of condenser shell length to mean condenser shell cross-section area. A long shell of small diameter does not lend itself to good longitudinal steam distribution; a short shell of large
MARINE ENGINEERING
466
Fig. 16
,
Condenser with circular array of tubes and peripheral steam inlet and central air cooler with venh cascaded to inlet water end of flnt-pass tubes for final cooling
diameter may result in objectionably heavy or deep tube banks. Either of the geometries can result in excessive pressure losses on the steam side of the condenser. Single- and two-pass condensers are the designs most frequently applicable to marine condensers, with the single-pass more commonly used with main condensers to accommodate the hydraulic characteristics of scoops, and multi-pass condensers more commonly applicable to auxiliary condensers. Constructions using three or more passes are only occasionally needed to meet the performance requirements associated with marine steam power plants. Multi-pass condensers require less cooling water but more surface than single-pass condensers for a given heat load. The smaller cooling water piping associated with the smaller flow may be desirable in some cases. b. Condenser Waterboxes. Waterbox designs used for marine condensers may be divided into two classes: the bonnet type and the ring-and-cover type. In recent years, the bonnet-type waterbox has been used almost to the exclusion of the ring-and-cover type. Figures 1, 2, 3, 4, and 5 illustrate bonnet-type waterboxes. A typical ring-and-cover type of waterbox construction is shown in Fig. 16. The bonnet-type waterbox is constructed with all elements of the waterbox, including waterpass partitions, flow dividing partitions, and inlet and outlet cooling water nozzles, combined as a unit structure. The ringand-cover type of waterbox is constructed with all elements of the waterbox, except for the cover, but including waterpaas partitions, flow-dividing partitions, and inlet and outlet cooling water nozzles, combined as a unit structure comprising the ring section only. The covers are separate and attached by means of bolted flanges. It is obvious from s comparison of the figures shown that ring-and-cover waterboxes provide the greatest accessibility for tube and tube sheet maintenance and tube replacement. There is little or no advantage in
using ring-and-cover waterboxes unless they also are equipped with cooling water inlet or outlet connections, or both. Normally, bonnet-type waterboxes without cooling water connections are just as easily removed to provide maintenance accessibility as their ring-and-cover counterparts. Condensers with an even number of passes have cooling water connections on one end only, and are almost invariably built with bonnet-type return waterboxes. The nozzle-end waterboxes may be of either the bonnet or ring-and-cover type. Nozzle-end waterboxes of the ring-and-cover type provide complete access for servicing tubes, tube sheets, and tube replacement without the need for disconnecting the inlet and outlet cooling water piping. Condenser maintenance is less d i c u l t than with nozzle-end bonnettype waterboxes, and tube repair, cleaning, and replacement can be easily done. Nozzle-end waterboxes of the bonnet-type must be removed for servicing tubes and tube sheets and the cooling water lines disconnected, except in the case of very large condensers wherein maintenance personnel may enter the waterboxes for tube servicing. For retubing, however, the waterboxes must be removed. The disadvantage of the bonnet-type waterbox is mainly the limited accessibility to tubes and tube sheets for maintenance. The advantages are lighter weight, better hydraulic design, lower cost (especially when corrosion-resistant materials are required), and fewer bolted joints. The selection of the type for a given installation should take into account the preceding, and also the materials used for tubes, tube sheets, and waterboxes. The use of 90-10 copper nickel for these parts with its excellent corrosion resistance (rather than aluminum brass tubes, naval brass tube sheets, and steel or coated steel waterboxes) has been a major contributing factor to the use of bonnet-type waterboxes for marine condensers. c. Single- and Divided-Circulation Condensers. Most condensers used in marine services, and especially
.
MAIN AND AUXILIARY CONDENSERS
those used for auxiliary applications, are of the single-circulation type. Single-circulation condensers are characterized by having but one cooling u-ater circuit from inlet to outlet. Condensers of this type must be taken out of service when maintenance is required; and if the main condenser is involved, the ship may also remain oet of ~erviceuntil all repairs are made. Condensers provided with divided-circulation coolingwater circuits may be operated with one half of the condenser in service while the other half is out of service. Maintenance is obviously restricted to t,he cooling-water side of the condenser. Condenser designs with open center steam lanes, such as are characteristic of two- or four-bundle tube arrangements, are characteristically ndaptable to a divided-circulation design. The cooling-water system must also be divided and valved so that each half of the condenser may operate as s separate unit. The design, when operating with both halves in service, performs in identical fashion to a singlecirculation type of condenser. When one circuit of a divided-circulation condenser is isolated, the condenser operates with one half of the normal cooling-water flow and one half of the condensing ~urfaceactive. The other half is out of service and the wsterboxes of the condenser may be opened to provide uccess to the tube sheets such that urgently needed minor repairs may be accomplished, especially temporary correction of leaking tubes or tube-end connections. Repairs may be made by plugging and/or welding as needed. Ordinarily, the condensing temperature at full load wiil not exceed 105 F for condensers designed for 2.25 in. Kg absolute with 85 F cooling water. However, with one half of the condenser out of service and with the full steam load condensed by only half the surface and half of the cooling water supplied to the condenser, the back pressure and condensing temperature will increase. At equivalent steam loads the cooling-water temperature rise will approximately double. Should the condensing temperature reach levels which impair maintenance activities, the load on the turbine may be reduced to provide more acceptable working conditions. Waterboxes for divided circulation may be made as single boxes with a partition or as separate boxes. If partitioned single boxes are used, each side of the waterbox must be capable of withstanding the design working pressure when the other half is out of service. Dividedcirculation cooling circuits are almost never used for auxiliary condensers. d . Pressure Losses in Cooling-Water Systems. I n ostablishing the total dynamic head required to circulate cooling water through the condenser cooling-water ~ystem,the pressure loss resulting from flow through the condenser must be determined. For marine power plants the pressure loss in the cooling-water circuit of the main condenser is solely from flow friction; but the pressure lbss of auxiliary condensers may have a static-head component, such as in those cases where the overboard discharge is above the waterline of the ship.
W
467
0.90 0.80 0.m
5 *L 0.60 O
0.50
$ 0.40
L
3 0.30 K
ki
Ilz5
IL ,Q20
3 '
%
r
0.15
ui
g
0.10 ,0.09
0 0.08
g
0.07
L
0.06 0.05 3
4
5
6
78910
WATER VELOCITY, FPS TUBE SIZE CORRECTION TUBE SIZE 5/8" 3/4" 7/8"
Fig. 17
,
16 BWG 1.10 1.06 1.06
18 BWG 1.00 1 .OO 1 .OO
.,. 20 BWG 0.94 0.95 0.96
ridi ion loss in condenser tubes [3]
The most reliable data available for computing condenser cooling-water flow losses are those published by the Heat Exchange Institute, "Standards for Steam Surface Condensers" [8]. These Standards provide graphs for determining the friction loss through tubes for various water velocities with correction factors for gauge and tube diameter and for cooling-water temperature. The standards also include graphs for determining waterbox and tube-end losses for velocities through tubes or nozzles for single- and two-pass condensers. The head loss resulting from circulating-water flow through tubes and in the condenser waterboxes is shown in Figs. 17 and 18 respectively. The data in Figs. 17 and 18 are limited to those tube sizes and gauges normally used in marine condensers; more complete informationis given in the Heat Exchange Institute Standards. 2.10 Mechanical Design practice
a. Tube Sheets and Tube Spacing. Tube sheets are drilled to a prescribed m&termto raeive the heat transfer tubes. The holes into which the tubes are secured have a number of geometries intended to meet specific needs. These geometries are described in paragraph c. The tube holes are first drilled and then reamed for finish and sizing. The shell sides of the tube holes are in. X 45 deg bevel, or chamfered, usually with a preferably with a in. radius. The finished dimension
x6
x6
468
MARINE ENGINEERING
2.0
I
I
I
I
I
I
I
I
I
I
Table 5
Recommended Tube Sheet Thickness [3]
Area of Tube Sheet, sq in. to 1965 1966 to 3739 3740 to 8495 8496 to 30791 30792 and up
Table 6
Thickness of Tube Sheet, in.
Diagonal Tube Pitch Data
Tube diameter, in. Tube pitch (30' X 60' diagonal), iw. No. tubes ~ e sa r ft of tube sh&t *
96
3/a
W
1H6
1x6,
1%
189
147
106
,
i
0
I
I
I
~ t
1
2 3 4 6 7 8 9 VELOCITY THRU TUBES OR NOZZLE, FPS
Fig. 18
t
O
Waterbox and tube end losses [3]
of the tube hole should be made equal to the nominal tube diameter plus a minimum of 0.005 in. and plus a maximum of 0.007 in. The thickness of tube sheets varies with tube sheet size (area). Tube sheets drilled for % in., % in., or 74 in. tubes which are to be rolled or packed should not be less than J$j in. thick. If the tubes are welded and rolled, the tube sheet should not be less than % in. thick. The earlier recommendations of the Heat Exchange Institute regarding tube sheet thickness are shown in Table 5. The Heat Exchange Institute no longer publishes dimensional standards; however, the standard established by Table 5 remains in general use. The location of drilled holes in tube sheets for receiving tubes is not only a function of the overall tube bundle design but also a function#of tube diameter and pitch. Most condensers are designed with a 60 deg triangular pitch, for which standard dimensions have been developed. ".'he accepted standard used by most condenser designers is given in Table 6. Condensing pressures higher than the normal values shown in Table 1 permit reducing the pitch to values lower than those shown in Table 6, except that the pitch should never be reduced to such that the ligament is less than in. The tube pitch in combination with the tube bank depth determines the internal pressure loss, and design conditions will determine if the values of tube pitch shown in Table 6 should be strictly followed. The relation of tube pitch to internal losses in condensers is discussed in Section 4. b. Support Plates. The condenser tube support plates serve a number of purposes. They are used to support the tubes between tube sheets in order to control tube vibration and to provide means for bowing the tubes for reducing stresses from differential expansion.
Also, they provide a convenient means for reinforcing the condenser shell against external pressure. Support plates should be sufliciently thick to provide a good bearing surface for the tubes they engage. Typito % in. for small condensers, cal thicknesses are including most auxiliary condensers. Thicknesses of to J$j in. are used with large condensers, including main condensers.' The tube holes should be chamfered to avoid cutting tubes on sharp edges. Accepted pracin. radius tice is to use a H6in. X 45 deg bevel or a on both sides of the support plate, or the equivalent by the use of wire brush deburring. The dimensions of the tube holes should be equal to the nominal diameter of the tube plus a minimum of 0.010 in. and plus a maximum of 0.015 in. To provide a smooth surface, the tube holes should be reamed after drilling. The spacing of the support plates, in most instances, will be about 40 to 45 tube diameters, but these dimensions should be verified using the procedures outlined in Section 2.4. Support plate areas which do not engage tubes should be cut out or relieved as much as possible to provide means for longitudinal steam distribution. When they are also used for reinforcing the condenser shell, support plates should be securely welded in place and provided with sufficient bearing area against the shell to properly distribute loading. Since the support plates act as edge-loaded diaphragms, they may require staying along the length of the shell to avoid buckling under heavy loading. c. Tube-to-Tube-Sheet Joints. Condenser tubes in marine condensers are usually one of tn-o standard sizes: % in. or yi in. outside diameter. The % in. od size is most commonly used for merchant ships and 44 in. for naval ships. The wall thickness most frequently employed is 18 BWG (0.049 in.). The method selected for securing condenser tubes to tube sheets should provide leak-tightness and strength as a supporting structure for the tube sheet, or a leaktight sliding joint to compensate for expansion. The various methods maytbe used in combination to provide characteristics consistent with design requirements for
x6
I
specific service conditions. Details of the more commonly used methods of securing tubes are shown in Fig. 19. Figure 19(a) shows a typical configuration of a tube expanded into a tube sheet with a rolled straight section and a swaged flared section. The flare is for irnpro9ing the tube inlet-end geometry to miniize cavitation effects from the entering cooling water. Also, the tubes mav be rolled and flared a t both ends. The cvlindrical section of the rolled joint in Fig. 19(a) is shown smooth, but it may be grooved for an improved holding ability. The grooves, which are machined into the tube holes, are usually two in number, spaced about in. apart, and are about 0.015 in. deep X in. wide. Figure 19(b) shows a typical configuration of a tube expanded into a tube sheet with a straight section only. When used, this design is applied to the tube outlet end. The tube hole is shown without grooves, but grooves may also be added to improve the holding ability as noted above. Figure 19(c) shows a typical design of a packed tubeto-tube-sheet joint using a combination of fiber rings and metal-foil packing. The ferrules are provided with clearance to allow for longitudinal tube expansion, but the clearance is limited to prevent excessive longitudinal tube movement. If the tube is packed on both ends, ferrules are required to prevent the tube from coming out of the tube sheets. Figure 19(d) shows a typical design of a packed tubeto-tube-sheet joint using a combination of fiber rings and metal-foil packing. The tube hole is threaded so that the metallic packing can be driven into the threads and hold without ferrules. This design is used for the outlet ends of the tubes only and must be used with inlet-end geometries (such as rolled or welded joints) which secure the tube in position. A geometry similar to that in Fig. 19(a) may be used to secure a tube to a tube sheet by welding; while the design incorporates a flared entrance to improve water inlet flow conditions, this design is alsp used at the outlet end of the tube. A geometry embodying a straight section only, with no flared section, may also be used, but specificallyfor the tube outlet end. The tubes should be welded first and then rolled. Welded joints without rolling me susceptible to fatigue failures from vibration. Properly rolled joints are substantially as strong as the tube itself. This is generally true only when the tubes are expanded into holes with grooves, or welded and rolled; however, the uniformity of holding strength will vary between these methods. d. Tube and Shell Expansion. The condenser operates at a temperature quite different from that prevalent during assembly. Even under normal operating conditions the temperature differences are sufficiently lwge to generate stresses high enough to cause tube failures. Three methods are commonly used to avoid the undesirable stresses which can result from thermal expansion differences. The use of packed tube ends, or tubes with one end
(a)
(b) CUTLET END
INLET END
EXPANDED ONLY
EXPANDED a FLARED METALLIC RING
,
METALLIC RINGS
-h
x6
(cIINLET OR OUTLET END PACKED WITH METALLIC WKING WITH FERRULES
(d)
OUTLET END PACKED WITH METALLIC =KING W W T FERRULES
Fig. 19 Tube-to-tube-sheet loinb
packed (outlet end) and one fixed (inlet end), allows movement of the tube longitudinally in the tube sheet joint, thus minimizing differential expansion stresses. This method has been used successfully for many years but does have a tendency to develop leakage (cooling water to condensate) at the packed ends. However, with proper design, assembly, and maintenance, the leakage is small enough to be tolerated with moderatepressure steam plants; but s > j han arrangement would not be acceptable for use with a nuclear steam plant. When tubes are expanded on both ends or welded and expanded on both eads, differential expansion between the shell and tubes resulting from different materials and temperatures may be accommodated by arranging the support plate positions so that the tubes are bowed. Under compressive loading the tubes act as eccentric columns which deflect easily and thereby hold the stresses caused by end loading to acceptable levels. A disadvantage of this method occurs if operating conditions place the tubes under tension. In this case there is little flexibility in the tubes and the expansion load can cause excessive tube sheet deflection. If the differential expansion movement is large enough, very high local stresses can develop and s'ome tube-to-tubesheet joints may fail. An alternative method, which is used when the tubes are rigidly fixed in both tube sheets, is the shell expansion joint, Figure 20 shows a toroidal joint which offers substantial flexibility. It is especially suited to condensers with long tubes &may_becused in multiples if one does not provide sufficient flexibility. Figure 21 is a diaphragm type of joint, shown with the diaphragm attached to a circumscribing ring. The flexibility is less than that of a toroidal joint of equal diameter, but it is less expensive to build. Where space permits, it may be increased in diameter and used as a single rather than a double diaphragm. I n such cases, the circum-
470
-
MARINE ENGINEERING Table 7 Condenser Shell Thickness for Cylindrical Shells (fabricated stbel) [3]
Cross Section Area of Shell, sq in. to 2460
2461 to 5670
Fig. 20
Toroidal shell expansion loinl
-
Fig. 21
Diaphragm shell expansion foint
scribing ring may be used as the tube-sheet bolting or welding flange. e. Condenser Shell. The important geometrical characteristics of marine condenser shells are described in Section 1.4. The methods for designing condenser shells to withstand external pressure generally vary with space requirements. Common practice is to provide compression stays to carry the external loading on large flat surfaces in preference to the use of ribs. While ribs may also be used, they are sometimes employed as secondary members or to distribute loads. Support plates may be used as full ribs, partial ribs, or as anchors for compression stays. I n the case of cylindrical shells, support plates may be used as circumferential supporting rings; however, cutouts must be provided for longitudinal steam flow. Also, cylindrical shells may be designed as self-supporting structures with the support plates used to carry external mechanical loading. The Heat Exchange Institute promulgated standards for the thickness of condenser shells. These standards, which are listed in Table 7, have been accepted by most condenser designers. While these thicknesses are recommended and establish an economic guide for average or normal conditions, they do not necessarily represent the optimum selection of plate thickness for marine condensers. The design principles described in Section VIII, Division 1, of the ASME Code for Pressure Vessels are applicable in addition to the thicknesses listed in Table 7. Stress levels for the materials used and the joint efficiency factors listed for vacuum or external pressure also apply. One exception is the pitch dimension for compression stays. Experience has shown that the 8% in. pitch dimension limit listed in the ASME Code may be exceeded and that the alternative dimension of 15 times the diameter of the stay is a safe and practical limit. Corrosion allowance is usually a matter of judgment;
MAIN AND AUXILIARY CONDENSERS Table 8 Condenser Construction Materials
Shell plate flanges, ribis, and reinforcements Tube support platea
Steel, ASTM, A-285, Grade C, Flehge Quality steel, ASTM,A-283 Steel, ASTM, A-285, Grade C, Flange Quality Steel, ASTM, A-283 Steel, ASTM, A-36 Steel, ASTM, A-7 Waterboxes and Iron. ASTM. A-278 Cast Iron, ASTM, A-278 waterbox cwers covers Cast N Nodular Iron, -&TM, ASTM, A-395 A-395 / , , cast && I;&, Steel, ASTM, A-285, Grade C, Flange Quality Steel, &EM, A-283 Aluminum Bronze. Bronze, ASTM. ASTM, B-169. E169, AUov Alloy D n Copper Nickel, 90/10,ASTM, E l 7 1 Tube sheets Muntz Metal, ASTW E l 7 1 Naval Brass, ASTM, E l 7 1 Aluminum Bronze, ASTM, E169, Alloy D Silicon Bronze, ASTM, B96, Alloy A Copper Nickel 90/10, ASTM, E l 7 1 Tubes Admiralty Metal, ASTM, Ell1 Aluminum Bronze, ASTM, Ell1 Aluminum Brass, ASTM, Ell1 Copper Nickel 70/30, ASTM, Ell1 Copper Nickel 80/20, ASTM, Ell1 Copper Nickel 90/10, ASTM, Ell1 Titanium. ASTM, B-338 NOTE: This is a re resentative list of materials suitable for marine condensers and is not intended to knit the application of other suitable materials.
Steel Plate Sbll Thicknem, in. Bf;
;
5671 to 11500 11501 to 29000 29001 and above
xs
in. allowance however, experience has indicated that a is usually adequate. f. Monitoring Condensate Purity. The need to control condenser condensate purity, because of the requirements of increasing boiler pressures or of thermonuclear steam plants, has resulted in the general use of monitoring systems in steam surface condensers. The monitoring devices are usually instruments which measure the specific conductivity of the condensate as an indication of dissolved solids concentration or measure condensate salinity as an indication of tube-to-tubesheet joint leakage or leakage through the tubes themselves. Monitoring systems designed as a means for measuring condensate impurities may also indicate the general area where leakage occurs. There are two areas where the monitoring of leakage is especially important. One of these is the tube sheet area. This zone is used for determining the soundness of tube-totube-sheet joints. The second area is that of the heat transfer tubes between the tube sheets. To monitor the tube-to-tube-sheet joints there are four methods in general use: (1) the double-tubesheet arrangement may be monitored by pressurization with condensate between the tube sheets, with leakage indicated by pressure decay; (2) an inner supporkbaffle plate, located close to the tube sheet, may be used to confine tube-to-tube-sheet joint leakage between the support-baffle plate and the inner surface of the tube sheet; (3) a catch dam may be used to catch condensate and any tube-to-tube-sheet joint leakage which drains down the inner faces of the tube sheets; and (4) a compartmented hotwell may be arranged with series flow, from initial compartments at the tube sheets through other compartments in sequence to the condensate outlet. The double tube sheet arrangement is used with marine nuclear power plants; it may be noted, however, that the high reliability of tube-to-tube-sheet welding has resulted in the almost exclusive use of welded tube-totube-sheet connections in the major applications of nuclear power. The inner support-bae plate used to confine condensate and leakage from the tube-to-tubesheet joints is effective and improves the reliability of monitoring. I t is more costly than the use of catch dams or weirs, described as method (3), but also is more reliable. Method (4), that of compartmenting the hotwell and providing series flow, is used in combination with methods (11, (21, or (3). Most designs divide the
I
I
I
I I
,
hotwell (if it is the whole condenser bottom) into four or more parts: two lengthwise and two or more crosswise of the condenser. Water flow is sequential, with the outlet from each compartment monitored, thereby indicating by the change in concentration of impurities the location and degree of leakage. With small condensers, such as auxiliary condensers, it is usually sufficient to monitor condensate quality at the condensate outlet only. 2.1 1 Materials of Construction. Materials used in marine condensers are selected to meet the requirements of strength, cost, and corrosion resistance. A number of material types and grades may be suitable for the same part or parts of a condenser, and the determining factors in making a selection should include,those criteria listed above as well as other considerations which may be controlling. Strengtfi, corrosion resistance, and in some instances weight are of major importance and may take precedence over economic considerations. Condenser shells, support plates, internal bracing, and nozzle connections which comprise the condenser shell structure, except possibly minor parts, are fabricated by welding. The materials most frequently used in the construction of marine condensers are tabulated in Table 8 and are described in the designated specifications. There are two major arem or zones in a condenser which are susceptible to significant corrosion attack. They are: (1) the steam or condensing area, comprised of the condenser shell, hotwell, tube sheets, internal baffles, piping and ducts, and the outside surfaces of the heat transfer tubes, and (2) the cooling-water side of the condenser, comprised of the waterboxes, tube sheets, and the inside surfaces of the heat transfer tubes.
The corrosion environments coincident with these two areas or zones are significantly different and the corrosion mechanisms involved, while similar in some respects, are vastly different in degree. The steamside corrosion in marine condensers is usually minimal and is seldom a major problem. The matsrials of construction, which are mostly carbon steel for'shells, support plates, baffles, and bracing, and nonferrous tubes and tube sheets, present no serious problem in this environment. The low operating temperatures, low dissolved oxygen levels in the condensate, and very pure condensate (in terms of dissolved solids) do not present a severely aggressive environment to the materials normally used. An exception, however, can occur due to carbon dioxide which, if present in significant quantity, may attack the condenser tubes, especially in the areas at the support plates and adjacent to the inside face of the tube sheets. The action may be chemical, with carbonic acid attacking the copper-base alloy tubes, or it also may be galvanic with the, carbonic acid acting as a local electrolyte at the interface between tube surfaces and support plates. The soyrce of carbon dioxide is the breakdown of carbonates and bicarbonates in the salt water evaporator which provides feedwater makeup (see subsection 2.5 of chapter 15). The evaporator temperature, point of admission of the evaporated vapor as makeup, and the provision for venting.~condenmt>lesall affect the concentration of carbon dioxide in the steam and condensate. Air cooler designs which eliminate short circuiting and stagnation of air-vapor mixtures, thereby reducing the carbon dioxide concentration, are helpful in controlling carbonic acid corrosion. Ammonia is seldom a by-product of saltwater evaporation at sea, but may be a factor with the evaporation of polluted harbor waters.
472
MAIN AND AUXILIARY CONDENSERS
MARINE EN
It may also be present in the condenser condensate if it is used for pH control in the feed system. In condenser service copper-nickel tubes and tube sheets are quite resistant to ammonia and may be considered to be a most reliable means for avoiding ammonia corrosion effects. The brasses, Admiralty and aluminum, and also aluminum bronze are susceptible to stress corrosion from ammonia, even at the characteristically low condenser temperatures. The susceptibility of these alloys to ammonia attack is reduced if the tubes, especially those in the air cooler, are not under significant tension stresses. Some control of this may be achieved by using packed tubes in the air cooler section, or by rolling air cooler tubes last, if rolled or rolled-and-welded tube-totube-sheet joints are used. Experience has shown that the corrosion of steel condenser shells and internal baWe parts is very minor and that a corrosion allowance of about in. is adequate on the steam side for the planned operating life of almost any condenser in marine service. Corrosion can be extremely severe on the coolingwater side of marine condensers if not properly controlled by design, selection of materials, and installation. Seawater, the condenser cooling medium, is an excellent electrolyte. It promotes galvanic and crevice corrosion in combinations of materials which are displaced from one another in the galvanic series. Not only must the difference in corrosion potential between materials be considered, but the relationship or ratio of their respective areas in contact with their environment must be taken into account. For example, highly corrosionresistant materials such as the nickel-copper alloys or high-nickel stainless steels may be safely used for bolting or for fasteners in combination with major parts made of cast iron or steel because the ratio of their respective areas (alloy to steel or iron) is small. The steel or iron is sacrificed by galvanic action in protecting the high-alloy materials. The steel or iron, by comparison, is so great in area that its loss through galvanic corrosion becomes insignificant. Some basic principles and recommended practices with respect to mitigating galvanic corrosion in condensers are as follows: (1) Where possible, use one material for all parts of the condenser cooling-water system, or select materials close to one another in the galvanic scale. (2) The materials used for bolting and fasteners, and in some instances welding, should be selected to be more corrosion resistant than the materials they join. (3) Provide a greater corrosion allowance for less noble materials in a galvanic system (such as carbon steel or cast iron as opposed to copper-base materials or high nickel-chromium alloys of steel). (4) The use of corrosion-resistant metals such as copper-nickel, aluminum brass, and naval brass has been an effective means for controlling corrosion. These materials resist corrosion by the formation of a durable protective oxide film on the exposed surfaces. Corrosion has also been mitigated by using ferrous sulphate to form
a protective iron oxide film on the metal surfaces. Such a film can be established by injecting a ferrous sulphate solution into the condenser cooling water circuit before the f i s t use; the solution is circulated for about three days. The coating may be maintained by providing anodes of soft grey iron or low-carbon steel which will continue the plating action on the metal surfaces in the system. If the coated surfaces are subjected to an erosive washout of the oxide coating by a high water velocity, cavitation, or suspended abrasive particulate matter, the protection can be damaged and concentrated corrosive action can result. The use of protective coatings or paints on cast iron or fabricated-steel waterboxes requires special consideration. Defects in the coating can cause accelerated corrosion where the seawater penetrates the coating. Applying the coating to the face of a nonferrous tube sheet used in combination with a cast iron or steel waterbox, and leaving the ferrous material exposed, frequently offers greater protection to the system than coating the iron or steel surfaces themselves. (5) Sacrificial anodes attached to the inside of the waterboxes are useful in protecting iron or steel surfaces and are especially needed when the ferrous material is protected by coatings or paints. Common metals used for anodes are soft iron, zinc, and magnesium. For these materials to function properly in the galvanic system they should be of high purity, except for magnesium which is frequently alloyed with manganese, or with zinc and aluminum to increase the duration of the required current levels. Iron or steel anodes have been used successfully with most systems which do not include iron or steel. If cast iron condenser waterboxes are part of the system, zinc anodes should be used to reduce the rate of wastage of the waterbox material. Zinc anodes should not be installed on systems which do not have iron or steel components; otherwise corrosion is usually accelerated. Combined with coatings, anodes improve the distribution of current and the protective coating used may be allowed to have some porosity without damaging effects (also see section 1 of chapter 14). (6) Galvanic corrosion can be accelerated by stray currents. It may be greatly reduced or eliminated by bonding. Usual practice with condensers is to metallically connect the condenser shell, tube sheet, and waterbox with a number of nonferrous metal straps, usually made of copper or high-copper alloy. Because of the low voltages associated with galvanic systems, an adequate number of bonding straps should be used (they may be spaced on 18in. to 24 in. centers around the periphery of the tube sheet). I (7) Crevice corrosion develops where there is an irregularity in surfaces, a junction, sharp bends, or other discontinuity where oxygen does not have ready access. The difference in oxygen concentration in the crevice and that outside the crevice sets up a corrosion cell. It may be manifested by corrosion within the crevice itself, or external but adjacent to the crevice. It may be avoided by design, or by the selection of materials. I n the cir-
1
I
I
culating-water system of marine condensers, the selection of materials such as 70/30 and 90/10 copper nickel for tubes, tube sheets, and waterboxes is an effective means of controlling crevice corrosion. (8) The water velocity in the cooling-water circqit is a major factor in the selection of materials in the design of marine condensers. The damaging effects of corrosion are greatly accelerated by the erosion effects of water velocity. The major concern is with the condenser tubes. They are subject to inlet-end attack and to general attack, both of which may be related to the velocity of the cooling water. The usual method of controlling this kind of attack is to select a tube material which not only is suitable for use in a seawater galvanic system, but is also compatible with the design velocity of the condenser. Table 9 lists some tube materials commonly used in marine service and the recommended design water velocity for which they may be considered suitable. Adherence to these values usually achieves satisfactory results with respect to the general corrosion of tubes but may not be satisfactory in terms of tube inlet-end corrosion. Tube inlet ends may be subject to accelerated corrosive attack should cavitation flow occur. Cavitation may result when the contraction of the flow stream a t the entrance is greater than the contraction contour of the tube at the entrance and the water flow stream leaves the side of the tubes. This causes an erosion action and wears away the protective film that the tube material develops. Thus, unprotected metal is continuously exposed and corrosion proceeds rapidly. Cavitation is increased with waterboxes of poor hydraulic design, especially those where the velocity of the cooling water entering the waterbox equals or exceeds the average water velocity in the tubes. The cavitation effect caused by the high entrance velocity of the water in the waterbox, with respect to the water velocity in the condenser tubes, can be
473
Table 9 Recqmmended Design Water Velocities Through Tubes Tube Material Water Velocity, fps Titanium 3-15 70/30 copper nickel 3-15 90/10 copper nickel 3-9 Aluminum Brass 3-7 Admiralty , 3-4 NOTE: The 3 fps lower limit is'lbecauae of heat transfer considerations. It is not a corrosion limitation for the materials listed. ., 0
reduced by special designs to absorb the velocity head energy; however, additional space is required. Where conditions, economic or otherwise, require high tube entrance velocities and high waterbox entrance velocities, the effect of cavitation may be greatly reduced by the use of tube inlet-end inserts. Tube inserts are usually made of a plastic material, although some are made of corrosion-resisting metal. They are inserted into the inlet end of -the tube, and cemented in place. They are made with a well-rounded entrance with their thickest wall just inside the entrance, forming an almost imperceptible throat. The wall then tapers gradually to a feather edge, offering a minimum discontinuity where it terminates in the tube. They are made in various lengths (about 4 to 10 tube diameters) and are selected on the basis of the extent of cavitation previously experienced or exppted. Tube-end inserts have proved to be effective %hen properly installed. It is important that they are securely cemented or wedged into place and that the feather edge is extremely thin and well bonded. Otherwise, the junction may set up its own cavitation effect downstream and transfer the damage to another tube area.
3
Section 3 Surface Cnsdenser Performance 3.1 General Considerations. The Heat Exchange through their application can be designed and built to Institute Standards for Steam Surface Condensers pro- meet the performance specified. vide a comprehensive treatment of surface condenser 3.2 Tdrms and Nomenclature. The terms and noperformance and describe levels of performance which menclature that are commonly used to describe steam oan be obtained from well-designed condensers. The surface condenser performance are as follows: aotual operating performance of condensers is usually Condenser Duty-Condenser duty is the net heat oompared for meanful interpretation with the HE1 per- transferred to the cooling water f r o u l l sources of heat formance standards. These standards do not serve as a entering the condenser. 16% also called the-hourly heat design tool except that they establish a guide for the /and is expressed in Btu per hour. application of rational design methods and procedures. Absolute Pressure-Absolute pressure is the measure of The Heat Exchange Institute Standards are suitable pressure with reference to a perfect vacuum. It is exfor determining condenser size and type and can readily pressed in inches of mercury, absolute. bo used as a basis for determining equipment costs for Static Pressure-Static pressure is the stagnation presline in economic optimization studies. Also, their valid- sure less the pressure effect of velocity. When the velocity is sufficient to assure that condenser selections made ity of the fluid is zero, the static pressure and stagnation '
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
pressure are equal. Static pressure is expressed in inches of mercury. Condenser Pressure-Condenser pressure is the absolute static pressure in the condenser shell, measured within one foot of the f i s t tubes in the inlet steam flow path, and with the distribution of measurement points in conformity with ASME PTC 12.2 Steam Condensing Apparatus. C o n d e ~ n gSteam Temperature-Condensing steam temperature is the saturation temperature of the condensing steam at the "condenser pressure" and is expressed in degrees F. Initial Temperature Diference-The initial temperature dierence is the difference between the "condensing steam temperatureJJ and the temperature of the inlet cooling water expressed in degrees F. Temperature Rise-The temperature rise is the difference between the cooling-water outlet temperature and the cooling-water inlet temperature expressed in degrees F. Terminal Temperature Dzference-The terminal temperature difference is the difference between the "condensing steam temperatureJJand the outlet temperature of the cooling water expressed in degrees F. Logarithmic Mean Temperature Diference-The logarithmic mean temperature difference is the ratio of "temperature riseJJto the Napierian logarithm of the ratio of the "initial temperature dierenceJJto the "terminal temperature differenceJJexpressed in degrees F. Condensate Temperature Depression-Condensate temperature depression is the difference between the "condensing steam temperatureJJand the temperature of the condensate leaving the hotwell. It may be either posi(heating) and is expressed in tive (cooling) -- or negative degrees F. Coolino-Water Velocit~i-Cooling (circulating) water velocitv is the average velocity of the-cooling water flowing through the h e 2 transfer tubes. It is expressed in feet per second. Heat Transfer Coemnt-The heat transfer coefficient is the average rate of overall heat transfer, from all heat sources, to the cooling water. It is expressed in Btu/hrsq fMeg F. Cleanliness F a c t o ~ T h ecleanliness factor is the ratio of the overall heat transfer of tubes in service which are
fouled, to the overall heat transfer of new clean tubes. It is dimensionless. C h s e r SurfaceCondenser surface is the surface measured on the outside of the heat transfer tubes between the inside faces of the tube sheets, including internal and/or external air coolers and that portion of the tubes within the support plates. It is expressed in square feet.
Nomenclature C = condenser tube heat transfer constant from Table 11 C' = ratio: gpm of cooling water to square feet of condensing surface, p m / s q f t D = outside tube diameter, in. Fl = heat transfer temperature correction factor from Fig. 22 Fa = tube material and wall thickness correction factor from Table 10 Fa = heat transfer tube cleanliness correction factor, selected as indicated from experience Hl = initial temperature difference : condensing steam temperature less cooling-water inlet temperature, deg F Hz = terminal temperature difference: condensing steam temperature less cooling-water outlet temperature, deg F h = heat transferred to cooling water, Btu/lb (of steam) Kl = surface constant for tube diameter and gauge. See Table 12 L = exposed length of tubes, ft N = number of tubes P = number of water passes in condenser P, = absolute static pressure in condenser steam inlet, in. Hg Q = quantity of cooling water, gpm S = exposed tube surface, sq f t Ti = cooling-water inlet temperature, F To = cooling-water outlet temperature, F Tt = cooling-water temperature rise, deg F
Table 1 1
Values of Tube Heat Transfer Constant, C [3]
Tube OD, in.
eat Transfer Constant
Tube Wall Gauge, BWG 20 1.02 1.00 1.00 0.94 0.90 0.87 0.77
18 1.00 0.97 Q. 97 0.90 0.86 0.82 0.71
16 0.96 0.94 0.94 0.85 0.81 0.77 0.67a
Corrections to 0, for the variable specific heat may be neglected. For calculations which require extreme accuracy, such as thdse associated with the establishment of condenser final designs, equation (8) should be modified to compensate for the changes in T, which result from pressure losses in the condenser and the effect of the partial pressure of noncondensable gases. This will reflect the true heat transferponditions as they exist in the condenser; see Section 4 for additional discussion on this 3.4 ~oeffiidentof Heat Transfer. The coefficient of subject.
30 Fig. 22
50 €0 70 80 90 COOLING WATER INLET TEMPERATURE, O F
40
100
heat transfek recommended by the Heat Exchange Institute is expressed by an equation which is a reasonably accurate simplification of the classical heat transfer resistance summation equation. The Heat Exchange Institute equation is as follows:
Heat transfer temperature correction factor [3]
T, = saturation steam temperature in condenser steam inlet, corresponding to P,, F U = overall heat transfer coefficient, Btu/hr-sq ftr deg F V = water velocity in tubes, fps W = steam condensed, lb/hr Y = ratio of initial temperature difference to terminal temperature difference, dimensionless 19, = logarithmic mean temperature difference, deg F
Equation (9) is limited to water velocities in the tubes of 3 fps minimum and 8 fps maximum. For velocities greater than 8 fps,'the equation usually yields high values for U. The factor FI may be obtained from the graph in Fig. 22, and Fais given in Table 10. The factor Fa is selected on the basis of experience. Values of C for different tube sizes are given in Table 11. +-
3.5
Heat Transfer Performance Standards
a. Fourier Equation. The Fourier equation for unidirectional heat transfer is used to express the hourly heat floly in terms of the hehkmtransfersurface, coefficient of heat transfer, and mean temperature difference. For condenser calculations it is written as follows:
3.3 Mean Temperature Difference. The true mean temperature difference across the heat transfer path from the temperature of the heat source to the temperature of the heat receiver is most accurately expressed as the logarithmic mean temperature difference. I t is as important in determining heat flow or required heat transfer surface as is the overall heat transfer coefficient. b. Heat Balance Equation. The heat balance equaAs commonly used in condenser calculations, the tion used for condenser calculations equates the heat logarithmic mean temperature diffexence is a close ap- given up by the heat source to the heat absorbed by the proximation of the true temperature difference between circulating water. It is usually written as follows: the temperatures of the heat source and heat receiver. The reasons for the errors are as follows: First, it is based on the condensing temperature, as previously defined, which is assumed constant throughout the condensing steam space; actually the condensing temperaNOTE: Wh in equations (10) and (11) may be from a ture reduces as steam flows through the tube banks. number,of heat sources in addition to the exhaust steam Second, it is based on the assumption of constant specific from the turbine. When considering all heat sources, heat; actually, a constant specific heat is not attainable Wh = W I ~ I Wahz . . . W.h,. The constant 512 when condensing impure vapor. is for a seawater coolant; for a freshwater coolant, the Fortunately, however, the total error is relatively small, constant is 500. and the logarithmic mean temperature difference as c. Equations Used-'&I Sizing-. Condensers. The generally used in condenser performance calculations is four basic equatioas used in sizing condensers are sufficiently accurate. The equation for the logarithmic equations (8), (9), (lo), and (11). Additional equations mean temperature difference is: derived from these and other relations of tube geometry and surface simplify the computations for determining surface, general dimensions, and other characteristics of condensers. The more commonly used supplementary equations are as follows:
+
Table 10 Values of Tube Wall Thickness and Material Correction Factor, Fp [3]
Tube Material Admiralty Metal Aluminum Bronee Aluminum Brass Copper Nickel 90/10 Copper Nickel 80/2@ Copper Nickel 70/30 Titaniumb a Estimated values. b Tentative values.
J
Table 12
Tube Diapleter, m.
% 94
Y
Values of Surface Constant for Tube Diameter and Gauge, K1 16 BWG
18 BWG
0.272 0.208 0.168 0.141
0.240 0.188 0.155 0.131
20 BWG 0.216 0.175 0.144 0.123
+
+
I
476
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
C'
=
V For values of Kl P L K ~ see Table 12
Tf = 75 F T, = 8 deg F D =%in. Tube gauge = 18 BWG Tube material = 90/10 copper nickel V = 6 fps L = 9 ft 0 in. exposed P =2 W = 7750 I b / h h = 1000 Btu/lb F8 = 0.85 (85y0 clean tubes)
P-, = from steam tables, corresponding to T,
Derive from the abbve assumptions F1 = 1.020 Fig. 22 F2 = d.90 Table 10 Kl = 0.240 Table 12 C = 267 able il To = T i + T, = 7 5 + 8 = 8 3 F
3.6 Design Criterici and Performance Standards a. Considerations in Determining Condenser Size.
The design of condensers, based upon a desired performance, is influenced by eight principal variables as follows: (1) Total hourly heat transferred, or condenser duty, in Btu, which is a function of:
The condenser surface, cooling water required, conWeight of steam condensed densing pressure, number of tubes, and cooling-water Enthalpy of entering steam less the enthalpy friction loss are computed as follows: of the condensate Heat loss or gain from drains and makeup v - = 1.389 gprn per sq ft Heat loss or gain from condensate depression C' = PLKl (2) Absolute static steam pressure at the condenser t in the inlet, (measured within one foot of the f i ~ stubes steam flow path) and the corresponding saturated steam temperature from steam tables. (3) Cooling-water quantity. (4) Cooling-water inlet temperature. (5) Cooling-water outlet temperature. (6) Cooling-water velocity through tubes. (7) Effective heat transfer surface, which is influenced by: T o t d area Number of tubes m b e diameter Tube gauge . Tube length Tube material Number of water passes (8) Service envirolllhent and maintenance; i.e. Tube cleanliness Air inleakage
All of the above factors must be considered in deter-
Waterbox inlet loss. . . . . . . . . . . . . . . . . . .0.82 f t (good practice requires waterbox inlet nozzle velocities of 0.70 to 0.80 times the condenser tube water velocity; in this case 0.75 X 6 fps = 4.5 fps was used) Waterbox outlet loss (V = 4.5 fps; 2 pass). . . . . . . . . . . . . . . . . . . . . . . . . . . . .0.15 ft Total condenser friction loss. . . . . . . . . . .l.70 ft of water
b. Establishing Condenser Size (Example 1). There are a number of mathematical procedures used in establishing the size of a condenser, depending upon the information assumed for the application. One of the more commonly used procedures is outlined in the following example for an auxiliary condenser. Assume
U
=
FIF$aCVO.' = 510.3 Btu/hr-sq ftrdeg F
wh
1892 gpm
512T,
Q = 1362 sq ft condenser surface 5 = C' -
P8
=
1
i
c. Condenser Designed for a Given Back Pressure (Example 2). When the condenser is to be sized for a given back pressure, the computation procedure is typically as follows for an auxiliary condenser: Assume Tf = 75 F D =%in. Tube gauge = 18 BWG i Tube material = 90/10 copper nickel V = 6 fps L = 9 ft 0 in. exposed P =2 W = 7750 l b / h h = 1000 Btu/lb Fa = 0.85 (85% clean tubes) P8 = 1.5 in. Hg Derive from the above assumptions F1 = 1.020 Fig. 22 F2 = 0.90 Table 10 Kl = 0.240 Table 12 C = 267 Table 11 Steam table for 1.5 in. Hg abs T, = 91.72 F HI = T, - Ti = 16.72 deg F
Q=-=
T, = Hz
Recapitulation of results Condenser surface = 1362 sq f t Cooling water required = 1892 gprn Condenser pressure = 1.450 in. Hg Condenser friction loss = 1.70 ft of water Number of tubes required = 925
v 1.389 gprn per sq f t PLKl U = E1F$aCV0.5 = 510.3 Btu/hr-sq ftrdeg F C'
= -=
1.450 in. Hg condenser pressure
DL
=
925 tubes
To determine the coolihg-water friction loss, refer Section 2, Figs. 17 and 18.
mining the she and characteristics of a steam surface condenser, and also when determining the performance Tube loss (0.325 X 9) X 2 = Tube end loss (V = 6 fps) = of a condenser on test.
3 825 DL
Hz .
Hl
= -G =
8.16 deg F
I
T, = HI - Hz = 8.56 deg F
'
= -=, 865 tubes
The condenser friction loss is the same as for Example 1, because the water velycity in the tubes, the tube lengths, and the waterbox characteristics were assumed to be the same. The amount of surface and cooling water required'change because of the difference in the two back pressures, one calculated, the other assumed. Recapitulation of results Condenser surface = 1273 sq f t Cooling water required = 1768 gprn Condenser aressure = 1.50 in. HE Condenser kction loss = 1.70 ft;f water Number of tubes required = 865 d. Determination of Condenser Performance (Example 3). In the process of testing a condenser, it is general practice to compare the condenser performance with a standard, such as those published by the Heat Exchange Institute. A procedure commonly used for determining condenser performance from test data is outlined below using the condenser from Example 1. Given
5 = 1362sqft Tube D =% gauge i n . = 1$ BWG , z2
Tube material = 90/10 copper nickel L = 9 ft 0 in. exposed P =2 Table 12 Kl = 0.240 F2 = 0.90 Table 10 C = 267 Table 11 Data from test (assumed to be typical) Ti = To = W = h =
70 F 77.9 F 7800 lb/hr 995 Btu/lb P, = 1.30 in. Hg
Determine the condenser surface, number of tubes, cooling-water flow, cooling-water velocity, and coolingwater friction loss as follows: \
+ T, + Tf = 90.63 F condensing temperature
N = -3'825
N
eke from test data T, T, Hl Hz Fl
To - Ti = 7.9 deg F 87.17 F from steam tables = T8-. &=17.174eg F = Hl - T, = 9.27 deg F = 1.00 Fig. 22 = =
-
. '
Determine the performance of the condenser as follows : Wh - 1919 gprn Q=-512T,
MARINE ENGINEERING MAIN AND AUXll.IARY CONDENSERS 1 @,,,=-I_=
HI
12.82 deg F
hl-
H2
U
=
Wh - = 444.5 Btu/hr-sq ft-deg F
e d
The value of Fa corresponding to new clean tubes is 1.00. The design value used for Example 1is Fa = 0.85 or 85 percent of clean-tube performance. However, the assumed test data show the condenser to be performing at Fa = 0.750 or 75.0 percent of the new clean-tube heat transfer rate. If the condenser had been cleaned prior to test, the 75.0 percent of new clean-tube performance could have resulted from a number of factors. First, condenser tubes which have been in service do not fully recover new
clean-tube performance after mechanical cleaning. Experience indicates that recovery after mechanical cleaning will be about 90 to 95 percent of new clean-tube values. If the tubes are acid cleaned. in addition to mechanical cleaning, recovery should be close to 100 percent. Second, poor performance may result from excessive air leakage, undersize vacuum pumps, or the combined effects of both. Third, poor performance may result from a poorly designed condenser, including an improperly designed air cooler. In general, fouled tubes (less than 100 percent of new clean performance) are always a factor and cannot be entirely eliminated without careful mechanical cleaning followed by acid cleaning. Most condensers are mechanically cleaned only because of the problems associated with chemical cleaning, and less than 100 percent of clean-tube performance during service periods is to be expected. In the actual design of a condenser, a design margin of from 5 to 10 percent (equivalent to 105 to 110 percent of new clean-tube performance) is good practice. Condensers designed with allowances of this magnitude usually show close to 100 percent performance (Fa = 1.00) if tested immediately after a thorough mechanical cleaning.
Section 4 Performance Predictions from Design Geometry 4.1 Introduction. The Heat Exchange Institute Standards recommend values for condenser performance standards which have been determined through field and laboratory test experience. These standards represent empirical values which have resulted from observation rather than from the development of theoretical considerations. The Standards list basic heat transfer values for the usual sizes of tubes used in condenser construction with the applicable factors for adjusting these values for tube material and gauge, for cooling water temperature, and for specific water velocity ranges (see Section 3). The performance of a condenser sized by the use of these standards is based on standardized heat transfer values. However, it is important to understand that these established values of heat transfer in the Standards include an adjustment for pressure losses within the condenser and therefore do not conform with normal heat transfer values established by rational means. Actually, the steam condenser is a thermodynamic device which incorporates both the principles of heat transfer and the thermodynamics of fluid flow, and both of these disciplines must be carefully considered in its design. The literature contains a large amount of data, on heat transfer which relate to water flowing through tubes and steam condensing on tubes. Generally, these data apply to single-tube condensers and they describe performance with good accuracy. If applied to multi-tube condensers, serious error will result unless other factors in addi-
tion to heat transfer are considered. The development of these factors and their application to condenser design parameters are discussed in the following paragraphs. 4.2 Modes of Heat Transfer. Three modes of heat transfer comprise the heat transfer system characteristic of shell-and-tube-type steam condensers. These modes are conduction, convection, and condensing heat transfer. As resistances, the reciprocals of their respective conductances, they are in series with the heat flowing from the higher temperature condensing steam through the tube wall and whatever fouling reeistances have accumulated on the tube surfaces to the colder cooling water. a. Condensing Heat Transfer. Condensing heat transfer, with respect to steam condensing on a colder surface, is characterized by a relatively high value of conductance. The usual method of computing condensing heat transfer coefficients for tubes is by the theoretical equation of Nusselt. Experimental data for organic vapors are in good agreement with the Nusselt prediction; for steam condensing on small tubes at low condensing rates, the conductance determined with the Nusselt equation for a single tube is about 65 percent of that empirically determined for a single tube. Experimental results with steam for a bank of 20 tubes is in reasonable agreement with the Nusselt equation for a single horizontal tube. The adjustment for additional rows of tubes is small, with the heat transfer coefficient reducing only about 10 percent for tube banks with as
c .r
many as 120 tube rows. The net effect on the overall heat transfer coefficient is minor, and the following classical form of the Nusselt equation for single horizontal tube condensing heat transfer is suggested for design where purposes: r b = resistance of tube wall referred to the outside diameter, hr-sq fedeg F/Btu ' t = tube wall thickney, in. Do = tube outside diameter, in. D, = tube inside diameter, in. where k = t h e r ~ h conductivity l of tube material, Btu/hrh, = conductance (steam to tube outside surface), q ffedeg F/ft Btu/hr-sq ffedeg F Values of roz, the resistance of the oxide film on new k = thermal conductivity of condensate, Btu/hr-sq clean tubes, are not generally available. Values which ft-deg F/ft are available will be found to vary with the source, p = condensate density, pcf largely dependent on the nature of the experimental g = acceleration due to gravity, 4.17 X lo8 ft/hrs data. Data published in the Heat Exchange Institute p = condetlsate viscosity, lb/hr-ft Standards for Steam Surface Condensers and other pubW = steam condensed, lb/hr-ft (length of tube) lished Heat Exchange Institute data on condensing heat NOTE: The physical properties of the condensate cor- transfer are possibly the most comprehensive and valid respond to the mean condensate film tempera- sources for this kind of information. The values for rozare ture. combined in the overall heat transfer datg given by HE1 b. ConductionHeat Transfer. The conduction mode but can be determined as separate values through the of heat transfer may be considered to be made up of application of analytical procedures. The values listed five resistances to heat flow arranged in series. The for rozin Table 13 have been derived from published HE1 data by such a procedure. relation is expressed as follows: The value used for r , the resistance due to fouling of the tubes in service, is comparable to the cleanliness factor generally used with the HE1 performance standwhere ards. It may be computed as a resistance from the HE1 re = resistance from outside to inside of tube inclu- standards; however, since i t h expressed as a percentage sive of all resistances e x c e ~ that t of the con- of new clean-tube heat transfer and is taken as a redensing steam and heatingAwater,hr-sq ft-deg sistance, it wi!l vary as a function of the HE1 heat transfer rate used. It is suggested that the fouling resistance F/Btu . . roz, = resistance of the oxide film on the outside of a be calculated on the basis of Admiralty metal tubes at an 8-fps water velocity with an inlet water temperature of new clean tube, hr-sq ftrdeg F/Btu roz, = resistance of the oxide film on the inside of a new 80 F. For example, when computed in this manner, for 85 percent clean tubes, rf is approximately equal to clean tube, hr-sq fMeg F/Btu 0.00023 hr-sq ftrdeg F/Btu. Values of rf derived from r~ = resistance of tube wall, hr-sg fedeg F/Btu m = resistance from fouling on the outside of the the HE1 cleanliness factor for a constant value of "percent clean" increase as the overall heat transfer upon tube in service, hr-sq ffedeg F/Btu r, = resistance from fouling on the inside of the tube which they are based decreases. The major controlling variables are water velocity, tube material, tube gauge, in service, hr-sq ffedeg F/Btu inlet water temperature, and tube diameter. General practice is to simplify equation (21) by combinc. Convection Heat Transfer. The mode of heat ing : transfer which applies to the cooling water flowing through' the condenser tubes is known as convection roz, Tom = roz and
+ rf, + rr* rr rc roz + rf + rtco =
so that
=
(22) The resistance of the tube wall is determined from the thermal conductivity of the tube material and its thickness. For thin-walled tubes there is no need to correct for tube diameter; nevertheless, the general equation for rt,', 1s: b
Table 13 Values of r., the Resistance of the Oxide Film Tubes on the Inside and OtikMe-Surfacas.of Clean Tube Material Admiralty Metal Aluminum Bronze Aluminum Brass 90/10 Copper Nickel 80/20 Copper Nickel 70/30 Copper Nickel Titanium
tor,
hr+q ft-deg F/Btu 0.000136 0.000153 0.000167 0.000178 0.000193 0.000243 0.000195
MARINE ENGINEERING
480
heat transfer. Since the flow is established by mechanical means, the type of flow is classified as forced convection. Further, the normal design velocities used maintain the flow in the turbulent range. The following classical Nusselt expression for determining the conductance of liquids being heated while flowing through horizontal tubes is recommended for determining the convection heat transfer value, for Reynolds numbers greater than 2100.
Nu
= C(Re)"(Pr)*
where
(24)
MAIN AND AUXILIARY CONDENSERS
48 1
where
U, = overall heat transfer coefficient as determined from rational methods, Btu/sq ft-hr-deg F h, = conductance of condensing steam, Btu/sq ft-hrdeg F h, = conductance of cooling water, Btu/sq ft-hr-deg F rtw= resistance of tube wall, sq ftrhr-deg F/Btu ro, = resistance of new clean-tube oxide film, sq fbhrdeg F/Btu rf = resistance of fouling in service, sq fbhr-deg F/ Btu
Re = Reynolds number = Nusselt number P r = Prandtl number
4.3 Steam Space Pressure Losses. In the process of condensing a vapor, the condensing temperature is a function of the vapor pressure in accordance with the With the exponents evaluated, equation (24) is com- saturation pressure-temperature relationship of the monly referred to as the McAdams equation, or the thermodynamic fluid involved, in this case steam. The Dittus-Boelter equation, which is as follows : steam flow from the condenser steam inlet, around the tube bundles, and through the tube banks is characterised by a pressure loss with the condensing steam temperature in conformity with the saturation pressuretemperature relationship. This pressure loss results in a where reduction in temperature in the direction of flow, and the h, = conductance (inside tube surface to water), Btu/ resulting change in temperature from steam inlet to air hr-sq ft-deg F cooler inlet affects the rate of heat transferred in the variD = inside diameter of tube, f t ous pressure zones of the condenser. k = cooling-water thermal conductivity, Btu/hr-sq The mechanism of heat transfer requires that heat fbdeg F/ft flow take place from one temperature to another lower p = cooling-water density, pcf temperature. To correlate with theory, heat flow must V = water velocity, ft/hr be based on the temperature of the steam in contact with p = cooling-water viscosity, lb/hr-ft the tubes themselves, rather than the commercially acC, = water specific heat, Btu/lb-deg F cepted definition of condenser heat transfer performance C = constant of the Heat Exchange Institute or the ASME Performance Test Code. The proposed values of C vary as follows: To simplify the analysis of the heat flow system, it can McAdams, 0.0225 be assumed that a single tube can be selected which will Dittus-Boelter, 0.0243 represent the average performance of the entire tube HE1 data, 0.0240 bundle, or bundles. Considering - such an average tube, the steam pressure at its surface is quite dieerent than For steam operating in the temperature that at the condenser steam inlet; and, consequently, range characteristic of marine power plants, c = 0.024. the steam temperature, in conformity with the saturation is recommended. pressure-temperature relationship will also be different. l-he physical properties of the cooling water spend to the mean bulk temperature of the cooling In addition, the performance standards of the HE1 and test measurements conforming to the ASME Performance rrm41.r wal,Gl. d. Overall Heat Transfer. The overall heat transfer Test Code for Steam Condensers are based on static prescoefficient U,can be calculated from the conductances sure. The condenser tube, however, senses a temperafor condensing vapor and heating water, and the total ture more closely related to total or stagnation pressure of the conduction resistances. In condenser practice than to static pressure. It is, therefore, necessary to conit is customary to calculate the coefficient U,based on sider these relationships in performing a rigorous heat transfer analysis. the outside surface of the heat transfer tubes. The heat transfer representing the contract performThe general equation for the overall coefficient U,for ance of the condenser and the actual heat transfer of steam condensers, with steam condensed on the outside of the tubes and cooling water heated on the inside of the the average tube in the tube bank are based on the same duty. The heat transfer corresponding to contract pertubes, is as follows: formance includes losses within the condenser steam 1 1 Do space and is based on a higher condensing temperature - - - +rho 70, 71 u, h8 h W a (26) than the average condensing temperature within the tube
Nu
m
H
"
(a) Tight tube spacing, low steam entrance velocities
(b) Moderate tube spacing, moderate steam entrance'velocities
-
+
+ +
I (c) Loose tube spacing, high steam entrance velocities Fig. 23
Condenaer design geometry alternatives
bank. Consequently, it is 10Ger than the heat transfer computed on the basis of the average condensing temperature in the tube bank. The two condensing temperatures can be related to condensing pressures, one representing contract pressure, and the other representmg the average pressure within the tube bank. It is deeirable to produce a design so that there is a minimum difference between the pressure representing contract condenser performance and the average pressure within the tube bank. To accom~lishthis it is necessarv to Droduce a design wherein thk internal losses are less than those included in the contract performance standard (inherent in HE1 performance). In order to illustrate the principles involved in the determination of condenser pressure losses, consider the condenstr design shown by Fig. 23. Figure 23 represents a condenser shell geometry which is based on a given available space. It is assumed that the condenser surface, water flow, and tube size, gauge, length, and number have been determined. The task remaining is to establish the optimum tu&g.arrangement. The steam flowing from the steam id& to the air cooler inlet incurs a pressure loss. This loss may be considered to be made up of two components: (1) the distribution loss, that is, the loss in pressure required to cause the steam to flow from the condenser steam inlet, throughout the length of the condenser, and around the tube bundle; and (2) the penetration loss, that is, the loss in pressure required to cause the steam to flow
482
MARINE ENGINEERING
UNITS OF PRESSURE
Loss
CONDENSER DESIGN GEoMETR~ Fig. 24
Determination of optimum tube arrangement
steam inlet lane and the loss due to the change of direction (at the hotwell) of the steam flow leaving the main steam inlet lane. The average of these three total or stagnation pressures may be considered to be a reasonably valid estimate of the average stagnation pressure of the steam entering the tube banks. The steam flow in a condenser is a compressible, variable-area flow with fluid friction and a transfer of heat. Except at the tube bank interface with the steam distribution passages, the flow is adiabatic with changing mass. It cannot be treated as isentropic because of fluid friction, nor as Fanno line flow because of area change and mass change along the flow path. In addition to the above, the flow entering the condenser is usually stratified (see Fig. 9). It has been observed that under some flow conditions, steam entering the condenser assumes a pressure gradient much like that of an expanding nozzle. To simplify the problem, it is customary to treat the steam flow as if it were one dimensional and uniform in velocity rather than stratified, incompressible rather than compressible, and adiabatic rather than nonadiabatic. Once the basic condense^ geometry is established using the above simplifications, adjustments may be made to compensate for stratification of flow. The other simplifications do not materially affect the validity of the design calculations. Since the Heat Exchange Institute Standards define the condenser design pressure as the average static pressure measured withii one foot of the first tubes in the flow path, the longitudinal pressure distribution loss is usually included in the specified condenser pressure. This is especially true with main condensers where the steam inlet area covers a very high percentage of the condenser shell area. The designer is cautioned, however, to check the center-to-end flow distribution losses; that is, the pressure loss from steam flowing around the support plates at those sections where the support plates are outside of the steam inlet area of the shell. The sum of these losses, including end flow in the condenser, should be significantly less than the distribution loss around the tube bundle. An estimate of this loss for each support plate may be made from the following equation, which is empirical and limited to area ratios greater than 0.70 in this application.
through the tube bundle, from its perimeter to the air cooler within it. The arrows surrounding the tube bundles in Fig. 23 represent the steam flow for distribution and their length represents the steam velocity. The arrows arranged radially within the tube banks represent the steam flow for penetration, and their length also represents steam velocity. Since the pressure loss is a function of the velocity of the fluid stream, the length of the arrows is indicative of the pressure loss. Figure 23(a) shows a design where the pressure loss from distribution is small because of a large flow area and low velocity, and the pressure loss from penetration is large because of the compact tube bundle and, consequently, high penetration velocity. Figures 23(b) and (c) represent modifications of the tube bundle in Fig. 23(a). Successively, the tube bundle is expanded reducing the steam velocity of penetration, but increasing the steam velocity of distribution. Consequently, Figs. 23(a), (b), and (c) represent design geometries of increasing distribution losses and coincident decreasing penetration losses. Some arrangement of the tube bundle, with given condenser shell dimensions, will result in a minimum pressure loss (distribution loss plus penetration loss) from the steam inlet to the air cooler entrance. Curves which permit a determination of the optimum tube arrangement are shown in Fig. 24. The letters A, B, and C represent the design geometries of Figs. 23(a), (b), and (c), respectively. It will be noted that the total pressure loss, the sum of the distribution and penetration losses, reaches a minimum value at a design geometry approaching that of B or Fig. 23(b). AP/support plate = By application of the preceding technique, the optimum condenser design geometry can be determined. where However, the actual values of internal pressure loss assoAP = pressure loss across support plate, in. Hg ciated with each design must first be established. The A1 = net area between support plates and condenser procedure for evaluating these losses follows. shell, sq f t a. Determination of Mean Total Pressure Around Az = net area between tube bundle perimeters and the Tube Banks. The mean total pressure around the condensch shell, sq f t tube bank is taken as a weighted-average total pressure. p = density of steam, pcf It is based on (1) the average total pressure immediately V,, = steam velocity between shell and support plate, above the tube bundle, (2) the average total pressure immediately above the hotwell, and (3) the average total fps g = gravitational acceleration, ft/sec2 pressure at the tube bundle centerline. Included in these C = conversion constant, psf to in. Hg = 0.01414 pressure determinations are the entrance loss to the main
MAIN AND AUXILIARY CONDENSERS
483
S r U M INLET
7 PRESSURE ZONE 1
&IR WM
5~hlON
Fig. 25
PRESSURE ZONE 2
Condenser pressure zones
1
PRESSURE ZONE 3 ,:
CONoENSATE OUTLET
ACWENSATE
PUMP
g = gravitational acceleration, ft/sec2 The steam flow and the resulting velocities around the C = conversion constant, psf to in. Hg = 0.01414 eupport plates must be determined by calculating the relative steam demand for each section between support plates for the entire condenser in order to properly apply In this example, all pressure losses around' the tube bundle periphery are based on the pressure in pressure equation (27). To compute the pressure distribution around the tube zone 1. The second step is to comgyte the entrance loss from bundle, the following procedure may be applied to each pressure zone 1 to the main'steam inlet lane as follows: section between the support plates and averaged as the sntrance steam pressure around the tube bundle. The cpv2 (29) difference betyeen this preeaure and the specified conAP,, (entrance loss) = 0.05 2s denser pressure when both are computed as total or ntagnation pressures represents the pressure loss for where ateam distribution. The process may be simplified without significant error, if the longitudhal pressure losses AP,, = entrance loss, in. Hgare small. The sim liication can be accomplished by V = steam velocity at entrance area, fps aomputing the losses around the tube bundle at that section of the condenser corresponding to the specified de- and other terms are as defined for equation (28). The third step is to calculate the pressure loss due to sign pressure. The procedure for computing these losses in this manner is as follows. (It is also applicable to each friction from pressure zone 1 to pressure zone 2. This is accomplished by computing the flow area at the respecindividual pressure section.) First, the steam flow area within one foot of the first tive zones and calculating the steam velocity at these tubes in the steam flow path is computed and the steam areas. In calculating all steam velocities, it is assumed velocity is determined assuming uniform unstratified that the' net steam flow reduces proportionately with flow. The stagnation or total pressure at this location, flow along flow distances 4, Lz, La, and Lq. The flow or pressure sone 1 (see Fig. 25), is computed from the areas are considered to be the distances between the tube following equation. banks at the various reference pressure zones times the condenser tube length. .+ The Reynolds number is next co&$uted-for the main tube bundle entrance area at pressure zone 1 and for the flow area at pressure zone 2, from the following relationwhere ship. .P,= total pressure, in. Hg P, = specified condenser pressw&static), in. Hg p = steam density, pcf V = steam velocity in pressure zone (pressure zone 1, Fig. 25), fps where
P
484
MAIN AND AUXILIARY CONDENSERS
MARINE ENGINEERING
where
AP,, I
= mean total pressure loss around tube bundle. 3
APla\=
APu
=
APa-2 = =
AP,* =
AP,
=
and where
in. Hg A = & ( P - D~)L, pressure loss at zone 1, flow path L1, (generP ally zero), in. Hg. Pressure at location 0 , in pressure zone 1 is to be taken aa P. where from equation (28). A = entrhnce area to tube bundle, sq f t prewure loss from zone 1to zone 2, flow path L = perimeter of tube bundle, f t L2, in. Hg P = tube pitch (for variable pitch use equivalent pressure loss from zone 2 to zone 3, flow path value), f t L,,in. Hg Do = tube diameter, f t pressure loss from zone 3 to Zone 4, flow path Lt , condenser tube length, ft Lr, in. Hn With this flow area the steam velocity entering the entrance loss, zone 1 to flow path L2, in. Hg, periphery of the tube bank is computed 'as equation (29) exit loss, from end of flow path La to zone 3, in. Hg, equation (29), except the coefficient becomes 0.25 (approx.) / where
-
V = steam velocity, fps W = steam condensed in tube bundle, lb/hr
REYNOLDS NUMBER: R.= Fig. 26
p =
Friction factor for steam distribution losses
p = steam density, pcf Vl = steam velocity, ft/hr p = steam viscosity, lb/hr-ft D, = equivalent diameter of flow area (i.e., the ratio of flow area to wetted perimeter multiplied by 41, ft Using the mean of the two Reynolds numbers for zones 1 and 2, refer .to Fig. 26 and determine the friction factor, f, on the basis of a roughness equal to that for smooth tubes. The pressure loss due to friction from pressure zone 1 to pressure zone 2 can be calculated using the Fanning equation as follows (the entrance loss from equation (29) must be added to the associated A P , in this example AP9-1) :
p = =
C
density of steam, pcf conversion constant, psf to in. Hg
=
0.01414
This procedure is repeated to determine the frictional pressure loss from pressure zone 2 to pressure zone 3, again taking into account the reducing mass flow along La. It is repeated along L4 to the centerline of the tube bundle where the terminal flow and velocity is assumed to be zero. In this case the flow area is calculated as the distance from the tubes to the maximum hotwell level (assutning no other obstructions) times the condenser tube length. An exit loss due to the change in flow direction from . the main steam lane to the underside of the tube bank must be added. For this purpose, equation (29) may be used with a coefficient of 0.25 instead of 0.05. Such losses, which are entrance and exit losses, are added to the friction losses calculated at the various zones of the main steam distribution lane; i.e., at pressure zone 1-2 (flow L2) and pressure zone 3 4 (flow path Ld) which are path where illustrated by Fig. 25. The process for computing the mean total pressure loss APa-l = pressure loss from zone 1 to zone 2, in. Hg around the tube bundle involves averaging the losses for f = friction factor from Fig. 26 selected incremental flow distances. This may be done L = flow length from zone 1 to zone 2, ft V = average velocity between zone 1 and zone 2, fps by graphic integration, by algebraic averaging, or by D, = average equivalent diameter (i.e., the ratio of other convenient methods. An acceptable result may be flow area to wetted perimeter multiplied by obtained by using the following equation, which is based on a three-pressure-zone analysis with flow directions 4) between zone 1 and zone 2, ft and distances as shown in Fig. 25. g = gravitational acceleration, ft/sec2
The preceding example illustrates the principles used. The specific treatment of flow and geometry in the example should not be considered to be limiting. Many simplifying assumptions were made and improved accuracy is possible by the application of more rigorous techniques. It is recommended that after a condenser is selected through the procedures described, a more rigorous check of APdm be made by applying the principles to the complete tube bundles and making an allowance for stratified flow a t the pressure measurement zone, pressure zone 1. b. Determination of Pressure Loqs in Tube Banks. For the calculation of the pressure loss through tube banks, the actual bank is replaced by an equivalent rectangular one having the same perimeter and number of tubes. The equivalent number of rows in the depth of such a tube bank may be computed from the equation
steam density at the mean static pressure around the tube bank, pcf
VI, the steam velocity in ft/hr, is used in calculating the Reynolds number. The )Reynolds number may be calculated from equation (30) modified as'
where Do = outside tube diameter (ft), and all other terms are as described for equation (30) at the temperature corresponding to the mean pressure at the tube bank entrance. The friction factor is determined by entering Fig. 27 with an adjusted Reynolds number equal to one half of the value calculated from equation (36). The tube bank pressure loss equation, modified for a uniformly decreasing mass flow, is [9]:
where
where
\
N o = equivalent number of tube rows N = total number of tubes in tube bundle being considered
P
= tube pitch (for a variable pitch use an equivalent
.L
= perimeter of tube bundle, ft
value), ft
The next step is to calculate the velocity through the minimum flow area between the outer tubes. The flow nrea may be calculated from the equation
AP, = pressure loss through tube bank, in. Hg f' = dimensionless fgsion factor (Fig. 27) . N, = equivalent number of tube rows - , p = mean density of steam at tube bank entrance, pcf V = entering steam velocity between tubes, fps g = gravitational acceleration, ft/sec2 The mean effective pressure loss through the tube bank is used to determine the mean condensing pressure in the tube bank and its corresponding temperature, which
486
MAIN AND* AUXILIARY CONDENSERS
MARINE ENGINEERING
487
determined by first determining the hourly heat or duty terms of total resistance or in terms of overall heat transper square foot of condensing surface (Btu transferred fer. It is more meaningful if the fouling resistance exper hour square foot). _The mean temperature of the pected in service is excluded. It is expressed as condensate film is equal to the product of the unit hourly heat and the sum of the following resistances from Fig. 2 7 Friction factor for steam penetration (tube bank) louer
is used to calculate the basic heat transfer. The mean transfer coefficient as related to the fluid flow characterpressure loss in the tube bank, with units of in. Hg, is istics, the procedure delineated in Section 4.3 may be used to calculate P,,, the mean total pressure correspondcomputed from ing to the average tube representing the tube bundle design conditions based on P,, the mean condenser total pressure. The temperature corresponding t o P, is the In order to calculate the pressure in the tube bank re- condensing temperature used in determining the heat lated to AP, the following procedure is suggested. All transfer necessary in,the tube bundle to meet the conheat transfer calculations in this procedure are based on denser performance requirements. Using the temperature corresponding to P,, to detertotal pressure and, therefore, the mean condensing presmine Om, the heat transfer coefficient can be calculated sure, P,,, in the tube bank representative of the conusing equation (10) of Section 3, modified as follows: denser as a unit is referred to the design condenser pressure plus its velocity pressure, i.e., P,. P, is calculated as follows:
P,
= Pt
- APhn - AP,
(39)
where the mean total pressure corresponding to the average tube representing the tube bundle design conditions calculated with P, as the reference, in. Hg mean condenser total pressure (see equation 28)) in. Hg mean distribution pressure loss (see equation 32)) in. Hg mean penetration pressure loss (see equation no\
.- n g
JUJ, 111.
TT-
4.4 Performance Prediction and Evaluation. Outlined in the preceding aragraphs are two elements of a procedure which may Fle followed to determine if a selected condenser design geometry is adequate to meet the condenser performance requirements. The first consists of an investigation of the heat transfer relationship as discussed in Section 4.2; and the second procedure involves an analyk of the fluid flow and ressure loss relationships as described in Section 4 . 1 A method of combining these two procedures in order to evaluate the performance of a condenser from its design geometry is outlined as follows: a. Heat Transfer Coefficientas Related to Characteristics of Fluid Flow. In order to determine the heat
In this instance, U, = heat transfer for the tube bundle related to pressure loss considerations, i.e., the mean bundle pressure, P,. All other terms are as geherally defined for equation (10). The value of U,, as computed from equation (40) may not include a cleanliness factor, except as the value of S, the condensing surface, reflects any cleanliness factor used in the determination of condenser size. b. Heat Transfer CoeflBcientas Computed by Rational Equations. The procedure for computing the heat transfer coefficient by rational means is delineated in Section 4.2. First the conductances for condensing and convection heat transfer are determined and converted to resistances. To these resistances, add the resistance due to conduction heat transfer; namely, that due to tube wall resistance, inside clean-tube resistance, and outside clean-tube resistance. If a cleanliness factor was included with equation ( a ) , the resistance due to fouling, calculated from this cleanliness factor, must be included here. Convert these resistances to UrJthe rational heat transfer coefficient representing the tube bundle performance, using equation (26). Note that in calculating U, the physical properties of the two fluids, condensing steam and heating water, must be evaluated at their correct temperatures, see equations (20) and (25). The mean temperature of the condensate film may be
oquaiion (26),(&
.- -
+ r, + r, + r i), ,added to the
mean bulk temperature of the cooling water. Using t,his temperature to cdculate h, from equation (20)) caluulate U, from equation (26). c. Performance Evaluation. By comparing the values of U, the heat transfer coefficient related to presaure loss, and Ur, the rational heat transfer coefficient, tbe adequacy of the condenser design can be determined. To meet specified performance requirements, the following relationship must exist: The greater the difference, the greater will be the design ~narginor excess performance characteristics of the condenser. U, represents the heat transfer needed to meet porformance requirements, U, represents the heat transfor attainable. Should the relationship shown in equalion (41) be the reverse, the condenser will not meet its performance requirements and other condenser configmaIions should be investigated. Up, can be significantly ahanged by modifications in the condenser design geometlry; Ur is only slightly affected by the condenser de~ i g ngeometry. If a correction cannot be accomplished by changing the tube bundle design, the condenser shell would normally be the controlling factor; in such an instance, it should be made larger. In some respects it is more convenient and more underntandable to use resistances rather than conductances for making the comparison of performance. In this case,
The PF should be greater than unity, and it is considered good practice to design for a performance factor of not less than 1.05. The predicted operating pressure for the condenser based on its d'esign geometry and on rational heat transfer can be determined using the previously calculated mean tube bundle pressure, PmJ as the base. Using equation (39)) cialculate a new value for P,(Udm and AP,, as previously determined are assumed to remain constant). Using equation (28) determine P,, the static pressure the condenser will produce in the pressure-measurement zone as defined in the H.E.I. Standards. As a check, P, will be less than the H.E.I. specified condenser pressure, provided Ur > Up, as required by equation (41). The methods and procedures outlined represent a reasonably accurate method for evaluathg condenser design geometry and performance. Accuracy can be improved by a more rigorous treatment of the fluid flow chaxacteristics. The degree of accuracy improvement necessary depends on the circumstances of a particular case. While the descriptions presented have been related to the normal geometry of the larger main condensers having two separate and distinc$kube bundles, the method is also applicable to single-bundle condensers such as those generally used for auxiliary condensers. The principles can also be adapted to condensens having more than two separate tube bundles or to any shell-and-tube apparatus which condenses a vapor on the shell side. References
where
1 Courtesy, Foster Wheeler Corporation. 2 Courtesy, Worthington Corporation. R r = -1= overall resistance frdm rational equa3 Courtesy, Heat Exchange Institute. tions, hr-sq ft-deg F/Btu ur 4 Courtesy, American Society of Mechanical Engi1 overall resistance from fluid flow char- neer, R ~m = -u,, = acteristics, hr-sq ft-deg F/Btu 5 J. E. Fowler and R. E Brandon, "Steam Flow DisIf fouling resistance is introduced (it must be con- tribution at the Exhaust of Large 'steam Turbines," lridered to be the same for both the resistance calculated ASME paper 54SA-62. by rational means and by fluid flow characteristics) equa6 J. P.&bald and W. D. Nobles, "Control of Tube tion (42) becomes Vibration in Steam Condensers," Proceedings of the American Power Cmference, ~llinbis1nstitute7of ~ e c h (43) nology, Chicago, Illinois, 1962. t~ndmay also be written as 7 -A. P. ~ i l b u r n "Problems , in Design and Research on Condensers of ~ a ~ o u i k T a nVapo3 d MixtgesJJJThe ' Rdm = (R, ry) - (Rr rf) (44) Institution of Mechanical Engineers, 1951. 8 "Standards for Steam Surface Condensers," Heat where Rdm= design margin as a resistance, hr-sq ft-deg lp/Btu, and must be positive if the condenser is to meet Exchange Institute, New York. 9 W. H. McAdams, Heat Transmission, McGrawi t ~required l performance. The performance factor PF may be expressed either in Hill Book Co., New York, 1954.
+
+
t
HEAT EXCHANGERS
489
CHAPTER X I V
Charles D. Rose Philip Liu
I Heat Exchangers Section 1
1.1 General. Heat exchangers are used throughout the marine power plant to transfer heat from one fluid (liquid or gas) to another fluid. The most widely used type of exchanger in marine service is the "shell-andtube" type. As shown by Fig. 1, the shell-and-tube heat exchanger consists of six basic elements: the bonnet, tube sheet, shell, tubes, baffles or support plates, and tie rods. The bonnet or channel is often referred to in the marine industry as the "head" or "waterbox" of the exchanger. Due to stringent space limitations imposed on marine heat exchangers and for ease of maintenance and cleaning, head inlet and outlet connections are arranged to permit access to the tubes and tube sheets without dismantling the attached piping. Only the smaller heat exchangers (under 100 lb) would have straightaway connections as depicted in Fig. 1. Figure 1 illustrates a "single-pass" fixed tube sheet exchanger. "Single-pass" is a term indicating that the tube-side fluid flows in one direction only. A "two-pass" or "four-pass" exchanger would have the tube-side fluid inlet and outlet connections a t the same end. The fixed tube sheet construction depicted in Fig. 1 (tube sheets welded to the shell) must often incorporate an expansion joint in the shell to compensate for differential expansion between the shell and tubes as a result of the relative temperatures of the fluids involved. The various means of providing for differential expansion are covered in Section 1.2. I n addition to fixed tube sheet designs, heat exchangers employing floating tube sheet and "U"tube designs are commonly used. Figure 2 illustrates the floating tube sheet construction which is employed in main lube oil coolers and electronic equipment coolers. In this design, one tube sheet is free to "float" against packing rings (usually made of neoprene). The packing rings are held in place between the head and shell flanges by a packing retainer ring. Details of this construction are described in Section 3.1 of this chapter. Zinc protectors, such as those shown adjacent to the division plate in the head of the heat exchanger in Fig. 2, are used in saltwater-cooled heat exchangers to protect headers, tube sheets, and tubes from galvanic corrosion by electrolytic action. When dissimilar metals that are connected together are immersed in an electrolyte, a simple galvanic cell is formed and an electric current
BONNET
TUBE SHEET
BAFFLE OR SUPPORT PLATE
TIEROD
TUB~SIDE NOZZLE Fig. 1
Typical single-pass conventional exchanger (flxed tube sheet)
flows from one metal to the other through the electrolyte which completes the circuit between the two dissimilar metals. The metal (anode) from which the current flows will tend to suffer rapid corrosion, often termed galvanic or electrolytic corrosion, and the metal (cathode) to which the current flows will tend to be protected from galvanic corrosion. The direction in which the current flows depends on the composition of the metals or alloys exposed to the electrolyte and also on the hardness of the metal, the cleanliness of the metal surfaces, and other factors. Thus, if a single metal is immersed in an electrolyte and one part of the metal surface is harder than another part, or cleaner than another part, there will be a flow of current from one part to the other and galvanic corrosion will take place. If several different metals or alloys are involved, current will flow in varying proportions between the surfaces exposed to the electrolyte. If clean metallic zinc is properly arranged within a heat exchanger waterbox, a current will tend to flow from the zinc to the adjacent metal surfaces exposed to the seawater which constitutes the electrolyte of the galvanic cell. The zinc protector plates are corroded as the current is generated; and the current flowing through the seawater to the metal surfaces of the heat exchanger tubes, tube sheets, and waterbox tends to protect these parts from galvanic corrosion. The elechc circuit is completed through the metal parts of the heat exchanger. Gaskets between the waterbox and manhole covers, to which zinc protectors are frequently attached, and
Fig. 2
Typical heat exchanger with floating tube sheet, disassembled
between the tube sheets and the water chests do not interrupt the flow of current, as the circuit is completed through the metallic bolts and collar studs which secure the joints. The Navy requires that the exposed surface of sincs (exclusive of edges) be a t lea& one square foot for each 1000 square feet of heat transfer surface. For now construction, the amount of zinc surface is based on rtn equation which is given in Military Specification MIL-A-19521. All zinc protectors should be thoroughly scaled once every 4 to 6 weeks to assure that active metallic zinc nurface, as opposed to corrosion scale adhering to the metal, is exposed to the seawater. Zinc surfaces which aflord proper protection are quickly detefiorated. In the design of the waterbox and application of zinc modes, care must be exercised that the zincs do not interfere with, or add turbulence to, the fluid flow within the waterbox. As illustrated in Fig. 1, the shell of the heat exchanger is usually cylindrical with flanges attached to each end. The tube sheets are either welded or bolted to the shell flange, and the heads are bolted to the tube sheets. The most critical joint in the exchanger, and that most likely to develop a leak, is the tube to tube sheet joint.
I
SERRATIONS Fig.
9
Expanded tube-to-tube-sheet joint with inlet end bell
Expanding the tube into the tube sheet with a "tube expander" is the most commonly em~loyedpractice; a typical expanded tube is &wn by Fig. 3. Care must be taken to follow the manufacturer's instructions as to limits of tube wall reduction and dial settings on the electrically or electronically controlled automatic tube expander when repairing a leaking tube to tube sheet joint. Seawater flowing into heat exchanger tubes a t high velocities tends t9 remove the thin protective film of
MARINE ENGINEERING
HEAT EXCHANGERS Table 1
T* Fluid
ITEM side
Velocity, fps
LWBICA~G OIL COOLER
Low-paEssmr~1 F ~ HEATER D
HIGH-PEEBSUBE FEEDHEATEB
Seawater
Feed
0-7
4-7 2-8
Typical Featuror
-
TANKCLEANING SYBTEM --HEATER DRAINCOOLER
5/811 OD
Tube Material H a d Material
per inch 90/10 CuNi Bronze
90/10 CuNi Steel
70/30 CuNi Steel
90/10 CuNi Steel
Tube Sheet
90/10 CuNi
Steel
Forged Steel
Bronze--Composition G
SluU Side
Fluid Velocity, f s Number 6 P Pas= Type of b d e
Flow
Cbnslrvdion
5/8" OD
x 0.049"
Steam Condensing 6 (baaed)
Oil
2-3
10-16 (baaed)
Expanded
m d e d
Expanded or welded
Thermal Ekpansion Gasket
Floating tube sheet Full Face
Expausion joint
U-Tubes
corrosion products adhering to the base metal of the tube wall. This protective a m is replaced a t the expense of further corrosion of the tube wall. As continued removal and replacement of the protective film of corrosion proceeds, the tube wall is gradually thinned and the tube to tube sheet joint is weakened and ultimately fails or the tube wall just beyond the tube sheet is perforated. This type of erosion is generally termed impingement erosion, inlet end attack, air erosion, or bubble attack. The occurrence and rapidity of the attack are governed by the water velocity (for recoinmendations see Table 1)) the amount of air entrained, and the design of the waterbox as it affect8 the velocity, ,direction, and turbulence of the fluid flow approaching the tubes. The inlet ends of the tubes are normally ground flush with the face of the tube sheets, and no gaps should be left between the edges of tubes and the radius of the holes on the inlet sections as gaps here tend to promote impingement erosion. The outlet ends of tubes may
Collar Bolts Bare
600
2.7 gpm seawater per sq ft
190°F 110°F rise in
200 180 lb drains per hr/sq ft
100°F 200°F drop in d r u i ~ ~ ~
extend Ne in. beyond the face of the tube sheet. Where s considerable temperature differential exists withim a heat exchanger, a packed-tube design may be employed. Packed tubes, such as shown by Fig. 4 and Fig. 8(c), allow considerable differential tube expansion since each tube is free to move independent of the others. The combined low-pressure feed heater/drain cooler/gland-exhaust condenser (which is described in Section 3.5) is a typical marine exchanger normally employing packed tube ends. Marine heat exchangers are designed in accordance with the Standards of the Tubulas Exchanger Manufacturers Association [I]' and the American Society of Mechanical Engineers [2] in addition to compliance with marine regulatory body wde requirements such as Lloyd's [3], U.S. Coast Guard [4], and the American Bureau of Shipping [5]. Many additional standard 1
Condensing
2
CuNi
I
{ lrrt~dsteam !rb~dst~sing fi rlr H (baffled)
Full Face
Ring (shell aide) Solid co per ring (tube d e ) Stud Bolts Bsre
10°F 1 0 0 9 rise in feed
i'i (bafled)
Steam
U-Tubes
Ring or flexible
800 3 0 0 lb feed per hr per sq f t
Heater dr-ains Segmental Counterflow, last pwa
...
Tube Joint
Bolting Collar bolts Surfaca Finned Therm41 PerfOnna?ca Heat T r a d e r Coefficients ~tu/hr-fd 40 0.3 to 0.4 gpm L.O. Surfaca Requirement per sq ft Terminal Temperature Diffknce (or LMTD) 18°F Temoerature 20°F d r o ~in L.0.
Steam Condensing 6 (baaed)
Segmental
Segmental counterflow
ELECTBONICB EQUIPMENT COOLER
Saltwater 0-7 1-4 3/4" OD x 0.049"
1 R ~ 2 ~ 5/8" 7lowODfin-19x 0.049" fins
x 0.049"
w1 Marine Heat Exchangers
Numbers in brackets designate References at end of chapter.
8;;h l t a
Steam
steam'
Condensing
Condensing
2
2
(bayonet tube) 90/10 CuNi Steel Steel or Cast Steel 90/10 CuNi BronzeSteel Composition G or steel Fresh Water
Chilled Water or Seawater
Boiler Makeup Water
1
:;"
34
5/8" OD X 0.049 (bayonet tube) BWG ,. Steel 90/10 CuNY Cast Steel Bronze 4
Water
1/4" OD X 0.035"
4-7 2 5/8" OD XO.049" BWG 90/10 CuNl 90/10 CUNI
/
Steel
90/10 CuNi or Bronm Composition G
...
90/10 CuNi
Steam Condensing
...
Open Shell
Lubricating Oil 2-3 1 (baf8ed)
Fuel Oil 1-3 1 (baffled)
Fresh Water 3-5 1 (baffled)
Seawater 1-8 None
None Boiling
Segmental Counterflow
Segmental Counterflow '
Segmental Counterflow
None , Counterflow
Expanded
Expanded
Floating head or U-Tube , Full Face
Bayonet
Expanded Expanded (buter) Ferrules (inner) Bayonet ...
Stud Bolts Bare
Stud Bolts Finned
10-20 60 6-8 gph 75 gph
1 (baffled)
Welded or brazed
...
U-Tubes
Ring
Full Face
Ring
Stud Bolts Finned
Stud Bolts Bare
Stud Bolts Bare
Stud Bolts Bare
45 7-12 gph per sq ft
400 150 h per Sq t
500 0.25 m boXr water per sq ft
150 hper sq%
100°F 100-150°F
20°F 5°F drop in fresh water
400°F 400°F drop in sample temp.
200°F 80°F rise in water temp.
Ring
, ,= 6-10 111 team
&r Irr/sq ft
8% nonucr~~tlansable
%-50 lb steam per hr/sq ft
80°F Boiling a t 100 pa aba
P Y = #?sq 120°F
250°F 50°F
T
500
Heads are designed so that it will not be necessary to disassemble piping to gain access to the inside of the heads The and tubes. tube bundle is usually of the removable type for easy cleaning and maintenance. I n the design of cooling water spaces and connections, a smooth flow path must be provided to minimize erosion-corrosion attack. Sharp comers and projecting edges are avoided. Internal fittings are arranged to result in a minimum of interference with the water flow and a minimum of turbulend? Y ' Rg. 4 Packed tube Cooling water velocities a t the design point must not exceed those specified or recommended by the material supplier. requirements have been incorporated in specifications for Heat exchangers having tubes of length exceeding marine heat exchangers due to stringent space limitations 4 f t are designed so that the o r d e ~ n glength of tubes and reliability requirements. The following points are will be in multiples of 6 in. The ordering length of tubes is determined by addbg M in. to the face-to-face emphasized in the design of marine heat exchangers:
MARINE ENGINEERING
HEAT EXCHANGERS
(g) Head integral with shell, no shell side
(a) Flat face ring shell flange with full face unconfined gaskets on both shell and head side.
(d) Welded shell tube sheet joint with semiconfined gasket on head side.
gasket.
support plates or between a tube ahget and a support plate does not exceed 36 in. Holes for tubes in bases and support plates, baffle clearances, and tie rod standards are usually required to be in accordance with the latest standards of the Tubular llxchanger Manufacturers Association [I]. In order to diffuse the entering stream and reduce erosion of tube ends, for single-pass coolers the waterbox depth measured normal to ihe-tube sheet should be not less than one-half the equivalent diameter of the tube sheet area exposed to the flow of the cooling water into the tubes. For cylindrical two-pass coolers, the waterbox head depth ~houldbe not less than 35 percent of the inside shell diameter. All heat exchangers must be provided with tuiequate foundation supports. When required by the oonditions of service, provision is made in the design of the supports to provide for expansion or contraction of the shell. Heat exchanger supports are usually independent of any att,ached piping. Supports must be given special consideration when designing for high-impact shock conditions [6]. Is the design of marine heat exchangers, consideration must be mven to the varvin~deerees of inclination encountered in service. In n a v z prLtice, heaters and ooolers are designed to perform satisfactorily under oonditions of 5 degrees trim, 10 degrees pitch, 15 degrees list, and 45 degrees roll (in commercial practice a 30degree roll is the design criterion). The conditions for permanent list and roll or for trim and pitch are not bonsidered additive. Adequate air vents must be provided on heat exchanger waterboxes to avoid the collection of air in the upper region of the waterbox, as air pockets can restrict the tube-side flow and render a portion of the heat transfer surface ineffective. Such air pockets can also result in overheating and expansion of the dry tubes and oause failures of the tube joints at the tube sheets. In feedwater heaters and condfnsers, wet steam at a high velocity must not be permitted to impinge on the tubes, otherwise the surface of the tubes will be rapidly eroded. ~ a f h e sor distribution pipes must be incorporated as necessary to preveht the direct iinpingement of wet steam on the tubes.
n , 00000 000000 0000000 000000 Seg,,n*l
(e) Double packed floating tube sheet with retaining ring and packing rings on both hell and head side.
(h) Double floating packed tube sheet with retaining ring and packing rings on both the shell and head sides.
7
(c) Ring shell flange with tongue and groove joint and fully confined flat ring gasket on both the shell and head side.
If] Outside-packed removable tube bundle with semiconfined gasket on head side. Fig. 5
(I) Ring joint flange, double tube sheet design with confined O-ring gaskets.
Shell, tube sheet, and heod ioint design
distance between the outside faces of the tube sheets. The minimum tube sheet thickness is usually specified to be not less than 31 in. When external fins (low fins) are applied to tubes, one end of the tube is usually gradually enlarged to the outside diameter of the fins to enable the removal and insertion of individual tubes. Holes in the tube sheet at the inlet end of the tubes are flared to allow for belling the ends of the tubes. Holes in the tube sheets are provided with at least one groove. The edges of the holes are rounded, usually on a
He-in. radius, on the inner face of each tube sheet and on the outer faces of the tube sheets at the discharge ends of the tubes. The inlet ends of the tubes are expanded and belled and the ends are finished flush with the face of the tube sheet. In no case should the ends of tubes be below the face of the tube sheets. Discharge ends of tubes can protrude up to He in. beyond the face of the tube sheet. a A number of the baffles are increased in thickness (usually H in.) to act as tube support plates and are located so that the maximum tube length between
1.2 Shell, Tube Shed, and Head (Channel) Joints.
The design of the means for attaching the head (channel), tube sheet, and shell are governed primarily by the operating pressures and operating 'temperatures to be ttccommodated. Figure 5(a), ' which employs gaskets, depicts the least costly means. Gaskets (which are usually compressed asbestos) aiqepositively positioned, thereby insuring alignment. This design could incorporate studded tube sheets (shown) or collar studs to permit removal of the head without disturbing the shellside pressure joint. In naval applications, the joint shown in Fig. 5(a) may be used with up to 150 psig and 375 F. Figure 5(b) illustrates the joint that is the most widely used in shell, tube sheet, and head joint attachments.
BaMe
(dl Triple Se~nental
P
(bl Single W m n t e l
-
-
(b) Flat face ring hell flange with ring type unconfined geskets on both shell and head side.
, l"
.
,
'(el D i x 81 Doughnut
(c) Double Segmental fig. 6
Fbw baffler
The design normally incorporates studded tube sheets (shown) or collar studs. The recommended gasket material L coinpressed asbestos, and the usual a plications will accommodate liquids and vapors to 3 psig and 450 F. The design indicated in Fig. 5(c) is widely used for high-pressure fuel oil and steam service. The gasket grooves completely confine and positively align pass partitions, and afford excellent protection against gasket blowouts and failures. The usual applications permit ' liquids and vapors to 1000 psig and 750 F with a proper gasket material and design (normally compressed .asbestos, jacketed asbestos, or spiral-wound gaskets). Figure 5(d) depicts a joint that is used primarily when no leakage is permitted o p h e shell &de, as in cases where the shell-side medium is either hazardous er Comsive. The semi-confined gasket allows for protection from blowing out and provides a more positive positioning of the gasket. The usual applications permit liquids and vapors to 600 psig and 500 F with proper gasket materials (normally compressed asbestos, jacketed asbestos, or spiral-wound gaskets). f i e design indicated in Fig. 5(e) is an excellent means
L
494
HEAT EXCHANGMS
MARINE ENGINEERING
(4Square tube pitch Fig. 7 ' Tube pitch
(a) Triangular tube pitch
Id) Floating head with backing device
{b) Bayonlrt Tube Fig 8
(d) Rotated square tube pitch
(b) Rotated Triangular tube pitch
ggmpatible fluids; should a tube joint fail, the leak can be detected immediately, avoiding contamination of the guids. The usual applications permit liquid and vapor prassures up to 1000 psig and 400 3 '. 1.3 Spseifle Construction Detuils. I n shell-andCube heat exchangers, baffles are generally used to guide flow and increase the velocity of the fluid flowing on the rrhell side of the heat exchanger. The most commonly uaed baffles are of the segmental type or its variations, as Illustrated by Fig. 6. Segmental baffles are formed by cutting out thin metal plates to an outside diameter slightly less than the b i d e diameter of the shell. A segment is cut out of the baffle to form a segmental opening, the size of which may vary from approximately 15 to 45 percent of the nhell cross-sectional area (with single segmental, double regmental, disk, and doughnut ba&) or higher than 46 percent of the shell area (with triple segmental baffles). The tubes must be supported at intervals dong their length to minimize tube vibrations excited by the fluid flowing across the tube or by pulsations of the flow rate. The maximum permissible unsupported tube longth will depend on the tube material and tube diameter. For nonpulsating flow, the maximum unsupported tube length is 60 to 80 times the tube diameter. For pulsating flow, the unsupported length is uonsiderably less, and welding or brazing is used in lieu of spacers and nuts to secure the tube support plates. When designing for high shock, the tube support spacing $
of providing for shell or tube thermal expansion. When the packing on one side becomes deteriorated and eventually develops a leak, the leak is readily detectable through the vent and drain holes in the retaining ring and the packing can be replaced. The studded retainer ring permits repacking the tube side with full p r e m maintained on the shell side. The usual applications permit liquids axid vapors to 300 psig and 300 F (the temperature is limited by the packing material). The design indicated in Fig. 5 0 is an alternative means of providing for shell or tube t h e d expansion. On the she11 side, liquids and vapors at pressures up to 500 paig can be accommodated but the temperature is limited (usually to about 300 F) by the type of packing. On the tube side, the pressure may reach 600 psig at temp&rattwesof 500 F with proper gaaket materials. Figure 5(g) describes a joint which is relatively inexpensive and is used when it ia desirable to eliminate the shell-rside gasket. This design employs only one
gaaketed joipt at each head end for servicing the tubes. There are no limits on pressure or temperature except for gasket considerations. The design indicated in Fig. 5th) is recommended where contamination of one fluid medium by the other cannot be tolerated and provides for excellent thermal ex~ansionof either the shell or tubes. The studded r e k n e r ring permits repacking the tube side with full pressure maintained on the shell side. The double tube sheet type of construction is often specified by the Navy in cases where the greatest assurance against mixing of the two fluids is desired. Liquids and vapors to 300 psig and 300 F (with due consideration to packing material) can be accommodated with this design. The joint described by Fig. 5(23 is excellent for vacuum and very-high-pressure service, and the temperature iR limited only by the O-ring gasket material. The design requires precise machining of the O-ring grooves. Tho double tube sheet design-is excellent for use with non-
T h m o l expomlon provisions
is made small to minimize thd%esponseof the tubes to shock loadings. There is considerable latitude in selecting the pitch for tubes. The four most common tube patterns, as viewed from the tube sheet end,, am shown in Kg. 7. The triangular pitch and rotated triangular pitch are the most compact forms, and the triangular pattern is the one most commonly used for marine heat exchangers. Square pitch and rotated-square pitch pattern have see-through lanes which facilitate manual or mechanical cleaning of the outside of the tubes. The square pitch is common in submerged-tube boilers where passage for fluid circulation is important. It is a h applied in s e ~ c e whme s minimum pressure drop is a paramount design criterion. Tube patterns other than' the four illuatrated in Fig. 7 could be used to dtisfy specific design co%iderations as to pressure drop or turbulence, but they would be more costly to manufacture. Tube center-to-center distances are normally 6.25 times the tube diameter, or greater, and uniform over the tube field. .*@ Since the shell and tubes operate at&fFerenbtemperatures, it is n q s a r y to provide meam to accommodate the diflerence in their thermal expansion, as h i stresses could otherwise be developed within the heat exchanger. The various types of construction which have been used to accommodate thermal expansion are illustrated in Fig.5(e), Cf), and (h) and Fig. 8; each has its own area of application.
I
496
MARINE ENGINEERING
HEAT EXCHANGERS
expansions present no difficulties with this type of design. When the tubes are firmly fastened to the tube sheets and the tube sheets are fastened to the shell (i.e., fixed tube sheet designs), shell expansion joints are used to lessen the stresses caused by the difference in thermal expansion between the shell and tubes. The shell expansion joints illustrated in Fig. 9 are adequate only for small differential thermal expansions (say 0.06 in. for a 5-ft tube). For large thermal expansions, expansion joints of the bellows type are normally used. To provide for thermal expansion when the shell of a heat exchanger operates at an elevated temperature, (a) Shell expansion joint using flanged only one "foot" or "leg" of the shell is anchored to the and flued heads foundation and the other is designed to permit free axial movement sg that the axial expansion of the shell is not resisted. Freedom for axial movement may be provided by either a sliding foot (slotted holes with ferrules) or a slender leg (in the case of heavy heat exchangers) which provides little resistance to lateral force. 1.4 Design Data Requirements. In order to specifically direct attention to the items which govern the design of a heat exchanger and which must be furnished to the design engineer (or assumed by him), those items which should be included in a specification for a marine heat exchanger are listed below. Substance to be heated (and cooled) (b) Split pipe shell expamion joint Quantity of substance to be heated (and cooled) within a given period of time Fig. 9 Shell expansion iainh Initial temperature of the substance heated (and cooled) Final temp rature desired for the substance heated The simple U-tube illustrated by Fig. 8(a) is the most (and cooled) economical and commonly used means to separate the When the heating and cooling media are other than thermal expansion in the tubes from that of the shell. water or commonly known substances, the following I t is widely used in small condensers and instantaneous should be specified : heaters. (a) Viscosity The bayonet tube, illustrated in Fig. 8(b), is suitable (b) Specific gravity for use with a tight, cross-flow baffle spacing. Bayonet ' (c) Specific heat tubes are used in viscous oil heaters. (d) Thermal conductivity Packed-tube joints, Figs. 4 and 8(c), permit some axial Working pressure of the heated (and cooled) movement of the tube. Packed-tube joints are used in substance rectangular (box type) low-pressure feed heaters and in Allowable pressure drop through the shell and tube smaller straight-tube condensers. sides of the heat exchanger In cases where a large differential thermal expansion Desired construction materials can be expected, a floating head arrangement may be Typical features of a variety of marine heat exchangers employed. As can be seen from Fig. 8(d), large thermal are listed in Table 1.
T9
t = temperature of cold fluid A, = tube outside surface area 2~ = rio rdio rw rdo ro r, = resistance across the fluid film on the inside of the tube rdh = resistance of the deposit or scale on the tube inside wall r, = resistance of tube wall metal rdo = resistance of the deposit or scale on the tube outside wall ro = resistance of the fluid film on the outside of the tube
+ + + +
1i I
i
!
.
Section 2 Heat Transfer in Shell-and-Tube Heat Exchangers Q =
Ao(T - t) Zr
where
Q
(1)
I1 I
= heat transferred
T = temperaturk of hot fluid
SCALEPR DIRT
Fig. 10 Heat transfer through a tube
b
2.1 Heat Transfer Relationships. The fundamental theory dealing with heat transfer was discussed in Chapter 2, and a discussion of heat transfer by condensation is given in Chapter 13. The application of this theory to the transfer of heat through the walls of a tube leads to the equation
TUBE ?ALL
These quantities are illustrated by Fig. 10, and the oorresponding temperature gradients across a tube are ahown in Fig. 11. I t may be noted that "new clean-tube oxide film resistances" have not been included on the inside and outside of the tubes as they are not of sufficient magnitude to warrant consideration in the design of viscous fluid or water-to-water heat exchangers. The film resistances in equation (1) are further defined
as 1 1 1 ri0 = =(2) hiAi/Ao hi, tube-side film coefficient 1 1 + To=-- (3) h. shell-side film coefficient A, and A; are the tube outside and inside surface areas respectively. The terms hi and ho are called specific conductances or heat transfer film coefficients and are generally a function of flow velocities and fluid properties. The scale resistance or fouling resistance terms, rdo u,nd rdo,do not lend themselves to analytical assessment and little is known regarding them. The resistance of the tube wall metal is readily analyzed and it can be quantified without difficulty. The overall heat transfer coefficient, U, which is defined as the reciprocal of Zr, is a convenient means of axpressing the resistance to heat transfer through tubes. The overall heat transfer coefficient can be stated as
Fig. 11
Temperature gradients auou'tubes
location within the exchanger. Consequently, the heat flow rate, Q, also varies from location to location. To obtain the total heat transferred per unit time for a heat exchanger, it is convenient to base calculations on the average temperature difference across the tube over the entire tube length. Such a temperature difference is defined as the log mean temperature difference (LMTD) and is expressed mathematically aa LMTD =
-1
1 .A0
(T - t)dA
(5)
By combining equation (5) with equation (1) and noting that Zr is equal to the reciprocal of U: the heat transfer through a tube is determined to be
Equation (6) is the basic analytical tool employed to establish the thermal design of a heat exchanger. The factors to be considered in assessing this equation are briefly discussed in the following. With known terminal temperature differences between the shell and tube streams, the mean temperature difference between the shell 'and tube flows can be derived [7] if the following aasumptions are made: The overall heat transfer coefficient is constant along the entire flow path. The flow rate and.pxific heaj of both the shell and tube streams are constant. The heat transfer surface is uniformly distributed along the flow paths. The temperature of either fluid is constant over 1 any cross section of its path (i.e., there is complete U = (4) (I/&) rdo rw rdio (Jlhio) mixing and no stratification). There is no internal leakage or bypassing of fluid In the case of shell-and-tube heat exchangers, the temperature difference (T - t) varies from location to around the tube bundle.
+ + + +
498-
MARINE ENGINEERING
499
HEAT EXCHANGERS
for C allows a considerable m a r k (as com~aredwith a value of 0.33) to allow for flowleakages asLs-would occur tube-side film coefficient between cross baffles and the shell, and should only be tube inside diameter used as an average value. thermal conductivity of fluid on tube side The term "ideal tube bank" came from a report [lo] 0.027 published as s result of a research program on shell-andtube-side mass velocity tube heat exchange? conducted at the University of tube-side fluid viscosity at bulk temperature Delaware. The research /program was supported by tube-side fluid viscosity at tube wall temperature various heat exchanger manufacturers and lasted more Prandtl number = CJ/K than 12 years,, An "ideal tube bank" came to mean a specific heat rectangular hube field with straight-through flow and no Figure 13 is a nomograph of the film coefficient inside wall effects or bypassing. Theresults from these tests tube walls for turbulent water flow inside No. 18 BWG can be presented in the form of equation (13) if the contubes as determined from equation (12). A means of stants C and m are replaced by functions of the tube oorrecting for other tube dimensions is indicated in pitch, P, the tube diameter, do, and the Reynolds number, (doG/Z), in the lower Reynolds number range. Fig. 13. The relationship used to assess the shell-side film That is C = b(1.33d0/P)" (14) ooefficient is similar in form to that used on the tube where aide and is as follows:
where
(c) T w singbshell and double-tube pass exchangers in series
Q) Ona shall pass and multiple of twa tube passer
RQ. 12 Flow arrongeaenh Light flow Pnes mpreaent lube-side flow and heavy itow lines represent shd-ride flow
There is no transfer of heat between the heat exchanger and its surroundings. For a single-pass counterflow heat exchanger such as illustrated by Fig. 12(a), the LMTD is
Single-shell heat exchangers with even numbers of tube passes may be arranged in series. With identical heat exchangers arranged in series, as shown by Fig. 12(e), the LMTD is computed as LMTD =
A
where
where
T1= shell-8ide inlet temperature T2 = shell-side outlet temperature €1 = tube-side inlet temperature ta = tube-side outlet temperature For the special case that
then equation (7)reduces to LMTD = TI- B, or TI- C
(9) For one-shell and multipl& of two-tube-pass heat exchangers as shown in Fig. 12(b), the.EMTD becomes LMTD = where
X Y
A
Y+X
- T s ) ~+ fh - t~)']"' = TI T2 - (Is + 51) = [(Ti
+
Values for b and m are as follows:
where
lia
m = number of identical single-shell heat exchangers
with an even number of tube passes in series The tube-side film coefficient, hi, may be computed from the Sieder and Tate equation [$I. This equation waa published in 1936 and remains in widespread use; it has the following form:
ho = shell-side film coefficient do = tube outside diameter C, m = constants
Reynolds Number b I
m
0-10 1.4 0.333
10-100 1.36 0.343
100-1000 0.593 0.533
The remaining terms are as defined for equation (12). In the higher Reynolds number range (Re > 1000), There are some variations in the technique8 used in C and m are taken as constants having the values of 0.32 applying equation (13); the variations primarily deal and 0.612, respectively. * ,* with the quantification of C and m. For calculations In a commercial shell-and-tube heat exchanger, there regarding the flow of a gas normal to the tube bank (i.e., is a gap between the tube bundle and shell. A portion a cross-flow heat exchanger), Colburn recommended in of the fluid in the shell will therefore bypass the tube 1033 that equation (13) be applied as follows: bundle, and not be cooled or heated. Also, there is flow leakage between cross-flow baffles and the shell. m = 0.6 It is essential that these non-ideal conditions be Z and Zware assumed equal accounted for, especially in any effort to reduce uncerG = mass velocity of fluid through the minimum flow tainties of flow pressure drop prediction. Tinker [ll) cross-sectional area of the @be tank and Bell [lo] each proposed simplified methods to C -- 0.33 for staggered tube pattern (triangular or correlate these complex flow phenomena to geometry rotated square pitch) factors. Their basic approach is to divide the shell C = 0.26 for in-line tube pattern (square pitch) flow area into three categories, namely, cross-flow area When making computations for the flow of a fluid (the flow area between tubes in the direction normal to through a circular baffled heat exchanger; Donahue the tube axis); leakage area (the flow area between the recommends that 'equation (13) be applied with the tube and tube holes in the cross-flow baffles and the area following assumptions [9]: betweed the b d e outside diameter and the shell inside diameter); and bypass area (the flow area between the m = 0.6 tube bundle and shell). Bell's is an overall approach; G = (G,G,)lln with known relative values of bypass and leakage area G, = mass velocity normal to the tube bundle to that of cross-flow, he, obtained empirically correction Gw = mass velocity through cross baWe window (i.e., factors to be applied tofdeal tubecbank heat transfer baffle cut out area, Fig. 6[a]) coefficient and pressure drop values. Tinker assigned The geometrical mean value of the mass velocity is used flow resistance constants to each flow area and calculated in recognition of the fact that the direction of fluid flow an effective cross-flow rate to be used in the ideal tube hank cross-flow heat transfer and pressure drop correlais not normal to the tubes in the baffle window. On the basis of test data with segmental baffled heat tions. Tinker's is a more useful approach in that refineexchangers, Donhue further recommends that C be ment and generalizations are easier with his method. I t may be seen that computations for the shell-side given the value of 0.22 as an average value. This value
MARINE ENGINEERING
NOTE: Nomograph is based on 518" O.D. #18 BWG t u k For tubes of other O.D., Q end I.D., di, COW ths hio read a follows:
fig. 13
Rlm mo(ficierd for water inside hbw
HEAT EXCHANGERS
NOTE: This nomograph is valid for plain tuber having a 114" O.D. on e 11132"- A pitch or a 3/8"O.D. on a 17/32" A pitch.
MARINE ENGINEERING NOTE: This nomograph Is valld for 518" O.D. plain tubes on a 13/16" A pitch
HEAT EXCHANGERS
NOTE: This nomograph is valid for 518" O.D. t u k with a triangular a d rotated-rquare pitch. . For other tube diameters, multiply chart reedings by do, in.
Multiplier
Fig. 15 Etfective flow area In wgmmtal b d e d flow
Fig. 16 Fltn coefficient for water in segmental b d b d flow over plain tubes
1
504
MARINE ENGINEERING
HEAT EXCHANGERS
II
film coefficient are tedious and time consuming, and a particularly troublesome aspect of the calculation is the assessment of the effective flow area for segmental baffled flow. To simplify calculations, Figs. 14 and 15, which are based on references [ l l ] and [12], were prepared to provide an approximation of the effective flow area. Figures 14 and 15 cover the tube sizes and pitches normally used with marine heat exchangers. It should be noted, however, that manufacturing tolerances on the shell, tubes, and baffles vary and their variation can appreciably affect the value of the effective flow area, especially in the case of small heat exchangers. Therefore, these nomographs, which are based on a particular set of tolerances nominally followed by heat exchanger manufacturers, are of value primarily as an indication of the influence of the various factors which enter into the design of shell-and-tube heat exchangers. .For instance, it may be seen from Fig. 14 that small b d e spacings (small in relation to the shell diameter) are progressively less effective in creating small effective flow areas (or high velocities). These nomographs are particularly useful as devices for checking a design and narrowing down some of the parameters (shell size, tube size, b d e spacing, etc.) in the course of designing a heat exchanger. Figure 16 is a nomograph which relates the shell-side film coefficient to the water temperature and mass velocity. Figure 16 is based on the data presented in reference [lo]. Two means have been commonly employed to allow for the additional thermal resistance when scale or deposit begins to accumulate on the tube wall. The most ri&rous means is to assign values to the tube inside and outside fouling resistances, rdio and rh, as a function of fluid type, temperature, velocity, etc., and compute the required heat transfer surface accordingly. I n applications with mild rates of fouling, however, the fouling resistances are difficult to separate and evaluate meaningfully, and furthermore, they are small. Consequently, the "clean factor" concept has found general acceptance in conventional steam power plant and marine heat exchanger applications. The clean factor is a factor less than unity by which the clean overall heat transfer coefficient, U, is multiplied to allow for fouling. I n order to illustrate the procedure employed when performing heat transfer calculations for a shell-and-tube heat exchanger, consider such a heat exchanger with the following characteristics: Shell size = 6.075 in. I D Tube size = % in. OD 20 BWG (plain tube) Tube material = Admiralty metal Number of tube passes = 2 Number of tubes = 36 U-tubes Tube length = 44 in. Tube pitch = 1 x 2 in. Cross baffle spacing = 2% in. Effective heat transfer surface = 24 sq f t
With 50 gpm of fresh water at 78 F entering the shell side of the heat exchanger, the outlet temperatures and heat exchanged when the fresh water is cooled by 37 gpm of chilled water at 50 F entering the tube side will be computed. First, it is established that the tube inside and outside diameters are 0.305 in. and 0.375 in. respectively. The water velocity through the tubes is computed to be 4.5 fps, and from Fig. 13, the uncorrected film coefficient inside the tubes is found to be 750 Btu/hr-sq ft-deg F. Correcting for the tube size. results.in a film coefficient hi, of 820 Btu/hr-sq ft-deg F. With the stated values for the b d e spacing and shell ID, the effective flow area is determined to be 6.1 sq in. from Fig. 14. The shell flow rate was given as 50 gpm which then corresponds to a mass flow rate, G, of 164 lb/sec-sq ft. Entering Fig. 16 with this value for G and a water temperature of 78 F gives a shell-side film coefficient, h,, of 1735 Btu/hr-sq ft-deg F. The tube metal resistance, i.,, is readily calculated by employing the following equation for heat conduction acrom a cylindrical wall: I, =
NOTE: Nomograph is valid for turbulent flow with a specific pmity of unity within 518" O.D. #18 BWG tuk "
A, log. (doldi) 27rLK
where L is the tube length, K is the metal thermal conductivity, and A, is the tube outside area. For a smooth tube, A, is equal to 7r dJ, and the tube metal resistance reduces to I, =
do log. (do/&) 2K
Since K = 70 Btu/hr-ft-deg F for Admiralty metal, equation (16) gives an r, value of 0.0000461 hr-sq ft-deg F/Btu. Based on experience, a clean factor of 0.9 is considered appropriate for this service; therefore, the scale resistances, r& and rdi,, are considered to be zero. This being the case. substitution of the com~uted values into equation'(4) gives an overall heat transfer coefficient, U, of
+ 0 + 0.0000461 + 0 + &
1
1735
= 544 Btu/hr-sq
ft-deg F
When the clean factor of 0.9 is applied, there results U
=
(0.9)(544)
=
490 Btu/hr-sq ft-deg F
and this is used as the design value. It may be noted that the effect of the term corresponding to the thermal conductivity of the tube wall material is small when compared to the effect of the film resistance; for this reason the thermal conductivity of the tube wall is often neglected. I n order to compute the quantity of heat transferred, equations (6) and (10) are used in conjunction with the following heat balance relationship :
ma. 17 R.roun drop idde tub-
MARINE ENGINEERING
HEAT EXCHANGERS
E = exp
(2:)
The two specific heats are unity in this case, therefore all terms can be evaluated. Tzis found to be 69.8 F and t a is 61 F. With these values known, the heat transferred is computed as:
2.2 Pressure Drop. The film coefficients (and consequently the heat transfer rate) increase with an increase of flow velocities, as noted from the previous discussion; but on the other hand, the flow pressure drops due to friction are proportional to the velocities raised to a power between one and two. Consequently, a design tradeoff must be made which entails striking a compromise between high film coefficients and high pressure drops. The design of a heat exchanger becomes a matter of balancing the saving of heat transfer surface with the cost of pumping power to the exchangers. As a result, a major portion of the design work is expended in evaluating flow pressure drops for various operating conditions and exchanger configurations. In practice, the system designer very often must specify the allowed pressure drops for heat exchangers without an accurate knowledge of their impact on the design of the heat exchangers. There are three courses of action which are frequently taken when it is necessary to specify the pressure drops permitted with heat exchangers: Allow a flat 10 psi pressure drop per stream per heat exchanger. This rule-of-thumb approach is popular in some areas of the industry. The engineer has preferred to stay within the 10 psi limit since higher Fig. 18 Tube diameter and specific gravity correction faclora for tube pressure drop pressure drops with higher velocities may approach the erosion/corrosion limit of the metal and furthermore the resultant high heat transfer coefficient may be excessively sensitive to scaling. The 10 psi maximum allowable pressure drop is also high enough to keep the thermal where designer from "tail-chasing " (i.e., the lower the velocity, W. = shell-side flow rate, lb/hr the lower the heat transfer coefficient; the lower the heat W t = tube-side flow rate, lb/hr transfer coefficient, the larger the heat exchanger; the C , = shell-side fluid specific heat, Btu/lb-deg F larger the heat exchanger, the larger the pressure drop, C p t = tube-side fluid specific heat, Btu/lbdeg F which requires lowering the velocity, etc.). Specify alternative pressure drop limitations. and the fluid temperatures are defined as before' With the wide application of the computer to heat Equations (6), (lo), and (17) can be combined to express the two water outlet temperatures in terms of known exchanger design, this approach does not incur an excessive engineering load in identifying alternatives, and values. The resulting expressions, arranged in the order selecting the optimum alternative can be advantageous. of the computational procedure, are as follows: Use design charts to obtain an approximate drop assessment of t i e heat exchanger size and involved before establishing the design criteria.
Fig. 19 Equlwbnt number of tube rows f w bdffled-fiow prenun drop calculation
MARHE ENGINEERING
HEAT EXCHANGERS
509
4 NOTE:
This nomograph is valid for 518" Wolverine tubes type SIT on a 13118" A pitch.
5 /
8, 7 '\
fig. 20
Pressure drop in segmental baffled flow of water w e r plain tuber
In order to illustrate the use of design charts to obtain approximate, but quick, estimates of the pressure drop through heat exchangers, the example problem in Section 2.1 will be continued. Entering Fig. 17 with a water velocity of 4.5 fps and temperature of 50 F gives a pressure drop within the tubes of 0.085 psi/ft. Since there are two tube passes, the total tube length is twice 44 in. or 7.34 ft, and the total uncorrected pressure drop is 0.624 psi. The inside diameter of the tubes is actually 0.305 in. and the specific gravity of the fresh water is unity; therefore, the corresponding correction factors to be applied to the pressure drop taken from Fig. 17 are read from Fig. 18 aa ad = 1.95 hnd a. = 1. Applying these to the uncorrected pressure drop of 0.624 psi gives a corrected pressure drop of 1.22 psi. In order to calculate the baffled-flow pressure drop on the shell side, the number of baffles must be established. The number of tube baffles can be estimated as tube length -1 = 44 -1 = 18 2.25 baffle spacing I n addition, the number of tube rows which the flow crosses between each pair of baffles and the baffle cut-out area (baffle window) must be established before the calculations can proceed. It is convenient to introduce the term N,, which is defined as the equivalent number NB =
of tube rows that the flow crosses between each pair of baffles, as the ensuing calculations are greatly simplified. The equivalent number of tubes citn be determined if details regarding the baffle window height, baffle spacing, tube spacing, and shell are known; but such is seldom the caae in preliminary design work. For estimation purposes, Fig. 19 has been prepared oh the basis of a commonly used heat exchanger configuration to give guidance in determining the equivalent number of tube rows crossed between baffles. Entering Fig. 19 with a shell I D of 6.075 in., a baffle spacing of 2.25 in., and a triangular tube pitch of 17dZ in. gives an equivalent number of tube rows of 11.5. Consequently, the equivalent total number of tube rows is computed as NR = (NB
+ 1)N. = 21.9 rows
The pressure drop per row of tubes can be evaluated from Fig. 20. A line is drawn between a water temper* ture of 78 F and a mass velocity of 164 lb/sec-sq ft (line 1)) and the interception with the reference line is noted. Next, line 2 is drawn from the reference line intercept and the second mass velocity curve, and the product of the fluid specific gravity and pressure drop per tube row of 0.024 psi/row is obtained. Since the fluid specific gravity is unity, the shell-side baffled-flow pressure drop is (0.024) (21.9) = 5.26 psi.
Fig. 21
Meciive flow area for segmental baffle flew
MARINE ENGINEERING
HUT EXCHANGERS
51 1
that the resistance to heat transfer increases rapidly initially, but the rate of increase subsequently drops to a very sm* value. The tests h indicated that the fouling characteristics of finned tubes are similar to those of plain tubes, and both types of tubes have comparable percentages of reduction in performance. At the end of the four-week test period, ,the fouling resistances for both fin-tube and plain-tube bundles were below the value of 0.005. During the tests, both the finned and plain tubes were cleaned by a kerosene-water-detergent emulsion and the heat transfer rate waa restored to that initially achieved. In order to illustrate the princjples involved in the e,pplication of finned-tube heat exchangers, consider such a unit with the following characteristics: Shell imide diameter, Di = 23.265 in. -- Tube size = %-in. low fin Tube material = aluminum Number of tube passes = 2 Number of tubes, N = 652 Tube length, L = 96 in. in. Tube pitch = B&e spacing = 9 1 % ~in. Tube length between tube sheets, L, = 89% in.
h = ,'
B~UIM~Z-F
1
% fig. 23
A check will be made to determine the suitability of this
+
Rdo
Fin resistance of low-fin tubes
unit for the following service: Shell Side
NOTE: Multiply ho mad above by (2/z1,-~)-14
when, Z = viscosiw at average tempenturn, centipoi~er & 5 vMcosiw at tube wall temperature, centipdnes
fie. 22
t
1
Rlm coeffidenf ftw oil in segmental baffled tlow wer low-tln tubes
2.3 Design of Fin-Tube Heat Exchangers. Another form of the'basic heat transfer relationship given as equation (1) can be written as
where Ra, and Ro are the scale resistance on the inside and outside of the tubes respectively, Rwis the resistance of the tube, and the remaining symbols are as previously defined. If the shell-side film coefficient, ho, is small (as is the case with gases and oils which have low thermal conductivities) in comparison with the tube-side film coefficient, hc the shell-side resistance, (h,A,)-1 will control or "bottleneck" the heat flow. A means of counteracting this circumstance is to increase the tube
outside surface area, A,, to the extent that hoAo= hiAi, in which case the heat flow would no longer be choked by the shell-side heat resistance. The heat transfer surface on the outside of the tubes can be effectivelyincreased by providing fins on the tube outside surface. A type of finned tube frequently used in shell-and-tube heat exchangers is the so-called "low-fin" tubes which have a ratio of finned surface to tube outside surface in the range of 3 to 4. In applications where the tube outside film coefficient is as small as 50 percent of the tube inside film coefficient, the choice of low-fin tubes will permit a more economical, compact, and light unit than a similar design employing bare tubes. Regarding the fouling of finned tubes, tests [13] were conducted with 160 F, No. 5 fuel oil flowing through the shell side of a heat exchanger. The test results showed
Fluid. .................. Flow rate, W, lb/hr.. .... Specific gravity at average temperature. . . Thermal conductivity, Btu/hr-fMeg F ....... Viscosity: 2,centipoises at 210 F.. ............ at 167 F.. ............ Specific heat, C, Btu/lb-deg F. ......... Inlet temperature, F. .... Outlet temperature, F . ...
oil 123,400 0.84 0.0665
106 280
Tube Side water 66,500
pute the tube-side film coefficient. The tube inside diameter is 0.402 in., the average fluid temperature is 158 F, and the tube-side ve10lcit"~ is computed as 1.03 fps. A value for h, of 435 Btu/hr-sq ftaeg F is taken from Fig. 13 for Yrin. OD 18 BWG plain tubes which have an outside/inside area ratio of 1.186. Low-fin tubes have an inside/outside ratio of 0.26; and therefore the correction for low-fin tubes is made as follows: hio =
( L ~ ) Ai
plain tube
(4) Ao
0.527
O''
finned tube
= 141 ~tu/hr-sqft-deg F
To compute the shell-side film coefficient, the effective flow area must be known; this is determined to be 146 sq. in. from Fig. 21 by entering with the stated values The quantity of heat to be transferred is computed as for the shell I D and the baffle spacing. The effective W,C,(T1,- Tt) = 1,720,000 Btu/hr, and the LMTD is flow area, in conjunction with the shell flow rate, gives determined to be 47.3 F from equation (10). an effective mass velocity of 33.8 lb/sec-sq ft. The Low-fin tubes having an outside diameter of % in. shell-side film coefficient can be determined from Fig. 22 have a finned surface area of 0.405 sq ft per linear foot; by drawing a line (line 1) between the viscosity and mass therefore, the total finned surface, A,, is 0.405 NL, = velocity scales and noting the intercept on the reference 1970 sq ft. line. Next, a line is drawn from the thermal conducI t is now possible to compute the required overall heat tivity scale through the ina&cepf poia on the rehrence transfer coefficient from the expression line, and the intercept with the h. scale is read as 22.5 Btu/hr-sq fMeg F. This value must be corrected, howU = Ao(LMTD) = 18.45 Btu/hr-sq ft-deg F ever, to account for the variation in viscosity at the tube wall. Since the temperature of the oil at the tube wall If in this service the unit will provide an overall heat is not known at this point in the calculation, it is necestransfer coefficient of this magnitude, it would be sary to assume a temperature and then confirm the considered satisfactory. The h t step in obtaining an assumption to be satisfactory when it is subsequently estimate of the heat transfer rate for the unit is to com- established. With an assumed tube wall temperature
i
I
I
11/ II
I
512
The assumed tube wall temperature upon which the calculation was baaed must now be checked. The calculated tube average wall temperature is 1, =
Tmg-
(LMTD) = 168.7 F
-
This is reasonably close to the assumed temperature of 167 F ; therefore, the viscosity correction factor of 0.873 is sufficiently accurate for practical purposes. The calculated clean overall heat transfer coefficient gf 17.2 is somewhat smaller than the required overall heat transfer coefficient value of 18.45; consequently, no allowance for fouling during service is provided. In an actual design, the heat exchanger design should be modified to improve the unit's heat transfer characteristics. This could be accomplished by decreasing the baffle spacing (so as to increase the flow velocity and increaae the heat transfer coefficient) if the associated increase in pressure drop can be accommodated; or an alternative solution would be to increase the amount of heat transfer surface. Tests which have been conducted to corroborate the calculation procedure outlined in the foregoing have shown the method to be valid. Measured heat transfer coefficients have been slightly greater than those calculated, which is a desired characteristic for the difference to possess. The fundamental relationship for computing the pressure drop on the shell side of a finned-tube heat exchanger has the form
this line, use Fig. 25 for turbulent flow
.2J
ffg. 24
5 13
HEAT EXCHANGERS
MARINE ENGINEERING
I
Pressure drop In wgmontal baffled Aow wer low-fln tuber
of 167 F, the corresponding viscosity is 280 centipoises. 19.7 (aasurning that the shell-side fouling resistance, R*, This, in conjunction with the average oil temperature of is zero) the resistance of aluminum h, rti., is seen to be 210 F and viscosity of 106 centipoises, gives a viscosity 0.00013 hr-sq ftdeg F/Btu. As may have been anticicorrection factor of 0.873 and a corrected shell-side hlm pated, the resistance of low-fin tubes is quite small for their usual applications. coefficientof 19.7 Btu/hr-sq ftdeg F. The thermal resistance, r., of the annular ring portion Another factor, in addition to those involved with the analysis of plain tubes, is the thermal resistance of the of the fin tube wall is not included in the fin resistance fins themselves. When computing 'the heat transfer taken from Fig. 23. However, it is readily computed coefficients of finned tubes, an allowance must be made from equation (15), and is found to be 0.00012 hr-sq ftfor the fins because, aa part of the tube walls, they offer deg F/Btu. The estimated overall heat transfer coefficient can now some resistance. An estimate of fin resistance may be made from Fig. be evaluated aa 23, which presents the fin resistance of low-fin tubes of 1 diierent ,materials in common use. The fin efficiency, = 1 = 17.2 Btu/hr-sq ftdeg F upon which the resistances shown were baaed, was taken h, rti. r, from reference [14]. With a shell-side film coefficient of
+ + +
NOTE:
1.000
nomograph is for E/S" low-fin tuber on a 13/16" A pitch and Reynolds numben greater than 1800.
800 600 400 300
a2 /
,' I''
0.0003
where sg is the specific gravity of the shell-side fluid, and the other terms are aa previously defined. The viscosity term is known aa it was evaluated in connection with the determination of the shell-side a m coefficient. The number of baffles is given aa 8 and the fluid specific gravity is 0.84. With the shell and t@e data previously given, the equivalent number of tube rows crossed per . b a e d space is found to be 27 from Fig. 19. Figure 19 is applicable to both low-fin and plain-tube configurcc Fig. 25 Pressure drop in segmental baffled flow wer low-fln tuber (turbu1e.d tions. Note also that the effective flow a r e a evaluated flow range) from Figs. 14, 15, and 21 are used for both beat transfer and pressure drop calculations; this is a result of the approach taken in correlating the test data. velocity pcale, and the intersection with reference line B Only the (AP X sg) term remains to be determined. is noted and defined aa reference point 2. Finally, A value for this term, can be obtained from Fig. 24 or line 4 is drawn through reference point 1 (line 2 and Fig. 25. Entering Fig. 24 with the previously established viscosity scale) and reference point 2 (line 3 and reference viscosity of 106 centipoises and mass velocity of 33.8 line B), and at the intersection with the sg X A P scale, lb/sec-sq ft, line 1 is drawn between the viscosity scale the value 0.02 psi/row of;Abes ia read. and the left-hand mass velocity scale. At the point line With all terms in equation (19) evaluated, the shell1 crosses reference line A, line 2 is drawn along and in side pressure drop is found to be 6.64 psi. between the inclined solid guidelines which are provided The tube-side pressure drop is calculated the same aa to ensure that line 2 has the proper slope as it intersects with plain tubes. With an average water temperature the viscosity scale; note the intersection with the of 102.5 F and a water velocity of 2.55 fps, the pressure viscosity scale and define it aa reference point 1 (it haa drop per foot of tube length is 0.026 psi/ft, from nothing to do with viscosity). Next, line 3 is drawn Fig. 17. The tube diameter correction factor, a d , is 1.4 between the viscosity scale and the right-hand mass and the specific gravity correction factor, a,, is unity.
514
HEAT EXCHANGERS
MARINE ENGINEERING
Therefore, with an 8-ft tube length and two passes, the tube-side pressure drop becomes The method employed to formulate the nomograph for rrssessing the heat transfer and pressure drop within heat exchangers is bersically that of Tinker [ll], Devore
[12], and Donahue 191. Details of the derivations and assumptions involved with the method have not been reviewed as they are not of primary interest to the ~racticine:engineer. The Drocess of mine: through the procedure ouiined will, hoGever, convey an appreciation of the oonsiderations involved in the hydrodynamic and thermal design of a shell-and-tube heat exchanger.
- -
P.T. WNN-NEAR & FAR SIDES
a
515 P.T. VENT-NEAR SIDE PLUGGED /
OIL W N N .
\
PIPE TAP (P.1.)
Section 3 Heat Exchanger Applications 3.1 Lubricating-Oil Coolers. In addition to serving as a lubricant between moving mechanical parts, lubricating oils generally also accomplish a second objective of removing the frictional heat generated. Therefore, some means must be provided for removing the heat absorbed by the lubricating oil. With small systems, the natural heat transfer by radiation and convection may be adequate; but with larger systems, particularly those employing forced circulation, lubricating-oil coolers are required. The main lube-oil coolers (i.e., those in the lube-oil system serving the main turbines, reduction gears, and main thrust bearing) are generally the largest used aboard ship. A shell and straight-tube exchanger with a removable tube bundle and a double-packed floating tube sheet type of construction (such as illustrated by Fig. 26) is the design most commonly used for main lube-oil coolers. Oil flows in a single pass via transverse bafaing in the shell, and seawater is the normal coolant flowing in one or more passes through the tubes. The floating tube sheet is centered between the shell flange and waterbox flange. A gland ring retains the packing. Separate packing rings are provided for the shell (lube oil) side and the coolant side. The gland ring is grooved around the inside, and leak-off holes are provided so that leakage past the packing rings on either side will be relieved to the outside and attract the operator's attention. The stationary tube sheet, baffles, and support plates are assembled and held in proper relative position by tie rods and spacer sleeves. Tie rods are threaded into but not through the stationary tube sheet. The waterbox and stationary tube sheet are secured by collar bolts or by stud bolts driven into tapped holes in the tube sheet so that the tube-sheet to shell-flange joint will not be broken when the waterbox is removed. The packing retainer ring is usually "scalloped" so that there is a stud hole for every second stud in the shell flange st the floating tube sheet end of the cooler. This enables the gland ring to double as a test ring and allows the shell to be hydrostatically tested without the waterbox in place. In some of the more demanding applications, such as submarine heat exchangers cooled by seawater, there are special requirements for bolting as well as for materials,
and such requirements must be considered in the cooler design. Limitations are also provided for maximum fluid velocities and pressure drop, and shock resistance requirements may also be specified for military applications [6]. In the late 1950's, it was determined that the performance, in terms of fouling properties and service life, of low-fin tubing is approximately equal to that of bare tubes, and a gradual change has taken place such that %-in. and %-in. low-fin 90/10 CuNi tubing has become the preferred tubing in lieu of the previous %-in. and %-in. bare tubes. With low-fin tubes, the greater amount of heat transfer surface within a given shell size outweighs the disadvantage of the somewhat lower heat transfer rates of finned tubes, and the net result is a smaller, more compact, and more economical unit for a given performance requirement as compared with bare tubes. Lubricating-oil coolers are also used aboard ship for auxiliaries such as main feed pumps and air compressors. These units are similar in design to main lubricating-oil coolers but are smaller, having a shell diameter of 6 to 10 inches. The thermal design of oil coolers is treated in detail in Section 2. 3.2 Fuel-Oil Heaters and Lubricating-Oil Purifler Heaters. The residual fuel oils usually burned aboard ship are so viscous at atmospheric temperatures that they must be heated before they can be pumped from the storage tanks (the properties of fuel oils are discussed in detail in Chapter 23). The more viscous fuels may require heating to 120 F or more before they can be pumped; and when supplied to the burners, the fuel oil must be at an even higher temperature in order to attain a fuel viscosity sufficiently low for proper fuel atomizrlr tion in the burners. Therefore, two stages of fuel-oil heating are provided; the primary heaters are installed in the fuel-oil tanks, and the secondary heaters are installed between the service pumps and the burners. The primary oil heaters are installed in the tanks in either of two forms: as steam-supplied pipe grids or coils or as open-ended tank-suction heaters. The latter alternative has become increasingly popular due to (a) lower initial cost, (b) lower maintenance costs, and (c) lower steam consumption (only the oil to be pumped is
\ SFE DETAIL "A"
LEGEND, 1. Shell 2. Water Channel 3. Channel Cover 4. Return Channel 5. Fixed Tube Sheet 6. Flpeting Tube Sheet 7. Tubes 8. Baffle Assembly 9. Retainer Ring 10. Packing 11. Shell Flange Gasket 12. Channel Gasket 13. W r r Gasket 14. Zinc Amdo# (wha~Imludd)
SHELL FLANGE
GI
DETAl L "A" Detail of Floating Tube Sheet End Section at pecking retaining r i ~ g w e p h o l e .Mixing of the shell and tube side fluids thru the packing is impossible with this detail. Any seepage from either side, rwulting from loosening of the bda, d r i p out thru the copper-lined packing ring weep-holes, thus putting the m t i n g personnel on notice. The nuts on both sides may be made up to tighten packing without hutdown or interruption of operation. fig. 26
mg. 27
Typical main lube-ail c o o k
Typicol tank rudon fuel-oil heator
MARINE ENGINEERING
heated as opposed to heating the entire tank). A typical tank-suction fuel-oil heater is shown in fig. 27. hi horizontal U-tube heat exchanger has an outer shell flange that is bolted directly to the tank. The oil is drawn in through the open end of the shell and across the tube bundle. As the heating medium (condensing steam) circulates through the tubes, the portion of oil in the heater shell in contact with the tubes rises in temperature with a corresponding decrease in viscosity. The entire process is continuous with the oil flow being
HEAT EXCHANGERS
CLEANING FLUID INLET
MOUNTING BOLTS
RELIEF VALVE CONN.
from those surfaces. Steam is delivered to the steam chamber, and from there enters a series of small tubes, cdled "inner" tubes, through which it travels until it is dpcharged into the annulus between the inner and outer tubes at the far end of the outer or oil-heating tubes [see Fig. 8(b)l. Thus steam (free of condensate) is in contact with the heating surface at that part of the surface where the highest oil temperature is desired. The space between the inner and outer tubes is small so that the volume of steam flowing in the annular space generates a velocity of flow sufficient to continuously sweep the condensate from the surface as it forms, thus reducing water-film losses and preserving high transfer rates on the steam side.
CLEANING FLUID OUTLET
,
PRESSURE GAGE CONN.
serves aa a shell. The Navy haa set forth specific design criteria for all three of the foregoing designs, and further describes an "evaporator" type of fuel42 heater which uses an intermediate fluid to transfer heat from a primheating coil to the bayonet oil-heating elerpents; this precludes any possibility of contaminating the condensate with fuel oil in the event of a leaky tube or tube sheet joint [15]. Steel construction throughout is standard practice. A minimum of two heaters is normally provided, sized and so arranged that either heater can provide full boiler requirements at overload with the other heater serving as a standby. Heaters are fitted with a relief valve which usually discharges to the settling tanks via a check valve. The automatic temperature control valve is normally installed in the steam supply line with the sensing element located in the oil outlet flow line immediately adjacent to the heater. The steam flow valve is usually installed immediately adjacent to the steam inlet connection to the heater. Adequate steam traps are necessary in the condensate lines from the heater since flooding of the heater will have an advem effect on its performance. Sectionalized heaters or banks of heaters are commonEy used in order to provide the flexibility of using all or a portion of the heating surface over a wide range of heating capacity (from minimum port steaming conditions to 120-percent ovedoad). With such an arrangement, individual heaters or sections of a heater can be cut in or out as the demand fluctuates and thereby maintain a steam supply suflicient to ensure adequate control of the oil outlet temperature. Such a control of the amount of heat transfer surface in service avoids cyclical heater operation wheh the heating requirements are very small as compared with the effective heating
517 I
I
I
I
SECONDARY
The design requirements for lubricating-oil purifier heaters are much like those for fuel-oil heaters in that the oil n~ustbe heated to a prescribed twperature range (normally 100 to 160 F) in order to attain a sufficiently low oil viscosity for effectivepurification. Lubricatingoil purifier heaters are generally of the tubular type and are similar to those used as fuel-oil heaters.
I
3-3 Boiler Feedwater and Desuperheater Leakage Test Sample Coolers. Daily tests of the condition of the
fig. 28
Bayonet-tube fuel-oil heater
boiler water and feedwater are necessary to ensure continued efficient operation and protection of highPressure steam generators. These tests are necessary to : Maintain the specified boiler water chemistry through chemical treatment to ensure that the correct proportions of the essential chemicals are present. Check the effectivenessof the blowdown procedure by the concentration of the soluble and suspended solids in the boiler water. Determine the amount of dissolved oxygen in the boiler water and feedwater to guard against excessive The collection and cooling of water samples is the first in the test procedure. The coolers required
mi. 29
Boiler water sample coder
are relatively s m d heat exchangers due to the small quantity of sample requifd for testhg. One of the common arrangements is a cooler system consisting of primary and secondary coolers connected in series. The coolers have a cylindrical cast bronze shell that contains a helical mil wound around a core positioned in the c a t shell; such a cooler is shown in Fig. 29. The sample p=es through the coil and cooling water flowa across the outer surfaces of the coil. The boder water sample cooler coil is usudy made, of 90/10 or 70/30 CuNi and
MARINE ENGINEERING
HEAT EXCHANGERS
66
519
Fig. 30 Saltwater heater and drain cooler
the cooler and valves are designed to the boiler working cooler uses seawater at 85 F as pressme. The a and reduces the sample temperature to cooler uses chilled approximately 140 F; the 50 F fresh water as the coolant to reduce the temperature of the sample to that desired for testing. The maximum temperature for oxygen determination is 70 F; in Navy practice100 F is considered the maximum temperature euitable for pH, hardness, and chloride determination. The coils far desuperheater leakage test sample coolers can be madeof copper or a similar material since they me designed for a'rnoderate pressure and are normally cooled with fresh water. 3.4
Tank
Cleaning System
Heaters
and
Drain
Coolerr. When a different grade of oil is to be carried in a tanker (e.g., gasoline vice crude oil), it is necessary to wash the cargo oil tanks before receiving the different grade. The tank cleaning systems use hot seawater supplied from heat exchangers. Cargo vessels utilize a smaller system than tankers for tank cleamgThe heat, exchangers usually consist of a seawater heater rwlddrain cooler connected in series SO that the seawater flows first through the drain cooler and then through the heater. Steam is used as the heating medium in the heater, and the resulting condensate is cooled in the drain cooler. The seawater is heated to a temperature of 180 to 200 F. As illustrated in Fig.30, the exchangers are typically of the horizontal shell-and-tube type; the heater (above the d r h cooler) is of the U-tube type, and the drain cooler has a fixed tube nest and a shell expan~onjoint. The tube nest in the coolers is usually fitted with transverse baffles to create a flow path for the drains perpendicular to the run of the tubes. Depending upon design
conditions and installation requirements, the tubes may be arranged for single- or m ~ l t i - ~ a flow s s of the seawater. Zinc anodes are provided in the water heads to minimize galvanic corrosion on other parts of the heat exchanger (galvanic corrosion is discussed in section 1.1). Occasionally the heater and drain cooler are combined preclude the in a single shell, but the arrangement possibility of the condensate rising above its normal level and submerging the tubes in the heating section, which would reduce the effectiveness of the unit. In some cases, only a heater is installed and the drains are discharged through a steam trap to the deaerating feed heater or a suitable vessel or receiver. Where a separate or combined heater and cooler is installed, a liquid level control is employed to ensure submergence of the drain cooling tubes. Relief valves are provided on both the shell and tube sides of the saltwater heater. For naval applications, there are a number of special design requirements (such as construction materials and shock requirements) which must also be considered [16I. 3.5 LOW-PressureFeedwater Heaters. h he classification (i.e., low pressure ar high ~ressure)of feedwater heaters depends upon their location relative to the boiler feedwater pump; low-pressure heaters are located on the suction side of the main feed pump, whereas highpressure heaters are located on the discharge side. Feedwater heating is accomplished in a number of steps or stages and the heaters are usually referred to as the first stage, second stage, third stage, fourth stage, etc. Multiple stages of feed heating are essential to the efficiency of a steam turbine power ~ l a n t ,as may be t balance estabnoted from Chapter 2. The ~ l a n heat lishes the number of heating stages, feed flow through
I each point and a k l i a r exhaust ~ pressures, furnished with the glsnd-exhauster fan and and the temperature of the feedwater entering each drain r e d a t oa rpackage mounted on the unit. Due to the of Since the heat transfer coefficient of condensing steam combined heattemperature exchangers,difference the outletbetween ends ofsections the tubes
is
of velocity, and feed pressures are usually quite high, the feed is generally in the tubes with the in the For a given steam Pressure, the heat is dependent upon the feed velocity through the tubes. of to 7 fps result in a reasonable pressure and satisfacto~heat h n s f e r conditions. The heater should avoid the Occurrence of dead spaces, and drain cooling sections should hold 'lose baffle-to-shrOud tolerances so as to avoid excessive bypasing beat transfer surface which would result in insdequate drain caoling. Both the shell and water sides should be self-venting. The pre-e CQntr0l.i the constlTl~ti0ndetails feedwater heaters with design pressures UP to PS~Fare considered low-pressure heaters. It is common practice to combine Several lowpressure heaters one shell to save space, C O S ~of equipment, piping, and installation costs. Figure 31 depicts a ty~lcalcombined low-pra4sure feed h e a t e d r a i n cOO1er/dand-exhaust condenser. similar heaters combing two Stages of h e a t h with the drain cooler are sometimes used. The heater is normally
me u s u d y secured in the tube sheet by of alternate rings of metalic and fibre packing (see 1,-ig. 8 ( ~ ) and ) are therefore free to expand independently. 3.6
Dim.Montact
Dwerating Feedwabr ~
~
Since marine boilem are operated at high temperatures and pressures, there is a hazard of attack due to the presence of dissolved oxygen or carban diofide in the feedwater. It is ~ t u a l l y impossible to prevent the entry of air into the feed system, particularly during plant st&tup; therefore it is necessary to provide deaerating equipment for the removal of air and co~osive gases from the boiler feedwater. Although &aeration can be largely acomplbhed in the condenser, ucondenser deaeration' is not suffci$n&during @ant stertup; and without further &aeration there be no-provision for the removal of air introduced later in the system, particularly a t the condensate pumps. Flash &aeration, whereby saturated water at 10 to 15 psig is introduced into a surge tank at atmospheric pressure, is simple and economical. H ~ the ~ resulting flashing of steam will not ensure the low dissolved oxygen content (i.e., 0.005 cc per liter) reqUired
~
HEAT EXCHANGERS
MARINE ENGINEERING
VENTED STEAM CONTAINING OXYGEN REMOVED FROM WATER
WATER SUPPLY WATER SPRAY
VENT CONDENSING WATER SPRAY NOZZLE PRIMARY HEATING
AND DEAERATION CHAMBER
SPRING LOADED ATOMIZING VALVE
Fig. 32
Direct-contactfeed heater
of the feedwater unless suflicient agitation is positively provided. There is also a loss of steam through flash deaeration. For example, a deaerator operating at 15 psig flashing down to atmospheric pressure loses about 4 percent as flashed steam. The flashed steam should not be condensed and returned to the feed cycle since it will normally have re-entrained a portion of the undesirable dissolved gases. Therefore, standard marine deaerators are more sophisticated in design than those of the flash type. The practical considerations involved in the removal of dissolved oxygen from boiler feedwater may be briefly summarized as: 1 Heating the water to the boiling temperature for the pressure under which the process is conducted (satura tion conditions). From the chemical relationship termed "Henry's Law," it is known that when a partial pressure of a liquid is equal to the total pressure above the liquid
(boiling conditions), the solubility of any gases in the liquid is zero. 2 Providing a design that ensures thorough agitation and scrubbing of the feedwater by the steam. Complete agitation of the feedwater and contact with the scrubbing steam ensures that equilibrium will be reached and that the zero potential solubility condition (Henry's Law) will be attained. 3 Continuously venting from the system a mixture of gases and steam. Through the use of adequate venting, the partial pressure of the noncondensable gases in the system will be kept low and the saturation boiling point of the liquid will be maintained. The heater immediately preceding the suction side of the boiler feed pump is usually the "direct contact" or deaerating feed heater (generally known as a "DFT" for deaerating feed tank). A typical direct-contact feed heater is illustrated in Fig. 32. Condensate and
makeup are sprayed into the steam-filled primary heating and deaeration chamber through a series of spray nozzles and a vent-condensing spray nozzle. The spray nozzles provide an even distribution of water over the entire heating area. The steam flow, which- is essentially counter-current to the water flow, heats the water close to the saturation temperature such that the solubility of the gases is zero, and approximately 95 percent of the oxygen content is thereby released. Water and condensate collect in the conical water collector and flow to the atomizing valve, where highvelocity steam strikes the mixture, atomizes it into a fine mist, and raises the temperature the last few degrees to its saturation point. The mixture strikes a deflecting baffle which separates the water and steam. The hot gas-free water drops to the storage compartment. The complete atomization and heating of the feedwater by the steam jet ensures that the dissolved gases will be released. After the atomization process, the steam and released gases flow through the primary heating/deaerating chamber where a large portion of the steam is condensed as it heats the incoming water. A small portion of the steam and all of the gases pass through the integral vent condenser which condenses the majority of the remaining steam. The small amount of steam vapor that is mixed with the released gases is then discharged to the atmosphere or to the gland leak-off condenser. The deaerator conditions feedwater such that its dissolved oxygen content is less than 0.005 cc per liter. In addition, it substantially reduces the carbon dioxide content of the feedwater. Since the feedwater is at saturation temperature and above atmospheric pressure, the arrangement of the deaerator is of great importance as there is a strong possibility of the feedwater flashing into steam at the pump suction. There are two means of ensuring an adequate suction head at the main feed pump. One is to position the deaerator high in the maqhinery space; or an alternative is to provide a booster pump between the deaerator and feed pump which will maintain an adequate suction head on the feed pump. Damage control considerations dictate that the booster pump arrangement be used in naval ships. The booster pump must be designed to handle condensate at saturation temperature, and it is important that the booster pump suction line be short with little or no turns and adequately vented so that pump cavitation and suction line flashing will not occur. The alternative of locating the deaerator high in the machinery space is the preferred arrangement with merchant ships as a pump is eliminated and damage control is not a design criterion. Proper performance of a deaerator requires correct sizing of components and control of the rate of flow to the storage tank portion of the deaerator. The first major consideration is the boiler steam output. This determines the size of the deaerator and affects the storage tank, the makeup valve, the transfer pump, and the number of water spray nozzles in the unit. The other
52 1
major factor is the temperature of the water delivered to the spray nozzle; this temperature determines the size of the steam-regulating valve which admits steam to the deaerator. This valve is sized as closely as possible to furnish the quantity of steam required to maintain the deaerator a t the operating temperature, plus about 10 percent additional capacity/of steam over that required to heat the inlet water at the design conditions as a safety margin to handle surges of incoming feedwater. However, sincpc the steam capacity is considerably affected by the pipe size of the regulator and the incoming steam pressure, it is difficult to provide a valve that exactly matches the desired capacity. Proper deaera tion requires that the temperature.of the incoming water be raised to the saturation point; therefore, the volume of the inlet water must be controlled in relation to its temperature to stay within the heating capacity of the steam supplied by the steam-regulating valve. An excessive flow of cool water will, of course, quickly condense the steam in the deaerator, making it difficult to maintain the desired pressure. This emphasizes the necessity to provide an adequate safety mapgin in sizing the steam-regulating valve so that its capacity and response rate are capable of handling surges of cool water. The storage tank is usually selected to retain about five minutes of storage. If high-pressure (high-temperature) returns are available, they may be returned directly to the deaerator storage tank. Here they will flash and provide a certain amount of steam for preheating the water introduced into the dyapator. If these returns exceed 25-30 percent of the total capacity of the deaerator, more steam will be available than is needed and some other means must be employed to use the returns. Deaerating feed heaters are normally equipped with two spring-loaded relief valves: one to prevent a high pressure from accidentally building up within the tank; and the second, known as a vacuum breaker, to prevent a high vacuum from developing in the tank by allowing atmospheric air to enter the tank in the event that the pressure in the tank drops below a prescribed value. The shell and majority of the internals are normally of welded-steel construction; however, the 'steam baffles, spray nozzles, atomizing valve, and vent condenser are generally manufactured from nonferrous alloys or stainless steel. , 3.7 High-Pressure Feedwater Heaters. A high-pressure feedwater heater may consist of one, two, or three sections (a three-section heater contains desuperheating, condensing, and condensate cooling sections). All sections are normally integrkkd in oneshell fpr conipactness and simplicity of piping. In addition to the marine regulatory body design requirements [3, 4, 51, the code requirements of the Feed Water Heater Manufacturer's Association [18] are often applied. The construction features of a typical high-pressure feedwater heater are shown in Fig. 33. The tubes are usually N-in.-OD tubes, arranged on a 1x6-in. triangular pitch. The tubes in the condensing
I
I
1 I
I I
I I
I
~
I
I I
I
I
~ II
1
I
522
MARINE ENGINEERING
1
I (a) HORIZONTAL INSTALLATION
1. CHANNEL 2. CHANNEL COVER 3. STATlONdWY TUBE SHEET 4. TUBES 6. TUBE SUPPORT 6.SHELL SKkRT 7. PARTmKlN PLATE IL PARTITLON COVER S. TRANSVERSE BAFFLES 10. tMPINGEMENl PLATES 11. DESUPERHEATIMG ZONE 12 SUBGOOLING ZONE 13. SHELL 14. HEATER SUPPORTS 15. FEEDWATER INLET 16. FEEDWATER OUTLET 17. DRIP INLET la STEAM lNLET r a CONOENSATEOWLET 20. WELL RELIEF VALVE CONNEmlm 21- TtlEE VALVE CONNEmlON - SIDE -.- - -RELIEF .- a.LIWLD LEVEL CONTROL CONwmlONS % *G GLASS CONNECTIONS 24. OPERATING AIR VENT Q)NNEmIONS
- - - --
(M VERTICAL INSCALLATIOM
Fig.
33 Typical fedwafer haatam
section &t supported by plates spaaed at intepvak n d exmetkg 48 in. to avoid tubs vibration Beetions me cut out of the support plates to pmvide passages for steam flow and drainage. If the steam which enters the shell side of the feedwater heater is hi&ly superheated (e.g., a superheat of 100 deg F or above), the tube amface in contact with the superheated steam will have a wall temperhigher than the saturated steam temperature; this lneans that the tube wall will not be wetted by condensate and thst the transfer of heat will be low d e s s special preeauticr11~are made. It is found economical sometimes essential to ban a desuperheatbg section to mntml the desugerhesting of the steam. The desuperhestiing section is h t e d at the feedwater exit end so that the leaving feedwater can be heated to
the highest possible temperature. By arranging tho
Row this way, the feedwater temperature may eve11 exceed the &teamsaturation temperature in the desuperheating section. The desuperheating section consists of a shroud wrapped around a group of the tubes so as to confine the inlet steam. Cmss baffles are provided with the shroud to decrease the dry-vapor thermal resistance. Other design features incorporated in the desuperheating section are means to shield the other regions of the unit from the high-temperature steam and means to prevent distortion due to unequal temperature distribution. The eonde~mate-coolingsection is Located at the feedwater inlet end of the beater so that the condensate (or drains) from the condensing section can be subcooled to approach the feedwater inlet temperature. The cotl-
i
523
HEAT EXCHANGERS
densate cooling section also consists of a shroud enclosing a portion of the tubes and cross-baffle plates. When designing the condensate cooling section of the unit, the possibility of condensate reheating must be considered. Condensate reheating is caused because the steam condensate inside the shroud, while being cooled by the feedwater inside the tubes, is dso heated by the steam condensing on the outside of the shroud. This reheating of the condensate is a matter of great importance at the drain outlet end of the heater where the temperature difference between the condensate and feedwater is often as low as 10 degrees, while the difference between the condensate and steam outside the shroud can be as high as 100 degrees. Several means can be taken to avoid excessive reheating of the steam condensate. One would be to increase theratio of the condensate cooling tube surface area to the shroud area; another would be to insulate the shrouds. The attitude of the feedwater heater as it will be arranged aboard ship must be established before the thermal design of the feedwater heater can commence because the various alternative arrangements impose different restrictions on the thermal design of the unit. When the unit is installed in a vertical position with the feedwater entrance and exit channel (or waterbox) on the top, as in Fig. 33(b), the bottom region of the shell can be used as a steam condensate collector. The shrouded condensate cooling section extends the full length to the top of the shell in this instance, and the height 'of the unit is relatively short. A vertical arrangement with the feedwater entrance and exit channel at the bottom [the inverse of that shown in Fig. 33(b)] is normally selected for long units that are designed for outdoor land installations; however, a lack of space usually precludes its application in marine plants. With this type of arrangement, the unit is designed such that the steam eondensate exitcl at the feed inlet end in order to take advantage of the colder feed temperature for cooling. To accpmpliish this, it is necessary to flood a portion of the tubes with steam oondensate, which results in poor ueilization of part of the heat transfer surface. When it is possible to do so, high-pressure feedwater heaters should be arranged horizontally aa illustrated by Fig. 33(a). Compared with a vertical position, a horizontal arrangement affords the following advantages: There is less restriction on length. A heater of longer length normally results in a smaller shell diameter and a more economical unit. The ieheat problem is less severe in the steam condensate-cooling section because this section need not be exposed to the steam and a short condensateoooling section can be used to achieve the proper proportion of the tube surface to shroud surface. w A higher condensing heat transfer coacient is achieved on a horizontal tube bundle than on a tube bundle that is vertical. Tbis fact is not reflected in most design analyses; however, it should be considered
(a) Torrgueand-Growe Flange Joint fig.
lbl Sheer-block d a r e
34 Tubedde cfoums
by the designer so aa to provide the most effective utilization of heat transfer surface. Of major importance in the design of high-pressum feed heaters is the adequacy of the c l o m s for the pressures and temperatures involved. A breakable joint should be provided for the shell-side c l ~ soethat the shell can be removed from the tube bundle for inspection and cleaning purposes. Under the usual operating conditions, a bolted flange joint ifil suitable for this purpose. However, under conditions which would involve temperature distortion or when there is an infrequent requirement for removal of the shell, a welded joint provides positive sealing lihd is economical to fabricate. A back-up ring is provided to protect the tubes when a flame cut is made to open the welded joint. The feedwater that enters the channel and the tube side of the unit is under a relatively high pressure which imposes a severe requirement on the tube-side closure and seal design. The bolta of a flanged joint are required to take the hydrostatic load (whieh depends on the closure diameter and fluid pressure) and at the same time maintain a pressme on the gesket s d c i e n t ts ensure a seal. This often results in huge bolts which require enormous torques to tighten, e q e d l y so when the larger shell diamet~?rsme involved with pressures over 1200 pig. Nevertheleae, bolted-flanged joints can be properly applied in the design of heaters with diameters as 1-e as 20 inches and for pnxmes of less than lux, psig. Flat metal or metal-jacketed gaskets are frequently used with bolted tube-ride closures. The force required to adequately compress the gasket is a mbt~dltiat percentage of the h y load ag~the end .clow~e ~ and may even exceed it. The force required toobtain an adequate gasket seat can be reduced by narrowing the width of the gasket and at the same time confining the gasket to an enclosed space to prevent the gasket from deflecting freely. An example of this design feature is illustrated by the tongue-and-groove hnged joint in Fig. %(a). A shear-block closure design that is used in some of the
MARINE ENGINEERING
larger and higher-pressure feed heaters is depicted by F i p . 34(b) and 35. The hydrostatic load is resisted by the shear ring, and the pressure on the gasket is maintained by the hydrostatic load. Figure 35 shows the method by which the tube-side operating pressure is used to seat and seal the solid copper ring gasket. With the larger high-pressure heaters, a point is reached where a simple bolted-flange closure must give way to a more elaborate high-pressure closure; experience indicates that when the product of the operating pressure (in psi) and the shell I D (in inches) exceeds 25,000, a bolted-flange joint is no longer economical. In actual design practice, the availability of a standard flange for the size and pressure intended often decides the question as to the specific closure design. The economic advantage of a ready-made versus a custom-made item will often influence the selection of the specific closure. 3.8 Gland Leak-Off Cqndensers. I n order to avoid an ingress of air into the steam system at the points where the steam turbine shaft penetrates the turbine casing, and similar locations, a gland-sealing steam system maintains a pure steam atmosphere at a pressure slightly above atmospheric just outside the turbine shaft-casing interface; this ensures that atmospheric air will not enter the turbine. The gland leak-off system consists of a fan which removes an air and steam mixture from the turbine gland leak-off pockets, and a condenser through which the mixture is drawn in order to condense the steam so as to recover the water and reduce the quantity of gas which the fan must handle. The gland leak-off condenser may be either furnished as a separate heat exchanger or combined as a section
HEAT EXCHANGERS
of the low-pressure feed heater (see Section 3.5). When it is a separate exchanger, it is of the U-tube design and is arranged to receive low-pressure steam in the shell, and cooling water is passed through the tubes to effect the necessary condensing of the steam. The fan which forms a part of the gland leak-off system is designed to handle the leakage air plus the uncondensed steam vapor. The gland leak-off exhauster (which is the common name for this fan) is usually mounted on top of or immediately above the condenser (or condenser section of the first-stage heater). The Navy has set forth specific design and material requirements for this condenser and similar condensers for other shipboard applications [17]. 3.9 Unflred Steam Generators. Unfired steam generators (which are also referred to as contaminated water evaporators or steam service evaporators) supply lowpressure steam at a pressure of 50 to 150 psig to a system that is independent of the main steam system. The independent system provides steam for services which could possibly contaminate the main system in the event of a system malfunction. Some of the "contaminated" services include the fuel-oil suction and service heaters, cargo-tank heating coils, galley and heating systems, and steam-driven deck machinery. The contaminated evaporator tube nest drains are normally piped to the deaerating feed heater via a trap. Figure 36 depicts a typical contaminated steam system. Contaminated evaporators operate on bleed steam from the high-pressure turbine, auxiliary steam, or in some cases, high-pressure steam (up to 500 psig). The bleed steam is a variable-pressure source which depends on the percentage of full power being developed. The requirements of the contaminated system also vary greatly. Operational requirements, as depicted on a normal-power plant heat balance, may be only a small percentage of the maximum performance requirements.
PRESSURE REDUCING (TYPICALLY 600 TO 150 PSIG)
SAFETY VALVE
FROM DESUPERHEATER'
, ' STEAM LINE
TO D.C HTR.
ORIFICE
*ONE OF MANY POSSIBLE FEED CONTROL ARRANGEMENTS FLOWS TYPICAL FOR NORMAL OPERATION 1
CONTAMINATED DRAIN TANK Fig. 36 Typical contaminated steam system
The maximum performance requirements must be taken as the design conditions with checks niade to ensure that the design is satisfactory for other operating arrange-
LOW PRESSURE STEAM OUTLET
SPLIT RETAINING RING
I N N E R COVER
POSITIONINGR I N G
(J SHEAR-BLOCK
TYPE CHANNEL COVER
Two sets of design conditions are generally set up: one for bleed-steam operation and one for auxiliarysteam operation. Both conditions must be considered to determine the effectson the evaporator design as well as the safety valves and orifices. The evaporator bundle is normally of the U-tube type as shown by Fig. 37. When the tube bundle becomes very large, straight tubes should be used, incorporating the outside-packed head type of construction illustrated by Fig. 38. Tubes are usually %-in. OD on a lxs-in. square pitch, except in the case of a low temperature difference when a closer pitch may be used (N in.). H-in.-OD tubes, when used, are placed on a 1%-in.
BLOWWW Fig. 37 Typical contaminated water evapaata
(b) CONSTRUCTION DETAILS OF SHEAR-BLOCK TYPE CLOSURE
Fiq. 35 shear-block closure
water gage glass, steam supply r e d a t i n g valve, ther-
The overall clean-tube heat transfer coefficients that
526
HEAT EXCHANGERS
MARINE ENGINEERING
I
are used to determine the evaporator bundle size are as
(T,*,, - t , ~ ) ... . . . . . . . Overall heat transfer coefficient .............
<50
deg F
I100 deg F
>laO deg F
550
520
500
OVERFLOW
A steam flow orifice is usually installed in the steam supply line to limit the amount of heating steam entering the evaporator and reduce the steam supply pressure (a high heating steam pressure is not always beneficial due to the critical heat flux or the vapor blanketing phenomenon iri boiling). The orifice size is determined from the Fig. 38
=
105f[P.(Pl - P2)I1/'
if P2
Outside-packed head construction used on large tube bundler
> 0.58P1
a drain to the bilge. A probe for oil detection and an associated alarm are also normally provided. The filter section is packed with a loose filtering medium, usually loofa sponges, arranged for easy access and replacement of filter material. A, = orifice area, sq in. The storage section is provided with a gage glass, P1 = upstream pressure, psi thermometer, removable cover, cooling coil, a drain and P2 = downstream pressure, psi overflow to the bilge, and a vent. W#,,= steam flow rate, lb/hr 3.1 1 Hot-Water Heaters. Hot-water heaters are The safety valve should be sized for the maximum used to provide water at controlled temperatures for showers, lavatories, and galley services. Hot-water steam flow entering the generator, i.e. heaters are generally steam heated and can be classified into two types: storage heaters and instantaneous heaters. W = 105.3 Ao[P4(Pa - Pr)]1'2 A storage type of heater has the advantage that the water can be heated and stored during nonpeak periods of hot if P4 0.58Pa W = 51.45 AZa, water demand, thus reducing the peak heating-steam requirement. Also a less sophisticated temperature control device is required for a heating rate that is Pa = Pressure of line (Assume independent of the rate of hot water withdrawal. The if P2
< 0.58P1
<
P,
= set pressure of shell safety valve
however, the heat loss is rarely a deciding factor. Space The other major item in the contaminated system, and weight are usually a t a premium aboard ship, in namely the drain inspection tank, is covered in Section which case an instantaneous hot water heater, which heats the water as fast as the rate of withdrawal, should 3.10 Contaminated Dmin Inspection Tank. The be considered since a large storage tank is not required. Figure @(a) illustrates the storage type of hot-water contaminated drain inspection tank is used in the heater and its accessories. The notable features of this contaminated-steam drain collecting system to act as the receiver for contaminated drains from fuel oil, lube oil, heater system are the pump which circulates the water and oil-tank heating systems. This system is kept from the tank to the heating element, and the thermal bulb [labelled "remote bulb" in Fig. 40(a)] located at the isolated from the clean-steam drain system. Normally one drain inspection tank is furnished per outlet of the heating element. It can be seen from this system; the tank consists of a welded-steel (at least % in. arrangement that the heat input to the system is thick) fabrication that issuitably stiffened and galvanized independent of the rate of hot water withdrawal. Tho after assembly and is fitted with a cooling coil to prevent flow circulation rate, together with the thermal bulb and h h i n g of hot drains. The tank, as illustrated by Fig. the temperature of the water to be circulated (i.e., 39, is sized to accommodate all heating drains a t peak temperature of the water at the lower part of the tank, load conditions and is usually divided into three sections which may or may not mix with the inlet cold water), will dictate the heat input rate through the action of the (i.e., storage, filter, and inspection sections). The inspection section is fitted with an inspection port control valve. The control valve regulates the amount with a light set a t the waterline. It is also fitted with a of steam admitted as well as the steam pressure. If tho scum drain, test connection, removable cover, vent, and water circulation rate is sufficiently larger than the peak
Fig. 39
Contaminated drain impadion tank
hot water withdrawal rate, then the storage tank is not flow friction inside the heating tubes, and therefore is necessary and the storage type of heater becomes an flow rate sensitive. Consequently, the thermal bulb, instantaneous heater. which ac$uates the steam control valve, senses the The most complex aspect of a hot-water heating system combined effect of the hot-water outlet temperature, the lies in the scheme of temperature control. Figure cold-water inlet temperature, and the hot-water with@(b) illustrates the principles involved in controlling the drawal rate; as a result the time lag of the steam controltemperature of a typical instantaneous hot-water heater. valve action is minimized. The shrouds around the The flow bypass gate, bypasses a small portion of the thermal bulb point out w t h e r crit&al aspect .of the inlet cold water, and the bypassed cold water is directed design; namely, that the flow around the sensing element to flow over the thermal bulb (or temperature-sensing should be guided to produce an accurate signal that gives element) located in the water outlet channel. The a prompt indication of an incipient temperature change. amount of cold water bypassed depends not only upon The permissible variation in the hot-water outlet the adjustment of the bypass opening but also on the temperature should not be over-specified. A hot-water pressure difference between the inlet and outlet sections outlet temperature variation of 10 deg F will result in a of the waterbox. The pressure difference between the hot-water heater system of less expense and complexity inlet and outlet sections of the waterbox is created by than would be the case if a 5 deg F variation were
i
528
HEAT EXCHANGERS
MARINE ENGINEERING
VENTS
INLET la) Stomp Type of Hot-Water Heater NULUS BETWEEN TUBES
\
HEATING STEAM
Fig. 41
FLOW BY-PASS GATE
COLD WATER INLET
fig. 40
Hot water heaten
specified. However, there is a tendency for hot-water heater procurement specifications to stipulate an outlet temperature control of plus or minus 5 deg F at varied flow rates [IS]. 3.12 Fail-safe Heat Exchangers. Some heat exchanger applications require a design that precludes the entry of one fluid into the other in the event of a tube or tube sheet leak. For example, the contamination of condensate systems by seawater or fuel oil must be, prevented. Electronic equipment coolers and hydraulic system oil coolers are also areas where "fail safe" heat exchangers are used. The "evaporator" type of fuel-oil heater whereby an intermediate fluid is used to transfer heat from a primary heating coil to the oil-heating elements can be used in lieu of conventional oil heaters and thereby preclude the contamination of primary steam condensate with oil. An evaporator section with a primary heating coil mounted
on its cover is bolted to the heater in place of a steam chest cover. Navy applications consider such arrangements since the general lack of "contaminated" systems on Navy ships increases the necessity to take adequate precaution to prevent oil contamination [15]. Double-tube-sheet and double-tube, double-tubesheet exchangers are often specified for Navy applications where a "fail safe" feature is desired [6]. Figure 41 is a cutaway view of a double-tube, double-tubesheet heat exchanger as provided for hydraulic-oil cooling aboard submarines. Any leak that develops at the tube-to-tube-sheet joint or as a result of a tube rupture will flow to the void between the double tubesheets and out the drain. Alarm devices may also be incorporated to automatically warn of leakage into this void space. Similar designs with only the double-tubesheet feature are used on seawater-fresh water coolers to reduce the risk of saltwater contamination.
Double-tube double-tube-sheet heat exchanger
References 1 'Standards of Tubular Exchanger Manufacturers Association, " Tubular ExchangerManufacturers Association. 'lASME Boiler and Pre8sure Division I, Section VIII, American Society of M e c h c Engineers. 3 Rules and Regulations for the Construction and ClasSiJication of Steel Ships, Lloyd's Re'gister of Shipping. 4 Code of Federal Regulations, Title 46-Shipping, Chapter I-Coast Guard, Department of Transportation, Subchapter F-Marine Engineering, published by Office of the Federal Register. 5 Rules for Building and Classing Steel Vessels, American Bureau of Shipping. 6 ucoolers, ~ l ~ i d~ ,~ d ~ ~ t~~~~l ~ i a shipboard l , Lubricating Oil, Hydraulic Oil, and Fresh Water," Military Specification MIEC-15730. 7 R. A. Bowman, A. C. Mueller, and W. M. Nagle, "Mean Temperature Difference, in Design," Trans. ASME, May 1940. 8 E. N. Sieder and G. E. Tate, "Heat Transfer and Pressure Drop of Liquids in Tubes," Industrial and Engineering Chemistry, Vol. 28, 1936. 9 Daniel A. Donahue, "Heat Exchangers," A Special Petroleum Processing Report, March 1956.
10 Kenneth J. Bell, " ~ i n i i l ~ e ~ofothe r t Cooperative Research Program on Shell and Tube Heat Exchangers," University of Delaware En+eeriog Bulletin No. 5, January 1963. 11 Townsend Tinker, ''Shell Side Characteristics of Shell and Tube Heat Exchangers-A Simplified Rating system for Commercial Heat Exchangers,,, Trans. ASME, 1958. 12 Devore, Try T~~~simplified Method for Exchangers,,, May Rating 1961. 13 D. L. Katz, J. G. Knudsen, G. Balekjian and 8. S. Grover, "Fouling of Heat Exchangers," Petroleum ReJiner, 1954. 14 H. H. Keller and E. V. Somers, "Heat Transfer from an $Annular Fin of Constant Thickness," Tram. ASME, 1959. 15 "Heaters, Fluid, Fuel Oil, Naval Shipboard," Military Specification M1L-H-16313. 16 "Heater, Fluid, Seawater; Shipboard Use, " Military Specification ~1~-B?T6423. " 17 "Condensers, Steam, Surface, Naval Shipboard," Military Specification MIL-C-15430. 18 "Standard of Feedwater Heater Manufacturers Association, Inc., " Feedwater Heater Manufacturers Association.
DISTILLING PLANTS
CHAPTER XV
BACK PRESSURE VALVE SET TO MAINTAIN VAPOR PRESSURE I N EVAPORATOR SHELL
( Distilling Plants
VAPOR T ~ M A I NOR
/ AUXILIARY CONDENSER
CIRCULATING WATER OUTLET
Section 1 Distilling Plant Designs 1.1 Introduction. Distilling plants are used aboard ship to produce high-purity distilled water from seawater. They are also used to produce distilled water from raw freshwater supplies of insufficient or doubtful purity. The capacity of marine evaporating plants varies from a few thousand gallons per day (24 hours) up to 100,000 gallons per day or more, depending upon the size and purpose of the ship. Distilled water is required to furnish high-purity makeup water for the boilers and potable water for drinking, cooking, dishwashing, ablutionary, hospital, and laundering purposes. Additional capacity is often incorporated to enable the application of freshwater sanitary systems. The standard Navy specifications and U. S. Maritime Administration specifications require a distillate purity of less than 0.065 equivalent parts per million of chlorides. This is equal to of a grain of sea salt per gallon, or aboub 4.3-ppm total dissolved solids. Most marine distilling plants can produce distillate substantially below these limits. The specifications for nuclear-powered naval ships require a distillate purity of one half that of the normal specifications, or % grain of sea salt per gallon. Tests on marine distilling plants demonstrate that the units are capable of producing water containing less than x-ppm total dissolved solids with purity determinations made by means of the sodium tracer method and a flame spectrometer. When operating in normal seawater of 35,000-ppm solids (a definition of "normal" seawater is given in "The Oceans" by Sverdrup [I ]I), a distilling plant should be capable of operating for a period of a t least 90 days a t rated capacity without shutdown for cleaning. Depending on the specific design, this can be accomplished with or without chemical feed treatment. Such extended operating periods without shutdown for cleaning is a requirement of both the standard Navy specifications and U. S. Maritime specifications for $hipboard distilling plants. "Normal" seawater is found in most areas of the Atlantic and Pacific Oceans. 1.2 Early Distilling Plant Designs. As shown by Fig. 1, the single-effect distilling plants designed before World War I1 consisted essentially of an evaporator to Numbem in brackets desigmte References at end of chapter.
evaporate the raw water by means of steam-heated tubes or coils, a condenser (called a distilling condenser or more simply a distiller) to condense the vapor, and the necessary interconnecting piping, valves, etc. Multiple-effect plants were generally used for the larger capacities so as to reduce to reasonable proportions the amount of operating steam required. In multipleeffect plants, the vapor from the first evaporator (firsteffect evaporator) is condensed in the coils of a second evaporator (second-eff ect evaporator), thereby serving as the heating medium, etc., with the vapor from the last effect only being condensed in the distiller. Figures 2 and 3 illustrate the two types of submergedtube evaporators commonly used in the early distilling plants. The horizontal-tube evaporator shown by Fig. 2 required conservative design criteria in order to attain a level of performance considered acceptable in its day. Below brine temperatures of 130 F the fouling of tube surfaces was minimal, and for brine temperatures up to about 180 F the scale formati011 was not considered excessive provided the temperature gradient across the heating tubes was maintained below 50 F. At higher temperatures, scale formed rapidly such that vapor temperatures above 250 F were not considered practicable from an operator's viewpoint. "Cold shocking" or thermally cracking the scale formation from heating tubes by suddenly changing the temperature of the tube metal was a technique used to minimize the extent of marlual tube cleaning required. Cold shocking greatly improved the heat-transfer rates of the tubes, especially for the so-called self-scaling tube designs which were in the form of helical coils or bowed tubes with restrained ends such that the tubes were distorted by a temperature change. However, the coldshocking process was only partially effective. Scattered patches of scale were left on the tubes and they sometimes built up to a considerable thickness before cracking off. The heat-transfer rates obtained after cold shocking were considerably less than for tubes that had been thoroughly cleaned. The high maintenance associated with submergedtube evaporators was recognized and, insofar as practicable, design provisions were made to facilitate their care and cleaning. The vertical-coil evaporator in Fig. 3 includes such features as a track for ready removal of the coil assembly, removable heating coils, and heating tubes
SAFETY VALVE EVAPORATOR
PRESSURE REDUCING VALVE TO MAINTAIN STEAM PRESSURE I N EVAPORATOR COILS
SHELL VAPOR, PRESSURE GAGE
DISTILLER
COIL STEAM PRESSURE GAGE
GAGE GLASS TO INDICATE WATER LEVEL I N SHELL
LINOMETER POT FOR BRINE SAMPLING
FF ELL
CIRCULATINQ WATER INLET
DISTILLATE OUTLET
STEAM TRAP 6 $ H RV.PACS
PET COCK FOR AIR VENT, TO BE KEPT OPEN WIDE ENOUGH TO PREVENT AIR BINDING IN COILS
Fig. 1
Early single-effect submerged-tube distilling plant arrangement 1
!
in the shape of a helical coil. However, despite the efforts to improve the design of submerged-tube evapom tors, they continued to be characterized by high operating costs, high space requirements, high maintenance costs, complex piping systems, and low performance under continued service conditions. The state-of-the-art of distilling design has, however, changed considerably since World War 11. Since World War 11, a variety of distilling plant designs has been developed from which the most appropriate system for a given marine power plant can be selected. I n lieu of the previously predominant distilling plant8 that employed submerged-tube evaporators, the distilling plant dwigns considered for marine use emerged as : Plants employing flash evaporators Planta employing thin-film evaporatora Plants employing the vapor compression cycle Plants employing basket-type evaporators Plants employing a membrane process 1.3
Flash Evaporator Designs
a. Marine E'lash Evaporators. Since the initial appliaation of flesh evaporation to marine distilling plants, a vast majority of all steam-driven ships, commercial and naval, have been equipped with flash-type distilling plants for the supply of makeup feed and potable water
" s-
[2]. Flash evaporators became widely accepted for application in marine distilling plants because their low operating temperatures result in virtually scale-free performance with normal seawater. Chemical treatment of feedwater is not required in connection with low-temperature flash evaporators as they will operate three months or more without feed treatment and without cleaning. In addition to being smaller and lighter for a given capacity, the flash-type evaporator is also more simple to operate-an advantage from an automation standpoint. The flash type of evaporator produces distilled water by dischqrging heated seawater into a chamber that is maintained a t a pressure lower than the vapor pressure of the entering heated seawater, and subsequently condensing the water vapor thereby produced. Figure 4 is a simplified diagram that illustrates the basic internal constructiop-sf a two-shge flash-type distilling plant. The majority of the marine unit's are of this multistage type with integral horizontal condenser tube bundlea. The heated aeawater is introduced and released via "spray pipes" in each stage to achieve the most effective flashing of a portion of the brine into vapor. The development and application of the "spray pipe" design for flash evaporators enabled substantial reductions in the size and weight of the flash plant [3].
MARINE ENGINEERING
-
DISTILLING PLANTS
-
-
, SEA WATER FEED
3
.
HEA~ER
533
-
-
a
"t"
PORT END ELEVATION AND SECTION " B E
BLOW CONNECTION
Fig. 2
Early horizontal straight-tube evaporator DISCHARGE
7
Fig. 4 /VAPOR OUTLET CONNECTION
TOP COIL MANIFOLD
I
GAGE
CONNECTION
TRACK FOR REMOVING COlL ASSEMBLY
li]
Multistage flash evaporator process
&OW OFF CONNECTION
HALF END VIEW
Fig. 3
Early vedical-coil type evaporator
HALF SECTION "AA"
Feedwater leaving the spray pipe is directed downward in curtains of water by a spray cap located above the spray pipe. The first separation between the flashing vapors and the brine occurs at this point. The downward velocity of the feedwater direots all but the fine droplets to the lower section of the flash chambers. The small droplets that are entrained in the flashing vapor are removed in the demisters located in the upper portion of the distilling plant, after which the vapors pass to the stage condenser. It is most important that the design incorporate a positive seal aroynd the demisters to preclude the passage of any salt-laden vapors around the separators. Since the vapor pressure of the heated feedwater entering the first-stage spray pipes is higher than the firststage chamber pressure, a portion of the feedwater flashes into vapor, thereby removing heat from the feedwater, until the temperature of the feedwater reaches the saturation temperature corresponding to the chamber pressure. I n other words, the heated feedwater is evaporated by a pressure reduction instead of a temperature elevation. After an equilibrium condition is reached in the first stage, the remaining brine is introduced into a second stage that is maintained at a lower pressure than the first, and the process is repeated. The distillation process can operate from a positive pressure in the first stage to a high vacuum in the last stage, with a stage-tostage pressure differential being the key to repeated flashing. Figure 4 shows that the entering seawater feed is used as the condenser coolant, thereby reducing the quantity
of heat which must be added in the seawater heater. This feature is shown in more detail by Fig. 5, which is a flow diagram for an 8000 gpd,"two-stage, flash distilling plant. As indicated by Fig. 5, the initial vacuum in the stages is normally cre'ated by a high-pressure steam-driven air ejector vacuum system, but vacuum pumps are sometimes used where high-pressure steam is not available. The stage condenser bundles are internally baffled to direct the flow of noncondensible gases to vent connections leading to subsequent stages or to the air ejector. The air ejector normally provided is of the two-stage noncondensing type. It is suitably sized for the efficient removal of air and noncondensible vaDors from the d i e tiller condensers. Air ejectors are covered in more detail in Section 2.6 of this chapter. The feedwater heater and separate or "built-in" air ejector coydenser are of the multipass shell-and-tube type with removable flanged waterboxes. The seawater flows through the tubes. The water velocity through the heat exchanger tubes should not exceed 6 fps at rated capacity. The condensers and saltwater heater may be provided with means for ci@mdic protgction in a manner similar to that described in Chapter 14 in mnhection with other heat exchangers in seawater service. The saltwater heater is usually designed to maintain the temperature of the feedwater entering the first stage of the evaporator between 165 F and 175 F. The steam condensate temperature from the saltwater heater does not exceed 200 F under normal operating conditions. A shell-and-tube distillate cooler is installed when the
MARINE ENGINEERING
536
MARINE EN
evaporator is to provide potable water. The distillate cooler is designed to provide a distillate temperature leaving the distilling plant of 95 F or less with an initial seawater temperature of 85 F. Occasionally, this cooler is incorporated as an integral part of the last-stage condenser bundle. Shown in Fig. 6 is a feature frequently provided for distilling plants that must operate in polluted harbors. When in polluted harbors, it is desirable to use water from shore as makeup to the distilling plant. To reduce the amount of shore water that is required to operate the distilling plant, the feedwater is recirculated through the saltwater heater. Temperature levels are maintained in the distilling plant by circulating harbor water through the stage condensers and then overboard. Thus, the polluted harbor water is not put into the evaporator heater and the flash chambers; this is an important feature in some harbors that are highly polluted and contain debris and mud which can seriously impair normal plant operation by clogging and fouling high-temperature tube surfaces. Pressure gages are supplied on marine flash plant packages in the steam supply line, evaporator shell, evaporator inlet, high-pressure steam to air ejectors, lowpressure steam to orifice, and at each pump discharge. Thermometers are provided on the saltwater heater shell, steam line to the saltwater heater, feed inlet and outlet lines of the saltwater heater, distillate outlet line from the distillate cooler, and on the evaporator shell (one per stage). A positive-displacement type flowmeter is also installed in the distillate discharge line to provide an accurate recording of plant production. Horizontal close-coupled centrifugal pumps with motors are normally provided for distillate, brine, and condensate services and require special design considerations due to the low net positive suction head usually involved. The salinity equipment provided includes salinity cells installed on the saltwater heater condensate outlet, after condenser condensate outlet, first-stage distillate outlet, and final distillate outlet. The marine distilling plant incorporates a solenoidoperated, three-way dump valve that is located in the distillate discharge line. The valve discharges the distillate to the bilge if the salinity of the distillate exceeds 0.25 grains of sea salt per gallon (or at any other specified preset control point). The solenoid is energized by the salinity-indicating system and operates whenever the salinity system is energized, regardless of the position of the cell selector switch on the salinity panel. The valve also diverts the flow of distillate to the bilge upon interruption of electric current to the salinity indicator. A higli-salinity alarm is usually installed to indicate the occurrence of a salinity level in excess of a preset value. All interconnecting piping between packaged components (including valves and fittings) is usually provided by the manufacturer of the plant. Orifice plate and desuperheater nozzles for installation in the steam supply
DISTILLING PLANTS
line are normally furnished separately for installation by the shipbuilder or owner. The following items are also common to a packaged flash evaporator such as illustrated by Fig. 7: Feedwater pump and motor with pressure gage at pump discharge usually separately mounted from the distilling plant package. Optional condensate return system: 1. A drain regulator is used when the condensate is returned to the condenser by vacuum drag. 2. A condensate pump with drain regulator which is required when discharging to more than one system. Controls for unattended and automatic operation: 1. Steam pressure-reducing valve which is required when the steam supply pressure is subject to fluctuation. 2. Seawater temperature-sensing switch and timer for dumping the distillate to the bilge in case the seawater temperature leaving the heater falls below 165 F. The timer will not permit the distillate to be returned to the storage tanks until the unit is thoroughly flushed out; adequate flushing normally requires at least 15 minutes. 3. High brine level alarm to be located in the last evaporator flash chamber. 4. Desuperheater failure alarm switch. 5. Automatic temperature control valve maintaining the saltwater heater outlet temperature at 170 F. Instrument panel. Control panel. b. Brine Recirculation Systems. The "long-tube" type of flash evaporator, which is illustrated by Figs. 8 and 9, is a design normally applied to large-capacity evaporators (over 100,000 gpd) and is therefore commonly used for large-scale "land" desalination plants. The design of the long-tube type of flash distilling plant is similar to that of the marine flash evaporator except that the condenser tubes traverse uninterruptedly through several stages and over the full length of a multistage evaporator vessel. Marine flash distilling plants have individual cross-tube condenser bundles for each stage. A "brine recirculation" system, such as shown by Fig. 8, is designed to operate at temperatures up to 250 F. The brine recirculation feature permits operation at higher evaporator temperatures and, at the same time, reduces the amount of feedwater chemical treatment required as compared with that which would be required for "once-through" operation at the same temperature. In a brine recirculation system, filtered raw seawater is pumped through the tubes of the last few evaporator stages and is discharged. The raw cool seawater flowing through the tubes removes heat from the last stages; therefore, this region is termed the heat rejection section of the plant. The major portion of the seawater coolant is discharged baclc to the ocean; however, part of it is re-
Fig. 7
photograph of flash evaporator 1 -" sz
I
b,
tained as makeup water. As indicated b i Fig. 8, the makeup water is chemically treated (with sulfuric acid) for scale control, then deaerated to remove the noncondensible gases for corrosion control purposes. Chemical treatment is necessary to retard the formation of scale on the heat-transfer surfaces and to perm$ long-run operation of the plant without the necessity of shutting down for cleaning. The deaerated and chemjcally treated makeup water is mixed with the recirculated brine and is introduced into the last heat regenerative stage. The brine mixture is then pumped through the tubes of the heat regeneration stages and the brine heater by the recirculation pump. In order to maintain a material balance in the system, a portion of the recirculated brine from the last stage is discharged to the ocean. The blowdown flow rate is controlled by the brine level in the last stage. The brine density is controlled by the ratio of seawater to blowdown flow rates. As the brine flows through the condenser tubes of the heat regeneration section, it is heated progressively in each stage by the vapors which condense on the outer tube surfaces. The brine then passes through the tubes of the brine heater for final heating to its terminal temperature. Low-pressure steam is used to heat the recirculating brine flowing through the tubes of the brine
heater. On many large land-based plants, &&-pressure steam is used to drive the feed and/or brine pumps, and the turbine exhaust steam is used in the brine heater. The condensate formed from the heating steam is discharged from the heater by the condensate pump. The heated brine is returned to the shell side of the first stage for flashing. The recirculating brine flows successively from the first to the last stage, flashing in each stage. Distillate is formed on the stage condenser tubes by condensing the flashed vapors. The distillate collects in a distillate trough which is common to all the stages and flows from the first stage to the last stage. The distilbte is pumped from the last stage into the distillate storage system. c. Once-Through Systems. The operating principle of once-through systems is the same as that for recirculating systems except that once-through systems are designed to operate at tempe&ures up Do 195 F. ; In order to achieve long-run uninterrupted operation of the plant, it is necessary to retard the formation of scale deposits; this is accomplished by continuously injecting a measured amount of polyphosphate feed treatment compound into the feedwater. Anti-foam compound can also be introduced when needed because of pollutants which cause excessive foaming in the evaporator chambers.
DISTILLING PLANTS
MARINE ENGINEERING SEAWATER CHEMICAL IN TREATMENT
i
SEAWATER OUT
+
I
H.P. STEAM VENT
1
t.
H.P. STEAM
STEAM
I
t
SEAWATER IN
L.P.
CHEMICAL TREATMENT
H.P. STEAM VENT
H.P. L.P. STEAM STEAM
1 VACUUM SYSTEM
TURBINE (IF USED)
I
I
t
!
FEE0 HEATE R
1
1
STAGES RECIRCULATE
t
4
1
I
I
Rg. 8
>
BRINE BLOWDOWN
CONDENSATE
Flow diagram of a long-tube tlash evaporator, brine reci~ulationsystem
As indicated by Fig. 9, treated seawater is pumped through all of the condensing tubes, from the last stage to the first, enroute to the feedwater heater. As the feedwater flows through the condenser tubes, it is heated progressively in each stage by the vapors which condense on the tube outer surfaces. The feedwater then passes through the tubes of the feedwater heater where it is heated to its terminal temperature. The heated feedwater is discharged to the shell side of the evaporator first stage for flashing. The remaining brine flows successively from the first to the last stage, reflashing in each stage, and is pumped from the last stage back to the ocean. 1.4 Thin-Film Evapora?ors. Section 2 of Chapter 14 discusses details relative to the transfer of heat from one fluid to another through tube walls, and Fig. 11 of Chapter 14 illustrates the resistance to heat transfer encountered when transferring heat through a tube wall. By analyzing the temperature gradients involved with the transfer of heat through tube walls, it becomes apparent that if thin films can be created and maintained on the tube walls, a relatively higher overall heat-transfer coefficient can be maintained. Doubleflute tubes have been employed in several marine evaporator designs as a means of maintaining thin liquid films. The configuration of the so-called "thin-film" tube design employed is shown in Pig. 10. The arrangement of the double-flute tubes in the evaporator is such that the tube orientation is vertical, with steam condensing on the outside and a falling film of feedwater evaporating on the inside, as indicated by Fig. 11. The operating principle of the double-flute tube is such
EVAPORATOR
i
I
I
1
DISTILLATE
( 1
FEED HEATER
that the condensate which forms has surface tension forces acting to drain it from the crests into the grooves. This feature results in the major portions of each crest on the evaporating and condensing side having a very thin film of liquid to greatly reduce the resistance to the heat flow through the crest area. The liquids in the grooves are channeled off by gravity, and the heat flow through this area is somewhat less. Seawater is introduced on the inner surface of the tubes by a spray nozzle; and the falling film of brine on the evaporating side tends to collect in the grooves by surface tension. A secondary but important function is performed by the flutes in organizing and controlling the falling film to assure uniform distribution of the brine down the length of the tube. Two-effect thin-film evaporators, such as illustrated by Fig. 11, have been installed on a number of ahips. Such t h i n - h evaporators offer the advantages of good heat-transfer characteristics, easy tube removal (provided by O-ring joints), and low probability of distillate contamination due to brine carry-over (only a small quantity of brine is maintained in the evaporator). However, chemical feedwater treatment is required with these evaporators, and experience has shown that the feedwater chemistry must be diligently maintained in order to avoid a rapid accumulation of scale on the thinfilm heat-transfer surfaces. The "spray-film" evaporator is a popular form of the thin-film type of marine evaporators. The spray-film evaporator is especially adaptable for application in an "in-cycle" or "condensate-cooled'' distilling plant SYStern. Spray-film evaporators, employing condensate
t
1
t
DlSTl LLATE
BRINE
CONDENSATE
Fig. 9
Flow diagram of a long-tube tlash evaporator, once-through system
from the main propulsiori plant condenser as the coolant in the distilling plant condenser, have been installed aboard ship. With such an arrangement, the condensate-cooled distillirig plant accomplishes a secondary objective of functioning as a low-pressure feedwater heater in addition to its primary objective of producing potable and high-purity makeup feedwater from seawater. The combination results in a higher overall cycle efficiency [4]. A typical flow diagram of a condensatecooled spray-film evaporator is shown in Fig. 12. The heating bundle of a spray-film evaporator consists of a horizontal tube bundle with the heating medium on the inside of the tubes. The heating mediumcan be bleed steam from steam turbines or steam produced from waste heat from diesel engines or gas turbines. The brine level is maintained in a hotwell below the evaporator tube bundle and a brine pump provides the necessary pressure to recirculate the brine to the spray nozzle header. Recirculated brine from the spray nozzles is "rained" over the top of the heating surface where it flows from tube to tube in thin films, resulting in relatively high "thin-film" heat-transfer rates. An a u t e matic chemical feedwater treatment system is provided to retard the formation of scale on the heating surfaces. The vapor generated at the tube surface counterflows through the curtain of recirculated brine and fine liquid droplets are thereby removed from the vapor. Removal of the liquid droplets in this manner, coupled with the secondary entrainment separation in the demister column, enables the production of distillate having a very high purity. A steam-motivated air ejector or mechanical vacuum
HORIZONTAL CROSS SECTION
-
EVAPORATION SlDE
.
WALL
THIN FILM REGION
/
--
THIN FILM REGION
4
'CONDENSATE
HEAT FLOW CONDENSING SlDE
Fig. 10
Double-tlute thin-tllm heat-transfer surface
pump can be used to maintain a low shell pressure in the evaporator. The air ejector after-condenser can be a separate heat exchanger or it can be incorporated within the evaporator heating bundle, thereby utilizing the heat content of the air ejector steam without necessitating a separate after-condenser heat-exchanger assembly. Since a spray-film evzqm-ator opeates at relatively higher shell pressures arid lower feed rates than a marine once-through type flash evaporator, the vacuum equip ment handles a lower volume of entrained air and corrosive gases and at a lesser vacuum level. As a result, the application of vacuum pumps in lieu of air ejectors becomes more attractive. To ensure conformance with the regulations of the U. S. Public Health Service [5], the shell temperature
SEA WATER FEED
SECOND EFFECT HEADER \
n
54 1
DISI'ILLING PLANTS
MARINE ENGINEERING
540
/FIRST
DISTILLATE COOLER
EFFECT HEADER
ORING TUBE JOINT
BOILER FEEDWATER c3 INLET 135.000 LBIHR 910F 270 GPM 150 PSlG
LEGEND w
STRAINER I* SALINITY CELL
LP STEAM SUPPLY
SECOND EFFECT VAPOROUTLET
NONPOTABLE DIST. TO WASTE
SECOND EFFECT MESH SEPARATOR
FIRST EFFECT ESH SEPARATOR
GATE VALVE GLOBEVALVE NEEDLE VALVE CHECK RELIEFVALVE VALVE
DISTILLATE TO STORAGE WW LWHR 14 GPM
fp
GAUGEGLASS THERMOMETER
9
PRESSURE GAUGE TEMP. SWITCH
PSlA 4 5INLET PF BLEED 26.5 STEAM SECOND EFFECT VAPOR CHEST
81264 BTUILB 7330 LBIHR EXHAUST STEAM INLET DESUPERHEATER WATER 547 LBIHR 8 91°F 150 PSlG CONDENSATE TO 1ST STG. HTR.
FIRST EFFECT DOUBLE-FLUTE SECOND EFFECT DOUBLE-FLUTE TUBE NEST ORlNG TUBE JOINT DISTILLATE PUMP.
SECOND EFFECT BRINE SUMP
CH SECOND'EFFECT BRINE DRAIN
Fig. 1 1
EJECTORCONO. OUTLET 250 LB/H BRINE PUMP 160 GPM 35 P.S.I.
SECOND-EFFECT FIRST~FFECT FIRST~FFECT DISTILLATE DRAlN BRINE DRAIN STEAM CHEST DRAIN
Two-effect marine thin-tllm evaporator schematic SEAWATER INLET 41 GPM O( 850F 20850 LBIHR 40 PSlG PRESSURE
Fig. 12
must be maintained at a minimum temperature of 165 F so as to pasteurize the vapor and distillate. A temperature switch is provided to sense the shell temperature and actuate the distillate dump valve, diverting the distillate to waste should the shell temperature fall below the 165 F minimum. Where lower operating temperatures are desirable or necessary due to a low-temperature heating medium (such as engine jacket water), a separate distillate sterilizer must be incorporated. As indicated by Fig. 12, a blowdown cooler cools the hot blowdown while simultaneously preheating the incoming makeup feedwater. The brine concentration in the evaporator sump is maintained at the proper density by setting a continuous blowdown rate. A rotometer is provided to indicate the blowdown rate. A shell-and-tube type distillate cooler is used to reduce the temperature of the distilled water produced. Seawater is normally used as a coolant in the distillate cooler in order to reduce the temperature to a "potable" water temperature. Blowdown and distillate coolers have sometimes been combined into one plate-type heat exchanger utilizing a 3-fluid flow pattern, i.e., hot brine, hot distillate, and cold makeup feed. Condensate-cooled distilling units are often used in conjunction with a combined firstr and second-stage feedwater heater/gland exhaust condenser in a manner similar to that discussed in Section 3.5 of Chapter 14. The resulting packaged spray-film evaporator feedheater is illustrated in Fig. 13, which is a photograph of a unit ready for installation aboard ship. The advantages of compactness and increased operat-
ing economy of the packaged feedwater heater/condensate-cooled distilling plant are of significant importance from the viewpoints of machinery space arrangements and overall cycle efficiencies. 1.5 Basket Evaporators. Evaporators of the basket type are specifically designed to employ "cold shocking" as a means of preventing an accumulation of scale on the evaporator heating surfaces; also, the deeply corrugated basket provides a large amount of heatrtransfer surface in a given area as may be noted from the evaporator sectional view in Fig. 14. The diagram of a double-effect basket-type distilling plant illustrated by Fig. 14 shows that the cycle is commenced by supplying steam to the basket-type heating section of the first effect. The latent heat of the steam is transferred through the basket wall to the seawater and the condensate formed is returned to the boiler. A circulation baffle or "skirt" is located a fixed distance around the basket on the seawater side. As a result of the skirt, violent boiling action takes place between the skirt and the basket. The heavy foaming characteristic of the boiling seawater inside the skirt produces high velocities and a low static head, thereby reducing the pressure in the region of boiling. In addition, the high velocity over the basket surface has a scrubbing effect which tends to retard rapid scale formation. After leaving the evaporator section, the vapor passes through a cyclone separator and a "snail" where particles of brine entrained with the vapor are removed. From the snail the vapor passes through the feedwater heater and into the basket of the second effect. The
BRINE BLOWDOWN TO WASTE 27 GPM 120°F
-
.-.-
16 .- Pel(:
I
DISCHARGE PRESSURE 73800 LBMR
Condensate-cooled spray-tllm evaporator tlow diagram
j
latent heat is given up by the vapor and the condensate produced is discharged to the flash tank on the distiller condenser. The vapor produced from the seawater in the second effect passes through separators, similar to those in the first effect,and then goes to the distiller condenser where its latent heat is transferred to the seawater in the condenser tubes. After absorbing heat in the condenser, the majority of the seawater coolant is discharged overboard; however, a small quantity (equal to about three times the distillate produced) passes through the tubes near the top of the condenser where further heating takes quantity of seawater, now called feedwater, place. T h i ~ is discharged from the condenser to the feedwater heater where the feedwater is heated to within several degrees of the evaporating temperature in the first effect; finally the feedwater enters the bottom of the firstrstage shell. The excess feedwater (about twice the evaporation rate) maintains the required level of seawater in the shell. The entrained excess,feed is carried up with the vapor produced, but it js removed from the vapor in the separator and falls into a brine gutter in the bottom of the separator section. The brine collected is discharged to become feedwater for the second effect. The quantity of brine carried off from the first effect by entrainment is sufficient to feed the second effect and to maintain the correct brine density in the first effect.
t
I-
The distillate collected in the flash tanlc of the distiller condenser is pumped through the distillate cooler after which it is either discharged to a distillate storage tank or dumped to waste. Returning to the design of the basket itself, the basket is made in sections of 0.043 to 0.050 in. moilel sheet that are given a corrugated configuration in a press. The sections are then welded to form the basltet shape shown in Fig. 14. On the seawater side, the basltet surface is highly polished; this is done to minimize the tendency for scale to firmly adhere to it. In order to control the accumulation of scale on the basket surface, it is necessary to execute a cold-shocking procedure every 75 to 100 hours of continuous operation. To do this, the brine is drained from the unit and steam at about 15'psigis admitted to the basket heating section. The flat sides of the corrugations expand, and the 250 F steam dries the scale. Next, the steam line is secured and cold seawater is allowed to cascade down over the basket. As the steam insidsthe b a s k t condenses, the pressure inside quickly drops from 15 psig to-approximately 28 in. Hg vacuum, causing the flat sides of the basket to contract. The scale is consequently cracked off the basket and drops to the bottom of the shell where it is removed through the cleanout door. Repeating this process several times removes the majority of the scale from the heating surface. In addition, at infrequent intervals, cleaning with a solvent is recommended so as to
-
I
MARINE ENGINEERING
543
DISTILLING PLANTS SECOND EFFECT AIR OFFTAKE TO EJECTOR OR VACUUM WMP
1
HEATING SECT10
CIRCULATING WMP
LEGEND EQUALIZEA
STEAM
SKIRT
IST EFFECT VAPOR
CORRUGATED BASKETTYPE HEATING SECTION
ZND EFFECT VAPOR SEAWATER CONDEWSATE
Fig. 13
Spray-fllm evaporator
dissolve any accumulation of scale on the basket, heat exchangers, separators, and pipes. 1.6 Vapor-Compression Distilling Plants. Vaporcompression distilling plants are designed for service where low-pressure steam or diesel engine waste heat are not available in sufficient quantity to operate an evaporator. The major advantage of a vapor-compression evaporator is its high thermal efficiency and that it operates on a self-contained thermodynamic cycle that is dependent only upon a source of power to provide the input energy required. The power supply may be in the form of electrical energy to operate the electric boiler, motor drive for the vapor compressor, and motor-driven pumps; or the compressor can be driven by a diesel engne or gas turbine. When the compressor is driven by a diesel or gas turbine, the exhadst gases can be utilized as auxiliary boiler heat. Special arrangements can also be made to use steam heating coils for the boiler if a small amount of steam is available for this purpose. Most offshore drilling rigs are equipped with vaporcompression plants as are some gas turbine ships and submarines. The evaporator in a vapor-compression distillation plant can be any of a variety of designs. Possible types
of evaporators include the spray film, submerged tube, basket, or vertical type of tube bundle with seawater in the tubes. Figure 15, which is a diagram of a vaporcompression distilling plant that employs a spray-film evaporator, illustrates the principles involved with a vapor-compression distilling plant. Feedwater is pumped through a solenoid valve, control valve, flowmeter, heat exchanger, vent condenser tube bundle, and then into the spray pipe manifold from which it is sprayed over the tube bundle in the evaporator shell. Some of the sprayed brine striking the hot tube bundle evaporates into steam vapors which are drawn through the demisters into the vapor compressor. The brine that is not vaporized collects in the bottom of the evaporator shell and flows into the evaporator sump. The recirculation pump takes suction from the evaporator sump and returns the majority of the brine to the spray pipe manifold. Incoming feedwater is used as cooling water for the recirculation pump mechanical seal and is then combined with the recirculating brine. The recirculation pump also pumps a portion of the recirculating brine through the evaporator sump liquid level control to the blowdown side of the heat exchangers, and then to discharge. For seawater use, the blowdown
EVAPORATOR SHELL SECTIONAL VIEW OF EVAPORATOR
Fig. 14
Double-effect basket-type distilling plant
flow rate is normally set at twice the total distillate flow for best operation. The boiler section integral with the evaporator provides the small quantity of starting and makeup heat required for the operation of the distiller. If the boiler is steam heated, low-pressure steam is piped into a small U-tube bundle to provide the necessary heat. A lowpressure (15 psig) steam supply free of contaminants is required to maintain a compressor suction pressure of 0.5 psig and is usually regulated by a diaphragm-type control valve. If the boiler is electrically heated, three electric immersion heaters are generally used with two of them being manually controlled and the third one operated automatically. All three heaters are put in service during start-up to get the plant up to normal operating temperatures as quickly as possible. During normal operation, the manual heaters are used as required, along with the automatic heater, to maintain a compressor suction pressure of about 0.5 psig. The automatic heater is controlled by an evaporator shell pressure switch. A'fter start-up, some of the distillate is used as boiler makeup water. A float-operated control valve in the distillate circuit regulates the correct amount of makeup water into the boiler.
The vapors produced from the brine on the shell side of the evaporator are drawn through demisters into the vapor compressor. Vapor compressors of a centrifugal design, that operate at a relatively high speed and low noise level, are most suitable; however, many plants have been equipped with a single-stage, positive-displacement, three-lobe rotary-type compressor. The compressor normally operates with a 2 to 4 psi differential pressure between the suction and discharge; the maximum allorvable differential pressure is about 5 psi. During the compression process, the steam vapor increases in' pressure and temperature after which, in a spray film evaporator, it is discharged into the tube side of the evaporator tube bundle. The latent heat of the steam vapor is transferred through the walls of the tubes to the brine being sprayed, over the tube bundle.= This transfer of heat condenses'tffe steam vzpor inta distillate which flows out of the tubes into the bottom of the steam chest, A spray pipe assembly is the standard means whereby recirculating water is sprayed over the evaporator tube bundle. A steam chest and vent condenser are bolted to the evaporator tube bundle. The steam chest channels the flow of steam vapors from the compressor jnto the tube bundle and the flow of distillate from the tube
DISTILLATE S W A Y CONTROL VALVE ,
PRESSURE RELIEF I
bundle. Incoming feedwater flows through the tubes of the vent condenser, where it gains additional heat from the vapors condensing in the vent condenser section. The vent condenser is vented to the atmosphere to dis* charge noncondensible gases from the steam chest. The distillate pump takes suction from the steam chest and pumps the majority of the distillate through the heat exchanger, the flowmeter, and then to the discharge connection. A small portion of the distillate enters the boiler t h r o u b the boiler water level control valve as makeup water. Another small portion enters the compressor suction duct and serves as compressor desuperheating and sealing water. A spray-film vapor-compression distilling plant is furnished by the manufacturer as a package complete with all interconnecting piping, electric wiring, automatic controls, and insulation; such a unit is shown by Fie. 16. Over a period of time, scale-forming elements in the feedwater gradually accumulate on the evaporator tubes and lower the rate of heat transferred from the compressed steam to the recirculated water (scale control is discussed in Section 2.5). Normally the amount of steam compressed is constant. Therefore the compressor differential pressure rises (causing an increase in the temperature difference) to counteract the effects of the scale accumulation. 'As the compressor differential pressure rises, there will be a slight decrease in distillate production; therefore, an acid cleaning system must be provided and used when the rated distillate capacity can no longer be maintained or when the com~ressordifferential Dressure exceeds 4.5 psi. When the lpressure differentiil across the compressor reaches 5 psi, the unit is operating at its minimum rated capacity and the evaporator tubes must be cleaned to raise the distillate o u t ~ u t lower . the electrical energy input, and prevent overloading the compressor motor. When it is desirable to use a diesel engine instead of an electric motor to drive the com~ressoka conventional four-cycle, in-line, medium-speed, industrial-type engine is best suited for continuous operation. The engine directly drives the compressor and both water pumps via V-belts. Engine jacket water can be used as a heat source to the boiler thereby providing a means of returning the heat in the jacket water to the system. The boiler section can also use the engine exhaust gas as a heat source; by doing so, maximum use is made of the engine waste heat. Detailed material and design requirements for distilling plants of the vapor-compression type for Navy applications are contained in reference [6]. The plant installed on some submarines operates on the principle described in the foregoing, but the plant is completely electrically operated and incorporates a vertical-tube bundle as illustrated by Fig. 17. Referring to the plant shown by Fig. 17, the entering seawater is preheated (for arctic operation or cold feed due to deep submergence) in an assembly that consists of two cylindrical shells, each containing an electric
-
EVAPORATOR DESUPERHEATER
u
DISTILLATE OUT
FEEDWATER SOLENOID VALVE
1 2 Fig. 15
---
FEEDWATER IN
-.
BLOWDOWN OUT
Vapor-compression distilling plant with spray-fllm evaporator
Fig. 1 6
Spray-fllm vapor-compression distilling plant
545
DISTILLING PLANTS
MARINE ENGINEERING
immersion heater. The temperature control unit automatically energizes one or both of the heaters when the incoming seawater temperature is below 55 F and turns off the heaters when the temperature is above 65 F. A rectangular shell-and-double-tube heat exchanger, or one of similar features, is located in the line between the feedwater preheater and the yent condenser. The shelland-double-tube heat exchanger is vertically divided into two sections of equal size, a distillate side and a brine side. Inner anc)'outer tubesheets are bolted to each end cover plate. Larger, straight outer heat-transfer tubes are roller-expanded into the inner tubesheets and smaller straight tubes, inserted through the outer tubes, are roller-expanded into the outer tubesheets. Incoming feedwater flows in the annular space between the inner and outer tubes while the distillate and brine, discharge from the evaporator, flow through the inner tubes in their respective sides of the heat exchanger. The feedwater flow is counter to both the distillate and brine flows, providing good heat transfer between the flows. The unit serves to further heat the feedwater and simultaneously cool the distillate and brine. The feedwater leaving the heat exchanger next passes through the tubes of the vent condenser, where it gains additional heat as the hot noncondensible gases vented from the evaporator section are cooled. The hot feedwater leaving the vent condenser is discharged into the large circulation tube located in the center of the evaporating section and collects in the brine recirculation sectioq. .- The temperature of the brine in the recirculation section is raised to the boiling point by heat coming from the boiler section. The boiler section is located on the floor of the evaporator. Its source of heat is three electric immersion heaters. During operation the electric heaters keep the distillate at the boiling point, which in turn further heats the brine in the brine recirculation section located above the boiler section. Since the feedwater is preheated when it enters the evaporator, comparatively litt~e heat is required from the immersion heaters to keep the brine at the boiling point. A brine overflow tube, located in the center circulation tube, runs through the feed section to a sidewall connection. Two distillate drain tubes from the evaporator section to the boiler section also run through the feed section, astdoes a steam vent tube from the boiIer section to the evaporator section. When the brine boils in the recirculation section, it passes into the evaporator section tube bundle where more heat from the compressed vapor is applied. This action "o'f the ly-irie, and heating increases the boil* about two thirds of the brine is vaporized at a pressure of about 1 psig in the tubes. The brine not vaporized flows into the brine overflow and is piped out of the evaporator into the heat exchanger. Steam vapors from the boiling brine in the evaporator section tubes rise into the vapor section and are drawn into the suction side of the compressor through the baffle arrangement and demisters. The baffle arrangement and demisters remove *
MARINE ENGINEERING
DISTILI.ING PLANTS
entrainments from the vapor and allow relatively ('clean" vapor to flow to the compressor suction. The vapor compressor, which is the "heart" of the plant, is a centrifugal, liquid sealing ring type that is directly driven by a constant-speed a-c motor. The unit is mounted in the evaporator top cover plate and extends down into the vapor section. It operates at a suction pressure of about one psig at 215 F and a discharge pressure between 1.5 and 3 psig in the saturation temperature range of 217 to 222 F. When in operation, the compressor reqqres a continuous flow (about 0.25 gpm) of distilled water for sealing purposes. The seal water is necessary to provide proper suction and compression and to maintain the vapor discharge temperature within predetermined limits. The distillate discharge line, downstream from the heat exchanger, is tapped to provide the seal water. The compressor raises the temperature and pressure of the vapors and discharges the vapors into the evaporator section where a number of heat-transfer tubes are arranged in a pattern surrounding vertical baffle plates. The baffle plate arrangement directs the flow of vapors around the tubes and directs the noncondensable gases to a vent tube. A system of perforations in the vent tube collects the gases which are passed through an external vent condenser to the atmosphere. In the shell side of the evaporator section the vapors condense on the tube outer surfaces to form distillate which collects in the bottom of the evaporator section (on the lower tubesheet) and flows through the distillate return tubes into the boiler section. The excess distillate flows out through the outlet pipe into the distillate pump and is pumped through the heat exchanger and into the ship's storage tanks or to waste, depending on its purity. 1.7 Membrane Processes
a. Reverse Osmosis. Of the various membrane processes used in desalting, reverse osmosis has the most promise for shipboard applications. Developments in reverse osmosis have made membrane dpsalting processes widely accepted, and they are considered for seawater desalination. To understand reverse osmosis, it is necessary to review the basic phenomenon of osmosis. Osmosis depends on the existence of a membrane that is selective in the sense that certain components of a solution (ordinarily the solvent) can pass through the membrane, while one or more of the other components cannot do so. Such a selective device is called a "semipermeable membrane"; it is usually, though not always, in the physical form suggested by the word "membrane." As illustrated by Fig. 18, if a semipermeable membrane separates a solution from a pure solvent, or two solutions of different concentrations, the tendency to equalize concentrations will result in a flow of solvent from the less concentrated phase-that is, the phrtse richer in s o l v e n t t o t h e other; it is this flow of solvent that is termed "osmosis." I f an attempt is made to impede the flow by exerting pressure on the more saline solution (assuming for simplicity that the other phase is pure solvent), the
547
rate of flow will be decreased. As the pressure is increased, a point will be found at which the flow is brought to a complete stop, the tendency to flow being in equilibrium with the opposing pressure. This equilibrium pressure (actually, the equilibrium-pressure difference between the solvent and solution phases) is called the '(osmotic pressure"; the osyotic pressure is a property of the solution and cannot depend in any way on the membrane, so long as the latter has the necessary property of semipemeability. A further increase of the pressure on the solution causes reversal of the Ssmotic flow, and pure solvent passes from the solution, through the membrane, into the solvent phase; this phenomenon is the basis of the reverse-osmosis method of desalination. Most osmotic membranes have little mechanical strength and must be supported if they are to withstand large pressure differences. The development of supporting media to provide the necessary strength without seriously impeding flow is essential to the practical application of reverse osmosis. The most common semipermeable membrane used in reverse osmosis plants is cellulose acetate, the acetate ester derivative of cellulose. This is a chemical modification of cellulose in which some of the hydroxyl groups are replaced by acetate groups. Water selectively dissolves into the membrane and is transpoi-ted through the membrane by pressure-motivated diffusion. This is called "solution-diffusion." The membrane will reject trivalent ions better than divalent ions and divalent ions better than monovalent ions,. ,,Dissolved gases tend to pass through the membrane with very low rejections. Membranes have been produced and tested in such various forms and configurations as flat membranes (called plate and frame membranes), spiral or rolled membranes, and the tubular concept. The tubular membrane concept has emerged as one of the most popular due to its moderate space utilization, ideal pressure containment, light weight, ease of cleaning and servicing, and ease of membrane replacement. 'J'he application of reverse osmosis to shipboard desalting of seawater poses a somewhat different set of circumstances than those encountered on fixed-base installations. The variation of feedwater temperature from 28 to 85 F presents system design complications, because the osmotic membrane is sensitive to the water viscosity variation that occurs with temperature change. Permeation of solvent through a cellulose acetate membrane varies approximately 1.56 percent per degree F from a standard design temperature of 77 F. This change in rate of product water is partially offset by the change in average brine concentratio* As the membrane permeation rate is increased, with a constant feed rate, the salinity of the brine being circulated is also increased. The average brine concentration is defined as the initial brine concentration plus the effluent or waste brine concentration divided by two. For every 1000-ppm salinity change of the average brine in the unit, permeation through the membrane will vary 1.8 percent due to the change in osmotic pressure of the brine.
MARINE ENGINEERING
DISTILLING PLANTS
""U When fluids of different concentrations in a venoel are separated by e membrane, the dilute solution will flow thmugh the membrane into the concentrated solution.
"""U
1ST STAGE
Q MEMBRANE BANKS
9
2ND STAGE
) MEMBRANEBANK
FRESH WATER
F!!l
OSMOTIC PRESSURE The level of the dilute solution dmpr end the level of the concentrated solution rims until an "equilibrium" is reached The pressure difference between these two levels is the "osmotic pressure."
/
BRACKISH WATER
POTABLE WATER
WATER
I
MEMBRANE
SEA WATER
Fig. 2 0
REVERSE OSMOSIS
of the ounatic I f a pressure in expressure is applied to the concentrated solution, the flow is reversed from the concentrated solution to the diluted solution. This is "reverse osmosis." FRESH WATER Fig. 18
Fig. 1 9
I MEMBRANE
Osmosis, osmotic pressure, and revene osmosis
80,000-gpd barge-mounted reverse-osmosis desalination plant
Seawater desalination by revene osmosis
The first large reverse-osmosis plant to go to sea was an 80,000-gpd plant supplied to the U. S. Navy in 1969; it is depicted in Fig. 19. The plant is of the two-stage design employing a tubular cellulose acetate membrane. The membrane is "cast" inside a porous composition paper tube which is housed in a linearly grooved plastic support tube. The support tube is contained in the "tube-sheets" by means of an "O-ring" seal not unlike a shell-and-tube heat exchanger. Seawater desalination by reverse osmosis to produce potable water that meets the requirements of the U. S. Public Health Service [5] has been dome in two stages as indicated by Fig. 20. The initial feedwater containing. approximately 35,000 ppm of total dissolved solids (tds) is reduced to about 3500-ppm tds in the firstrstage permeate and thence from 3500-ppm to 350-ppm tds in the second-stage permeate. Research to develop a "single-pass" membrane and laboratory tests on cellulose diacetate, cellulose triacetate and cellulose acetate butyrate show promise that the "ideal" membrane to produce potable water (under 500-ppm tds) from seawater in one pass will be developed. Reverse osmosis will then challenge the evaporative processes for the lead in seawater desalination for potable use. Additional "polishing," such as the addition of a polishing demineralizer in the permeate stream, would be necessary to produce boiler makeup quality water (under 1-ppm tds). b. Electrodialysis. Whereas other forms of desalination schemes entail the removal of the vastly greater bulk of water from the salts, in the electrodialysis process the salts are removed from the water. As illustrctted by Fig. 21, with the electrodialysis (ED) process, an electrical field is imposed on the water by positive (anode)
SALINE WATER IN
WNC
IATED BRINE WASTE
I
-
1
FRESH PRODUCT WATER
Fig. 21
Schematic diagram of the electrodialysis process
and negative (cathode) electrodes. This field forces the positive ions to move toward the cathode and the negative ions toward the anode. An ED desalting "cell" is formed by two smooth, rigid plastic membranes with a spacer between, which guides the water flow. One membrane carries a permanent electrical charge that is positive and the alternate membrane has a negative charge. Just as magnets of like charge repel each other, the positively charged membrane W*els positi3 ions, yet ljermits negative ions to pass through it. The membranes and electrodes are so arranged to permit ions to leave an ED cell but none to return. Therefore, water in one c a m partment of an ED cell is desalted while the adjacent "brine" compartment is made more salty. Electrodialysis plants have successfully refined brackish waters of up to 2500-ppm total dissolved solids to produce potable water in shoreside installations [7] but have not been
I
MARINE ENGINEERING
550
proven to be economically feasible for converting highly brackish or seawater (feedwaters in excess of 5000-ppm tds) to potable levels. The cost of operation of ED plants is in direct proportion to the amount of salt removed. Since the salts actually pass through the membrane, the problems of fouling or "polarization" can be much more severe than encountered in reverse-osmosis systems. The use of cation-neutral membranes in ED systems lessens the danger of fouling. The operation of electrodialysis plants at temperatures
DISTILLING PLANTS
up to 180 F has been considered. At such temperatures, the electrical resistance to the flow of current through a solution is considerably decreased, resulting in lower power requirements. A lower power consumption could place ED systems in an acceptable cost range for processing the more brackish waters, and seawater, for marine applications. Other membrane processes such as ultrafltration and piezodialysis show little promise of being able to economicauy convert the more saline waters to a potable level.
Section 2 Distilling Plant Design Considerations 2.1 Heat Transfer in Distilling Plants. The fundamental theory of heat transfer in heat exchangers and the application of this theory are discussed in Chapters 2 apd 14. The basic analytical relationship employed in the design of heat-transfer equipment is
Q
=
UA(LMTD)
where quantity of heat transferred, Btu/hr heatrtrmsfer coefficient, Btu/hr= sq ftrdeg F = heatrtransfer area, sq f t A LMTD = logarithmetic mean temperature deg F =
Q
U
The main consideration in the thermal design of an evaporator is that fluids in the evaporator undergo a change in phase. In a flash evaporator, beat is transferred to the seawater in tubes by condensing a vapor on the outside of the tubes. In a submerged-tube or sprayfilm evaporator, the heat i s trwsfemed from the condensing steam inside the tubes to the boiling seawater on the outside of the tubes. AB in the design of heat exchangers (see Chapter 14), the overall heat transfer coefficient, U, is given by 1 U= (2) (I/&) 9.a rw rdt~ (l/h@) where
+ + + +
h,
=;
reciprocal
shelhide film coefficient
resistance of deposit or scale on tube outside wall r, = resistance of tube wall metal rdto = resistance of deposit or scale on tube inside wall - = reciprocal of tube-side flp coefficient
r*
=
values of heatrtransfer coefficients encountered in service are presented in references [8] and [9]. The overall heattransfer rate of a submerged-tube or thin-film evaporat~rdepends upon the vapor pressure of the liquid being evaporated, the temperature difference between the condensing saturated steam and the vapor (which fixes the steam pressure for a given vapor pressure), the disposition of the heating tubes in the shell, and the character of the liquid being evaporated as well as the cleanliness of the heatrtransfer surfaces. Heatr transferrates increase with increasing vapor and steam temperatures, with the temperature difference remaining unchanged, except that for seawater evaporation at temperatures exceeding about 200 F and for some classes of raw fresh water at the higher temperatures (particularly for high-temperature differences), scale forms so rapidly as to offset the advantage that would otherwise be gained. Under laboratory conditions with a single acid-cleaned tube re-evaporating distilled water, the temperature difference has a pronounced effect on the heattransfer rate. Rates varying from about 1300 Btu/hr-sq ftdeg F a t a 20 deg F temperature difference up to over 3000 Btu/hr-sq ftrdeg F at a 100 deg F temperature difference have been observed with the vapor at atmospheric pressure; the heatrtransfer rate falls off with further increases of the temperature difference. However, for scale-producing feeds, the scale resistance is such a large part of the total resistance to heat flow that the effect of temperature difference on continuous service rates is small. .Since t,he scale resistance is generally such a large ~ercentakeof the total resistance to heat flow for disiilling it is convenient to group all of the other together and express equation (2) in the folbwing form: 1 1 -+r Ue
us-
ho
All of the resistances listed in the foregoing are consistently based on the same area; by convention, the tube outside area L usually taken as the base. Typical
where
U,
clean-tube overall heatrtransfer coefficient; i.e., sum of film resistance on tube inside and outside surfaces and resistance of tube wall, Btu/hr-sq ftrdeg F r = fouling resistance; i.e., resistance of tube depo&itsor scale, hr-sq ftrdeg F/Btu Figure 22 is a plot of the clean-tube overall heatr transfer coefficient, U,, versus tube velocity that can be used in the design of a saltwater heater or condenser for a flash evaporator. The correction factor which must be applied to the coefficient read from Fig. 22 to compensate for inlet water temperatures of other than 70 F is given in Fig. 23; and correction factors for the tube material and gage are given in Table 1. Typical values of the fouling factor are given in Table 2. For a stage condenser in a flash evaporator, the fouling factor, r, will typically vary from 0.0005 to 0.001 hr-sq ftrdeg F/Btu. In the saltr water heater where elevated temperatures are encountered, the fouling factor ail1 normally vary from 0.001 to 0.0015 hr-sq ftrdeg F/Btu; see Table 2. A comparison of overall heat-transfer coefficients and temperature differencestypical of flash, submerged-tube, and spray-film distilling plants is shown in Table 3. Heatrtransfer data for a submerged-tube evaporator are usually expressed as a relationship between the heat flux, &/A = UAT, and the temperature difference, AT, =
55 1
because in a boiling evaporator the heatrtransfer coefficient is also a function of the driving force or temperature difference. Most available data are based on experience and therefore include allowances for fouling. Figure 24 is a curve showing typical values for submerged-tuhe evaporators. I n the case of a condenser such as encountered in a flash evaporator, the ternperdure difference between the fluids varies from location to location within the condenser in a manner similar to the temperature distribution in a heat qx'changer. There is a difference, however, in that the heat transfer for the condensing fluid takes place at a constant temperature and only the temperature of the liquid varies. For a condenser, the logarithmic mean temperature difference (illustrated by Fig. 25) can be expressed as:
where
Table 1 Heat-Transfer Rate Tube Material and Gage Correction Factor TUBE MATERIAL^ Admiralty metal Arsenical copper Aluminum Aluminum b r w Muntz metal Aluminum bronze 90-10 copper-nickel 70-30 copper-nickel Type 304 stainless steel
-TUBE WALLGAQE17
16
15
Bwg
Bwg
Bwg
0.97 0.97 0.97 0.93 0.93 0.87 0.87 0.80 0.56
0.93 0.93 0.93 0.89 0.89 0.84 0.84 0.76 0.54
0.88 0.88 0.88 0.84 0.84 0.79 0.79 0.71 0.51
2
3
4
5
6
VELOCITY, FEET PER SECOND
Fig. 22
Condenser clean-tube heat-transfer rate
Rg. 23 Hwt-transfer rate inlet temperature correction factor
(3)
TEMPERATURE OF INLET WATER. DEGREES F
7
8
MARINE ENGINEERING Table 2
DISTILLING PLANTS
Seawater Brackish water Cooling tower and artificial spray pond: treated makeup untreated City or well water (such aa Great Lakes) Great Lakes River water : mnlmum Mississippi Delawa~e,Schuylka East Rlver and New York Bay Chicago Sanitary Canal Muddy or silty Hard (over 15 grainslgal) ~ n g i n ejacket Distilled Treated boiler feedwater Bbiler blowdown
1
1
0.003 0.003 0.001 0.0005 0.001 0.002
0.002 0.003 0.001 0.0005 0.0005 0.002
0.004 0.005 0.001 0.0005 0.001 0.002
0.003 0.005 0.001 0.0005 0.001 0.002
the heating medium of 240-400 F. If the heating * Resistances are based on a temperature ofcooling medium is known to scale, these res~stances
IS
medium temperature is over 400 F and the should be modified accordingly.
Table 3 Heat-Transfer Coefficients and Temperature Differences Typical of Feedwater Heaters in Various Types of Distilling Plants
450300 300-650 500-800
5-50 15-45 10-30
T,
= vapor temperature 1, = cold liquid temperature k = hot liquid temperature
In a flash evaporator there are a number of temperature losses which must be taken into account in the design of the equipment. These losses are associated with the brine equilibrium temperature deviation, the boiling point rise, the demister pressure loss, and the condenser pressure loss. The effect of these losses on the LMTD is shown in Fig. 25. The losses are defined as follows: Brine Eauilibrium Temperature Deviation (DEV). The amount of superheat remaining in the brine as a result of incomplete flashing. Boiling Point Rise (BPR). The elevation of the saturation temperature of the liquid above that of pure water caused by the concentration of salts in the brine. Demister Pressure Loss (DEM). The equivalent ~A
20
25
30
40
35
46
55
50
TEMPERATURE DIFFERENCE, OF
I
Flash Submer ed tube spray fifm
553
Fouling Resistances Typical of Various Types of Water, r, hr-sq ft-deg F/Btv
saturation temperature loss associated with the pressure loss through the demister. Condenser Pressure Loss (COND). The weighted saturation temperature loss associated with the pressure loss through the condenser. I n marine applications, the losses associated with the brine equilibrium temperature deviation and the condenser pressure loss are usually small and c m be neglected; however, they are usually significant factors in higher capacity and economy plants. Figure 26 is a curve of boiling point rise versus temperature. Since the demister pressure loss is normally less than 0.5 deg F, it is common to allow 1.5 deg F for the combined demister l n n ~and *"---- - boiling point rise loss in the design of oncethrough marine-flash evaporato~. Because the operating brine concentration is higher in submerged-tube and spray-film evaporators, a value of 2 deg F is usually allowed for the combined loss for such units. I n order to illustrate the principles involved in the determination of the size of a condenser and saltwater heater for a distilling plant, consider the two-stage, 8000-gpd, flash evaporator illustrated by Fig. 5. The condenser temperature differences are computed as follows: STAGE1 STAGE 2
F.
Brine temperature, deg . : . . . . . . . . . . . D e m t e r and boiling ~ nrlse t loss, deg F vapor temperature, F . . . . . . . . . .. . . Condenser inlet temperature, deg F . . . . . Condenser outlet temperature, deg F . . . . Inlet temperature difference, deg F . . . . . Outlet temperaturqdifference,deg F. . . .
C
Fig. 24
Heat flux vs. temperature difference for submerged-tube evaporators
Since the temperature differences are the same in both stages, the two tube bundles will be made to the same design. The LMTD in the two condensers will be
LMTD = 33 - 8.25 = 17.85"F 33 In 8.25 The condenser tubes selected are %-in. o.D., 18-BWG (0.049-in. wall), 90-10 Cu-Ni tubes, and the tube design velocity is 5.5 fps. From Fig. 22 the heat-transfer rate for a % in. tube with a 5.5-fps tube velocity is 633 Btu/hr-sq ft-deg F. Applying an inlet temperature correction factor of 1.06 (since the condenser for the two stages will be the same, the correction factor read from Fig. 23 is based on the lower inlet temperature of 86 F) and a material correction factor of 0.90 (read from Table 1) gives a corrected clean-tube, overall heattransfer coefficient of 604 Btu/hr-sq ft-deg F. With a fouling resistance of 0.000675, the overall heat-transfer coefficient is computed from equation (3) as
U
=
1 1 604
+ 0.000675
=
FEED, T i ----c
TI,
+ + ----+ -------+ ---- ----DEV.
Tb' #
Tv Tv
f
-----,--,CONbAp
429 Btu/hr-sq ft-deg F
With a tube I.?. of 0.527 in., the number of tubes required to pass 60,000 lb/hr of feedwater having a density of 64.1 lb/cu ft at a velocity of 5.5 fps is
fl
Fig. 25
Temperature losses in distilling plants
DISTILLING PLANTS
MARINE ENGINEERING
Consequently, the required tube length is tube length =
0.5
60
100
150
200 TEMPERATURE OF
260
350
300
NOTE: THE NORMAL SEA WATER CONCENTRATION USED IN THIS CHART HAS 34.483 G SOLIDS PER 1000 G SEA WATER.
Fig. 26
Boiling p i n t rke vs. temperature
The tubes selected are %-in., 18-BWG, 90-10 Cu-Ni tubes and the design tube velocity is 6 fps. Reading a clean-tube heat-transfer coefficient of 660 Btu/hr-sq ftr deg F from Fig. 22 and applying an inlet temperature correction factor of 1.1 from Fig. 23, and s tube material correction factor of 0.9 from Table 1 gives a corrected For a feedwater flow of 60,000 lb/hr and specific heat clean-tube heat-transfer coefficient of 653 Btu/hr-sq ftr of 0.96 Btu/lb-deg F, the temperature difference of deg F. With a fouling resistance of 0.0015, the gverall 24.75 deg F corresponds to a heat transfer of 1,426,000 Btu/hr in each of the two condensers. This being the heatrtransfer coefficient becomes case, the tube area required is 1 = 330 Btu/hr-sq ftrdeg F U = 1 0.0015 653 No. of tubes required =
+
The number of 0.652 in. I.D. tubes required to pass 60,000 lb/hr of feedwater having a density of 64.1 lb/cu ft without exceeding a 6-fps velocity is
and the required tube length is tube length =
A
w
nm.
No. tubes required =
Using a 6-pass (3 U-bend) design, the required bundle length is determined to be 36.8/6 = 6.13 ft. The &e of the saltwater heater is established in a similar manner. The temperature of the condensing heating steam is 198 F and the saltwater inlet and outlet temperatures are 138 F and 170 F respectively; therefore the LMTD is
LMTD
'60 - 28 6 0
= -= In
28
42'F
(3600) (64.1) (6)
5 (0.6W2
No. tubes used = 19 The feedwater is heated from 135.5 F to 138 F in the air ejector after condenser. Since the feedwater has a specific heat of 0.96 Btu/lb, the heat transferred to-the feedwater in the saltwater heater is Q = (60,000) (0.96) (170-138)
=
and the heating surface required is
1,843,000 Btu/hr
A
-
?rDn
With a &pass design, the bundle length becomes 35.6/6 = 5.93 ft. 2.2 Heat Sources. The thermodynamic design of a distillation unit is strongly dependent upon the heat energy source to elevate the temperature of the incoming seawater feed to that required for efficient vaporization. This energy can be supplied as steam, electricity, or waste heat in the form of exhaust gases or hot water. A distillation unit cab use either, or any combination, of these energy sources. The selection of the optimum heat source is dependent upon the total plant heat balance, desired evaporator efficiency,availabIe energy, and mode of operation contemplated. Many ships are powered by steam turbines and, therefore, steam is a heat source commonly used in shipboard distillation units. Low-pressure turbine extraction steam and auxiliary exhaust steam are the normal sources of heating steam for distilling plant operation. To permit the most efficient and flexible operation, it is standard practice to design the distiller to operate with either steam that has been bled from the main turbine or steam that has been exhausted from any of the various auxiliary turbines. When operating at sea, bleed steam permits the most economical operation as the majority of the heat remaining in this steam is otherwise rejected to the main condenser, contributing very little additional energy to the cycle before being condensed (see Section 3.4 of Chapter 18 for further discussion regarding bleed steam). Ror in-port operation auxiliary exhaust steam is supplied to the distiller. The use of this reliitively low-cost steam permits continued economical operation of the distilling plant. The pressure of bleed steam is dependent on the turbine design and the specific extraction point. Distilling plant designs are normally predicated on steam being supplied at a pressure of 9 to 11 psia. Auxiliary exhaust steam is usually furnished at a pressure of 15 psig and reduced to 5 psig by a pressure-regulating valve; a pressure of 5 psig is usually provided at the inlet to the critical-pressure orifice. High-pressure air ejector steam, where employed, is also used to preheat the seawater feed. Heat in the air ejector steam is reclaimed in the air ejector aftercondenser using distilling plant feedwater as the coolant. The air ejector motive steam pressure is dependent on the air ejector design and can range anywhere from 75 to 150 psig; pressures in the 135- to 150-psig range are most common. Diesel or gas turbine driven ships often employ waste
555
heat as a heat source for~distillingplant operation. The waste heat may be in the form of diesel engine jacket cooling water or it may be in the form of heat recovered from the turbine or engihe exhaust gases. Submergedtube and thin-film distilling units have been used when hot water is the heat source because these units operate with a lower shell temperajure and can use water at a lower temperature as a heat source than can distilling plants of the flash type. Distilling plants for this application are~commonlydesigned to use jacket water with an inlet temperature to the distiller of 165 to 170 F. With a flash evaporator, the hot-water inlet temperature to the heater must be at least 190 F to obtain an adequate temperature differential to pennit an efficient saltwater heater design. A supply of diesel jacket water above 180 F is not normally available; however, water at such temperatures is obtainable on ships driven by gas turbines by installing heat exchangers to recover heat from the exhaust gases. See Chapter 2 for further discussion on this subject. Electric power is often used to operate the efficient vapor-compression type of distilling plants; however, due to considerations of economy, electric power is not normally used with distilling plants of other designs except as a supplemental or emergency heat supply. For flash, submerged-tube, or basket designs, the direct use of electrical power for feedwater heating is not practicable. For some applications, it is advantageous to use a combination of heat sources to operate the distillation unit. Figure 27 illustrates a typical flow diagram for a unit using all three common heat sources; this design has been applied in a number of diesel-powered naval ships. For the unit shown in Fig. 27, the primary heat source is diesel propulsion engine waste heat in the form of engine jacket water. This primary heat source is used when the ship is at sea operating with sufficient engjne power to provide the necessary waste heat. When the ship is operating at reduced power, the engine waste heat is augmented by heat furnished from either the steam heater or electric heater. When operating in this condition, the jacket water is directed through these heaters, thereby increasing the feedwater temperature. The steam heater is a shell-and-tube heat exchanger and the electric heater uses immersion heating elements: When the ship is in pee, all heat required to make rated capacity is supplied by the steam heater and electric heater. The steam heater is often sized to provide sufficient heat to produce rated capacity and the electric heater is sizedefi produce"'% rated cipacity. Aside from the increased reliability and operational flexibility provided by this combined heat source arrangement, an additional advantage is that the steam heater and electric heater are used to warm the diesel engine prior to starbup. 2.3 Distilling Plant Economy. The generally accepted meaning of the term "distilling plant efficiency" is the pounds of distilled water produced per 1000 Btu of
MARINE ENGINEERING
tions for ships constructed under the puwiew of the U. S. Maritime Administration are similar [ l l ] and 90-10 copper-nickel is also required for the flash chambers, with the provision that steel evaporator chambers, suitably protected against corrosion by means of a protective coating, will be considered. Protective coating materials provide definite savings in initial cost but the protection they provide steel surfaces in a marine distilling plant is sometimes iinsatisfactory. It is significant that pradt,ide in the use of nonferrous materials for distilling plants in ships built abroad followed that of the United States. Prior to 1950, distilling plants built in England and Germany used cast iron as the material for the shells. If the shells were thick enough, they were generally satisfactory, but the units were much heavier than when more corrosion-resistant materials such as copper-hickel, monel, and titanium were used. Not only is it important to use nonferrous inaterial in the areas in contact with the brine, it is equally important that the condenser baffles and distillate troughs and vent sections be of nonferrous materials. These come in contact with the vapolg which are laden with oxygen and carbon dioxide released from the heated brine and also are subject to corrosion. Due to rising costs and long lead times experienced in obtaining copper-nickel tubing, the use of titanium tubes in marine evaporators may become more common. See Section 5.6 of Chapter 22 for additional details relative to the materials of construction for marine distilling'plants and heat exchangers.
R P
CONDENSATEOUT D dirtlllats V vapor Jw FDW feed-
w-
Fig. 27
Flow diagram of an evaporator having three heat sources
heat supplied. For multistage plants, fitted with such features as evaporators which are heated by vapor produced in an earlier stage, the overall economy of the plant may be estimated best by first preparing an approximate heat balance flow sheet. Such flow sheets are approximate, but the balance a plant will assume in sewice can be predicted with fairly good accuracy. There are several factors which complicate a precise prediction of the manner in which a plant will perform in sewice. A "clean" plant is capable of producing excessive quantities of vapor to such an extent that carry-over will result, thereby contaminating the distillate; jn such an event the steam or feed supply must be throttled or other means used to control the output. Tube surfaces of different units foul to a different degree; therefore, fouling cannot be predicted with exactness. Additionally, the circulating water temperature varies with the season and locality. All of these factors affect the balance and economy of a distilling plant. The overall economy of a distilling plant, as distinguished from the so-called efficiency, may be expressed in terms of the pounds of distillate produced per pound of additional fuel required for the distilling plant, over and above the fuel otherwise required to fulfill the ship's power requirements. Establishment of the distilling plant overall economy entails an assessment of the amount of fuel used to produce the steam required for the saltwater heater, the electrical power required for the various pumps in the, plant, the useful work that could have been performed by the heating steam had it been used for other purposes, and the heat returned to the boiler feed system from the distilling plant (a very important consideration for distilling plants which have
557
DISTILLING PLANTS
condensers that use the main propulsion plant condensate as the coolant). Since the overall economy of a distilling plant depends upon the ship's power plant, and the interrelationships between the two, it is not possible to accurately state the economy of a distilling plant independently of the power plant. However, the following gives an approximate idea of the overall economy characteristic of four types of distilling plants: Single-stage submerged-tube plants operating on boiler steam at reduced pressure, 13 pounds of distillate per pound of additional fuel. Two-stage flash plants operatipg on bleed steam, 50 pounds of distillate per pound of additional fuel. Three-stage flash plants operating on bleed steam, 75 pounds of distillate per pound of additional fuel. Spray-film vapor-compression plants with electric motor drive and electric heaters, 200 pounds of distillate per pound of additional fuel. Based op flash distilling plants operated with exhaust or bleed steam, the weight of additional fuel required, together with the operating weight of the distilling plant, is bpt a small fraction of the weight of the distilled water produced during an average ship voyage. For a given ship, investigations may show t h a t a worthwhile savings in tonnage can be effected by using distilled water rather than tank water; and that, all things considered, the cost of distilled water is lower. 2.4 Materials of Construction. The U. S. Navy has standardized on 90-10 copper-nickel as the material used for tubes, tubesheets, waterboxes, evaporator shells, and piping for marine distilling plants [lo]. The specifica-
2.5
Scale Control and Acid Cleaning
a. Scale Control. High-temperature, land-based evaporators-have a much inore serious scale control problem than do marine plants which operate at a low temperature and normally under vacuum. Nevertheless, the formation of scale is a major consideration in the design of marine distilling plants. One source of scale is the bicarbonate in seawater which decomposes when the seawater is heated and then reacts with the magnesium and calcium in the seawater to produce magnesium hydroxide and calcium carbonate scales as shown in the following reactions: 3
1
2 bicarbonate ions
=
water
+ carbon dioxide gas + carbonate ion
i
carbonate 'ion
+ water = carbon dioxide gas + 2 hydroxide ions
+ calcium ion = calcium carbonate scale 20H- + Mg++ + Mg(0H)Z 2 hydroxide ions + magnesium ion = carbonate ion
magnesium hydroxide scale
When an acid or acidic salt is added to a seawater supply, it will neutralize what is commonly termed the alkalinity of the seawater. The alkalinity is the sum of the hydroxide, carbonate, and bicarbonate ions in the seawater, although usually only bicarbonate ions are present in significant ,quantity. The hydrogen ions produced in the seawater by tfie added acids or acidic salts act to destroy or neutralize the bicarbonates by the following reaction : ~
~
bicarbonate ions
0
+ H+ 3
-
+
COz
+ hydrogen ion
t
+ Hz0
=
carbon dioxide gas
+ water
Therefore, by treating the seawater with acid and converting the bicarbonates to form carbon dioxide and water, it is possible to eliminate the source of carbonate ions and prevent the formation of magnesium hydroxide and calcium carbonate scales. Such treatment is common practice in the operation of land-based multistage desalters. During the acidification of normal seawater, for scale control purposes, approximately 100 ppm of carbon dioxide is formed through bicarbonate alkalinity breakdown. This."carbonated brine" is acidic and corrosive to fermus materials of construction. The carbon dioxide can be removed by scrubbing the seawater with air or steam. Calcium sulfate scales are ~grobablythe worst of the common scales which may bk encountered when seawater is heated and evaporated to form concentrated brine. Calcium sulfate scales are not readilv soluble in acid solutions; consequently, they must often be removed by mechanical methods. The mechanism of calcium sulfate scale forination is simply precipitation caused by the concentration of calcium and sulfate ions beyond the solubility of calcium sulfate. Two crystal forms of calcium sulfate are involved, the anhydrite and hemihydrate. The anhydrite is the most insoluble form; therefore, its solubility is first exceeded uDon concentration of a feedwater containing calcium and sulfate ions. Solutions supersaturated with respect to the anhydrite are stable for long periods of time. Supersaturation with respect to the hemihydrate is not as stable as that for the anhydrite, so that scaling does not usually take place until the concentration of calcium sulfate hemihydrate is exceeded. The solubility of both the anhydrite and hemihydrate decreases with increasing temperature. The most common w & w f preverrting calcium sulfate scaling in marine seawater evaporators is toMm&ntaina blowdown rate sufficiently high that the brine concentration will not be saturated or supersaturated with respect to calcium sulfate hemihydrate. With proper distilling plant operation, calcium solubility limits are avoided and no calcium sulfate scales will be formed. Both thin-film and submerged-tube low-temperature marine distilling plants incorporate a chemical feed treatr ment system. This feed treatment system basically
MARINE ENGINEERING
DISTILLING PLANTS
I
consists of a chemical mixing tank and a proportioning pump. A small amount of plyphosphate is automatically added to the feed stream so as to chemically react with scale-forming ions to produce a soft sludgetype scale that is more readily washed from the tube surfaces than the calcium carbonate and magnesium hydroxide scales which would otherwise be formed. b. Acid Cleaning, During the operation of a marine seawater distilling plant, scale will form on the heating supfaces of the evaporator as reviewed in the foregoing. This scale is usually 80-90 percent calcium carbonate. The remainder is a mixture of calcium sulfate, magnesium hydroxide, metal oxides, silica, and miscellaneous deposits. However, when seawater feed is improperly treated with polyphosphates, other deposits such as phosphate sludges may be formed in large quantities. If fresh feedwater is used, silicates or calcium sulfate may be present. The rate of buildup and the composition of the deposits depend on such factors as the operating temperatures, the brine density, and flow rates. Feed treatment and cold shocking reduce the buildup of scale, but do not prevent or remove deposits entirely. The symptoms of scale formation are: (a) Consistently rising temperatures in the evaporator stages (decrease in vacuum). (b) Heating steam pressure to the feedwater heater, or heating tube bundle, is required to be above the design value t o produce the specified temperature at the feedwater heater outlet.
Sealed heat-transfer tubes can be cleaned chemically by circulating a diluted acid solution through, or over the tubes. Hydrochloric (muriatic) or sulfamic acid are the chemical reagents commonly recommended. Hydrochloric acid should be used only if sulfamic acid is not available, and then only by qualified personnel experienced in its use. Chemicals used for acid cleaning are samewhat haaardous; they are hazardous in the sense that most acids are dangemus. I t is essential that the operator understand the potential danger involved with the use of acids. If proper precautions are taken, persome1 injury and equipment damage can be avoided and acid cleaning of the evaporator can be accomplished in a minimum of down time. Sulfamic acid hasbecome the chemical mast commonly carried on board ship. No extra precautions are necessary in storing or handling dry sulfamic acid; however, since the acid dust will irritate the nose, eyes, and skin, careless handling which may result in its dispersion should be avoided. Sulfamic acid is only mildly corrosive to metals and there is no fire hazard involved in its use. lulfamic acid is considerably less objectionable in all respects than dilute hydrochloric acid. The maximum acid concentration should not be numerically lesa than pH 2.0. A reference for the correlation of pH values with various acidic and basic solutions is given by Table 4. When one pound of sulfamic acid powder with color
indicator is dissolved in one gallon of water, the solution will turn a light red color, indicating sufficient acid concentration to dissolve scale. If the solution is heated to approximately 120-140 F, the color will change to a deep red. As the solution is circulated through the evaporator components and comes in contact with scale deposits, a chemical reaction takes place between the scale and acid which dissolves the scale and reduces the concentration of the acid. The by-product of the chemical reaction is the liberation of large quantities of carbon dioxide gas which must be vented to atmosphere at some point in the cleaning circuit. As the acid is circulated and the chemical reaction takes place, scale is dissolved, consuming acid in the process, and the solution will gradually change in color to orange or yellow, indicating that most of the initial charge has been dissipated. At this time a recharge of sulfamic acid will be required to increase the acid strength and change the color back t o red. Beriodic recharging or makeup of fresh solution will be required until the acid solution remains red for to of an hour after the last acid charge. When this color condition occurs, the operator can be assured that all sohble scale in the cleaning circuit has been dissolved since the acid strength is no longer being dissipated by contact with scale. The acid should then be completely drained from the plant and the entire cleaning circuit flushed with large quantities of fresh seawater. 2.6 Distilling Plant Vacuum Equipmenf. Proper venting of the distilling plant condenser is most important to prevent the buildup of noncondensible gases. Air is liberated in the evaporator from the entering feedwater which is saturated with air. Carbon dioxide may also be released through a breakdown of bicarbonates in the seawater. Since a low-pressure marine evaporator o p erates under a vacuum, small quantities of air also leak into the unit through the gasketed joints. If the condenser is not properly designed, these gases collect in low-pressure pockets and render these areas ineffective, thereby reducing the performance of the unit. Figures 28 and 29 @howthe vent baffling and shrouding in a typical condenser and saltwater heater. Three different types of systems have been used to establish and maintain the low pressure required in the evaporation chamber; these are: steam-motivated air ejectors, mechanical vacuum pumps, and water-motivated air eductors. Of the three, steam-motivated air ejectors have been most common; however, mechanical vacuum pumps and water eductors are suitable for limited applications. The service conditions of a vacuum system vary depending on the design of the distilling plant. On a submerged-tube or spray-film distiIler, the shell temperature and vacuum are maintained at a predetermined value regardless of the seawater feed temperature, and the vacuum systems for these distillers operate at one suction condition. I n a flash evaporator, the shell temperature varies widely as the seawater feed temperature
Table 4
pH Value INCREASING ACIDITY
0 1.0 2.0
T@;l Purple Red
2.5
3.0 4.0 4.5
Pink Orange
5.0 6.0
Gold Yellow
NEUTRAL
7.0
White
INCREASING ALKALINITY
8.0 9.0 10.0
Correlation of pH with Various Acidic and Basic %lutions
Strepgth Very strong
pH of Industnd Chemicals Sulfuric acid 4.9% (1.ON) Hydrochloric acid 0.37% (LON)
13.0 14.0
Dark Brown Black
lO,oOo,aa, 1,000,000
Fairly strong Too weak to dissolve sde Mild Neutral
100,000
/
Orange juice Beer
A&;
acid 0.87 (0.1N)
10,000 1,000
n
100
American cheese M i
ib i
- .
Distilled water at 77 - .F Egg white
10
Borax Fairly strong
1w
Milk of Magnesia '
Ammonia 1.7% (1.0N)
12.0
Lodalue arithmic to Pure Water
Lemon juice
Will dissdve average scale
Mild Grass Green Dark Green
pH of Household Items
Caustic soda 0.4% (1.ON) Very strong
1,000 10,000
100,000 1,000,000 10,000,000
Caustic soda 4.0% {l.ON) NOTE: The color assumed by the indicator (e.g. litmus or pH paper) varies depending on the type and r a w of indicator used.
changes throughout the range of 28 F to 85 F. The vacuum system must, therefore, be designed to provide proper operation at the suction conditions encountered with inlet seawater feed temperatures in this range. In a two-stage flash evaporator, the temperature in the second-stage evaporation chamber will be approximately 114 F at a seawater temperature of 85 F; however, a t a seawater temperature of 28 F,?the second-stage temperature will drop to 70 F. The specific volume of steam is 238 cu ft/lb a t 114 F, but at 70 P the specific volume is 869 cu ft/lb. Consequently, the 28 F seawater feedwater condition controls the design of a flash evaporator vacuum system. The vacuum system must be capable of extracting large volumes per pound of steam and noncondensibles removed a t the lower seawater feed temperature. The highly corrosive nature of the noncondensibles released in a distilling plant vaporization chamber must also be considered in the design of a vacuum system. In addition to air inleakage and seawater deaeration loads, the distilling plant vacuum system must also remove the COz which results from the breakdown of carbonates in the seawater when it is heated to 170 F as well as NH;I and HpS, which are also introduced by polluted seawaters. a. Steam-Motivated Air Ejectors. Air ejector systems used with most marine evaporators are of the twostage noncondensing type, but singlestage systems can
be used on submerged-tube and spray-Mm units where higher shell pressures are maintained. The elementary operating principle of air ejectors may be understood by reference to Fig. 30. High-pressure motive steam is led to the unit and passes through the nozzle where its pressure is dissipated in accelerating the steam to a high velocity as it passes from the nozzle throat through t$e expanding section of the norele. The high-velocity jet of steam issuing from the nozzle entrains the saturated air mixture entering the ejector element. f i c t i o n between the steam jet and the low-pressure air causes the latter to move with the steam into the converging section of the diffuser tube where the steam and air mix. The divergent section at the downstream end of the difluser tube serves to decrease the velocity of the moving gas and increase its pressure, thus converting the kinetic energy to pressure energy. I n this way the air and noncondensible gases removed from the vacuum system are compressed, which is the object of,the air ejector element. Por a d a o n a l dis<&ion regarding air ejectors, see Chapter 12. The elementary air ejector element shown in Fig. 30 comprises a single-stage ejector which is the type used for low-pressure type distilling plants. In the case of multistage £lash plants, two-stage ejectom are generally e m ployed in order that a high vacuum may be economically obtained. In this case a second air ejector element is placed in aeries with the primary air ejector element.
DISTILLINGPLANTS
MARINE ENGINEERING
SHELL VENT OUTLET
Fig. 28
Typical candenser vent baffle
Fig. 31
VENT OUTLET
'\
[
STEAM INLET STEAM
MAIN STEAM INLET
b
DISCHARGE TO INTER OR AFTER CONDENSER Fig. 29
Liquid-ring vacuum pump
Typical saltwater heater and air elector baffler
Fig. 30
Elementary air elector element
The first-stage air ejector compresses the steam and air complication involved and the fact that two-stage ejectors to an intermediate pressure, and the second-stage element are capable of evacuating to a sufficient vacuum for all provides the h a 1 compression to atmospheric pressure normal installations. The air ejector motivating steam is coridensed in an or slightly above to dispose finally of the air and nonafter-condenser which uses the entering feedwater as the condensible gases. A single-stage ejector unit is not cooling medium; such an arrangement is typically suitable for operation at a vacuum higher than about 27 inches of mercury, and thus all high-vacuum ejector illustrated by Fig. 5. This reclaiming of the heat in the installations are provided with two- or three-stage ejector air ejector motivating steam increases the overall distiller elements in series. Three-stage ejectors are not nor- efficiency. The air ejector after-condenser is often conmally employed in naval practice because of the extra structed as an integral part of the evaporator saltwater
heater or part of the heating bundle in the submergedtube or spray-film type of distilling plants. This arrangement simplifies the construction of the units and reduces maintenance requirements by avoiding the necessity of an additional heat exchanger. Since noncondensibles in the after-condenser section are highly corrosive, the material in the after-condenser portion of a tube bundle should be nonferrous. The material commonly selected for this service has been 90-10 or 70-30 Cu-Ni. The air ejector steam chest is usually made of steel; the nozzle is stainless steel, and a stainless steel strainer is provided. The air chamber and the diffuser must be of bronze or copper-nickel construction to resist the corrosive vapors exhausted by the system. b. Mechanical Vacuum Pump. Mechanical vacuum pumps have been used on submerged-tube and sprayfilm distilling units, which have less severe vacuum system requirements as compared with flash units. Submerged-tube and spray-film units normally operate with shell temperatures in excess of 125 F, and a single-stage vacuum pump can be employed. I n most cases, a liquid-ring pump using fresh water to form the liquid ring has been used to minimize the corrosion of pump parts. A typical liquid-ring vacuum pump is illustrated by Fig. 31, and is discussed further in Section 4.8 of Chapter 12. The fresh water is cooled by seawater in a heat exchanger which is built into the vacuum pump seal water tank. If a vacuum pump were used with a flash distilling plant, a two-stage pump would normally be required in order to provide satisfactory operation with low seawater
temperatures. The greater complexity of the two-stage system coupled with large,component sizes and high costs involved in maintaining the pump under severe corrosive conditions has limited the use of mechanical vacuum pumps to instances where motive steam is not available to operate an air ejector. c. Water-Motivated Air Eductors. A water-motivated air eductor has also been used for the submergedtube and spray-film designs which have shell temperatures above 125 F. In these distilling plant designs, more water is required for cooling than is actually used as feed to the unit. This water is normally directed overboard after passing through the condenser bundles; however, the excess cooling water can be used as motivating water for an air eductor, permitting operation of the eductor without havihg to provide an additional supply of motivating water. I n cases where an eductor can fulfill the vacuum requirements, its use permits a compact installation since some steam piping is deleted and an after-condenser is not required. Very few water-motivated eductofs have been used with flash-type evaporat'dT since it 'is difficult for an eductor to satisfy the stringent vacuum requirements of flash plants. 2.7 Distilling Plant Automation. The Maritime Administration Standard Specification for Cargo Ship Construction [ l l ] requires that each desalination unit be capable of unattended, automatic operation after being put on the line locally. Most marine distilling plants can be readily automated to such an extent with a few simple
DISTILLING PLANTS
. ,
I
controls. Since the steam supply is normally from a lowpressure extraction paint, or alternatively, fram a higher pressure source reduced to 5 psig and then controlled through a fked critical-pressure orisce in the steam supply line to the distilling plant, no autumatic control is normally required for the steam supply. It is necessary to maintain the brine temperature leaving the saltwater heater of a flash plant at a specified temperature, usually 170 F; this is done by means of an automatic temperature control valve. No control of the brine flow leaving the brine overboard pump is required since a11 of the water entering the unit is pumped averboard. The only control required on the distillate is that of conductivity; a 3-way valve automatically diverts the distillate to the bilge in the event that the conductivity is not within the specified control limits; the dump valve can be provided with an automatic reset. Provisions must also -be made to stop the feedwater flow in the event of a brine pump failure to prevent accidental flooding of the evaporator. To insure that the water pmduced will meet the requirements of the U. S. Public Health Service [5], an arrangement or control in the circuit must be incorporated to inaure that all of the distillate is heated to at least 165 F; otherwise a distillate sterilizer must be incorporated. In a submerged-tube evaporator or thinfilm double-effect unit, it is necessary to add a sterilizer to the unit since in these types the normal operating temperature throughout the system is below 165 F. The steam supply to this sterilizer must incorporate an automatic regulating valve. Figure 5 shows those controls necessary for a flash plant to meet the unattended automatic operation requirement after being put on the line locally. Fully automatic operation requires that the plant be equipped with automatic control devices to initiate all starting sequences upon activation of one pushbutton. The complexity of such a system is so much greater than a system in which each of the various starting and stopping sequences is initiated by an on/off pushbutton or switch that distilling plant automation is normally restricted to systems with manual starbup provisions. The practicability of completely automating the start-up of a distilling plant is questionable due to the variables encountered in vacuum and temperature conditions coupled with the necessary timing of the sequences dependent upon these variables. The following sequence, requiring various controls and additional features, is necessary to accomplish automatic starting of a typical two-stage flash plant by a single on/off pushbutton or switch: 1. The feed pump is started, the brine pump is started, an air-operated bypass valve opens in the inlet feed line to the first stage. Correct operation of the feed pump is proven by a 20-second time-delay pressure awitch; if the pump pressure does not build up, the plant is secured. A high-level switch proves the operation of the brine pump; if the pump does not remove the brine from the second stage, the high-level switch will
secure the plant. A float switch with a one-minute delay proves the opening of the bypass valve; if a level does not build up in the first stage, the plant will be secured. 2. The steam valve to the air ejector opens after a three-minute delay; the delay is obtained by the use of a timer. If the air ejector malfunctions and the vacuum is not reduced to the operating level in 35 minutes, a vacuum pressure switch and timer arrangement will trip the plant. 3. When the vacuum pressure switch indicates that the second stage has reached the operating vacuum conditions, the steam pressure-reducing valve opens to allow 20-percent steam flow. Also activated at this time is a timer which continues to open the steam valve, allowing 35 minutes for full steam flow to be reached. Opening of the steam valve is proven by a temperature switch and timer in the inlet to the first stage; if the feedwater temperature does not reach 120 F in ten minutes, the plant will be tripped. A solenoid valve in the "start-up" desuperheating water supply line is also opened by the vacuum pressure switch in the second stage. 4. The condensate pump is started by an electric float switch in the feedwater heater hotwell. Starting this pump secures the start-up desuperheater water supply solenoid. Proof of the operation of this pump is determined by a pressure switch in the feedwater heater. If the pump does not operate correctly, the level in the heater will rise and the pressure in the heater will increase. The pressure switch will trip the plant if the heater pressure reaches 3 psig. 5. The distillate pump is started by a float switch located in the condenser trough above the normal operating level. Proof of the operation of this pump is obtained by the use of a low-vacuum trip in the second stage. If the condensate level in the condenser becomes excessively high, the air ejector suction line will be covered, the vacuum will decrease, and the plant will trip, 6. Additional items also required include: Pump back-pressure control vaIves are required on the brine, distillate, and condensate lines. Steam traps are required in all steam Iines; these traps should discharge to a drain main in order to prevent water hammer, A panel is required at the local automatic start station to indicate the progress of the start-up procedure. If the plant trips during startrup, this is necessary to determine the component, switch, or relay which caused the trip. Simply automating the opening of the steam supply valve and air ejector steam supply valve can resdt in possible compromises to reliable and extended operation. These services should be put on the Iine slowly with the caution required by sound engineering practice when admitting steam to a plant. Therefore, additional control and timing devices to "bleed" warm-up steam prior to opening the valves are required. These two valves, once opened during an attentive starting sequence, need
not be touched or regulated to insure proper plant performance. The automatic timing of the stdrbup sequences is necessarily based on starting with a cold plant plus safety margins and, therefore, the timing of successive, sequences is set for a much longer time than normally required to manually start the plant under normal conditions. Consequently, it would be common for the operator to override the automatic start in many instances to decrease the time required to get the plant on the line. The advantages offered by a completely automatic start-up system may not be justifiable when considering the initial expense, complexity, and maintenance of the automation system itself. The maintenance of the complex automation system will usually require more man-houm than those required to manually open the necessary valves in proper sequence for start-up. Most distilling plants have been designed to operate automatically after a manual start-up. Reterences
Art Conference," Sacramento, California, June 18, 1968. 3 Gordon F. Leitner, "Marine Distilling Plants Current Practice for Specifications," presented to the Northern California Section of SNAME, January 13, 1966. 4 Charles D. Rose and Lee J. Hartenstein, "Condensate Cooled Evaps," Marine Engineering/Log, May 1968. J 5 "Handbook on Sanitation of Vessel Construction," U. S. Public pealth Service, Department of Health, Education arjd Welfm. 6 "Distillation Unit, Water, Thermocompression," Military Specification MIL-b-16196. 7 Richard M. Ahlgren, "Electrodialysis as a Chemical Unit ~ ~ e ~ a t i opresented n)' to Americaq Institute of Chemical Engineera, St. Paul, Minnesota, February 20, 1962. 8 D. Q. Kern, Process Heat Transfer, McGraw-Hill Book Co., 1950. 9 W. H. McAdams. Heat Tramission. McGrawBill Co., 1954.Unit@,Water; Steam, or Flashed 10Book "Distillation
1 H. U. Sverdrup, M. W. Johnson, and R. H. Flem- Vapor Operated, or Fresh Water Heated, Low Pressure, ing, The Oceans-Their Physics, Chemistry, and General Naval Shipboard," Military Specification MILD-18641. Bwbgy, Prentice-Hall, Inc., 1942. 11 "Standard Specification for Cargo Ship Construc2 F. A. Loebel, "The Flash Desalting Process," tion," Office of Ship Construction, Maritime Adminispresented at "A Seawater Conference--State of the tration, U. S. ~ e ~ k t m eofn tcommerce.
1
,!
\
HULL 'MACHINERY
CHAPTER X V I Irving W. Smith Archer M. Nickerson, Jr.
I Hull Machinery
Section 1 General Design Considerations .1.1 Scope. Broadly defined, "hull machinery" (or "deck machinery") includes all power-driven equipment located outside the machinery spaces that is not associated with the main propulsion plant. Included as hull machinery are such items as steering gears, anchor windlasses, all types of winches (including cargo, topping, vang, mooring, accommodation ladder, and boat winches), cranes, capstans, elevators, dumbwaiters, conveyors, escalators, bow thrusters, and active fin stabilizers. Very few of these equipments require any design techniques beyond those of standard machine design practices, which are adjusted to make the equipments suitable for the marine environment and to cope with the oftentimes unique situations encountered on shipboard; therefore, this chapter concentrates on the design considerations peculiar to marine equipment. Close cooperation is required between the marine engineer and the naval architect in the design of hull machinery. As a consequence,this chapter has been written to complement the Society's publications Principles of Naval Architecture and Ship Design and Construction, which emphasize the naval architect's role in the design of hull machinery. 1.2 Types of Drive. There are, for most practical purposes, only three types of drives for hull machinery: steam, electric, and electrohydraulic. However, one occasionally encounters a diesel engine drive, as for a bow thruster, or a manual drive, as for schooner guy, vang, or accommodation ladder winches. Frequently a manually operated item may also be arranged to be driven by a portable power unit such as an air or electric drill. Because of the emphasis on crew reduction and quicker turnaround for ships, manually driven hull machinery has largely disappeared from the scene. Steam-driven units are seldom installed except on some tankers or in the vicinity of cargoes that are of a hazardous nature. Steam-driven units offer the advantage of minimum fire hazards because they do not spark as does electrical equipment. Although steam-driven equipment is generally very rugged, simple, and easy to maintain, the difficulties encountered in keeping steam mains properly insulated and the problems associated with maintaining an adequate steam flow in cold weather tend to outweigh the advantages of ruggedness and simplicity (see Subsection 2.2 d). Then, too, the reliability of electric motors and controls, the desire for remote con-
564
trol of many hull machinery components, and the increased technical ability of ships' engineers have all pointed the development of hull machinery drives away from steam and toward electric or electrohydraulic powering. With the advent of electrically driven engine room auxiliaries, direct current was the almost universal choice because it was easier to obtain multiple speeds for pumps, fans, and particularly for such items of hull machinery as cargo winches, capstans, and anchor windlasses. Variable or multiple speeds were obtained by using either rheostats or power resistors. The power not used to drive the equipment was converted into heat in either the rheostat or resistors, and it was necessary to provide extra ventilation to dissipate this heat. Shortly after World War 11,economic studies indicated that very appreciable savings could be realized by changing from two-wire, 240-volt, direct-current ships' systems to three-phase, 450-volt, alternating-current systems. The methods of obtaining variable or multiple speeds for the several categories of hull machinery will be dealt with as the equipments are described. For the moment it is sufficient to say that these speeds may be obtained in any of several ways either electrically, mechanically, or hydraulically. 1.3 Mechanical Details. One of the obvious points to be considered in the design of hull machinery is that it is necessary to design for the pitch, roll, trim, and list of the ship. All machinery should be designed for the following conditions: A pitch of deg (bow up to bow down) A permanent trim of deg by either the bow or stern A roll of 30 deg (each side) A permanent liet of 15 deg to either side However, no two of these conditions are considered to occur simultaneously. Additionally, it should be noted that in the case of equipment which is not used a t sea under storm conditions (e.g., cargo, topping, vang, and accommodation ladder winches) the design conditions listed only apply when the equipment is in the stowed condition. Every effort should be made to mount the driving and driven equipment on a common bedplate in order that alignment may be more easily maintained. In some few
I
instances this is impossible, as in the cases of capstans and windlasses where the driving and driven assekblies are frequently mounted on different decks. When such two-deck arrangements are employed, provisions must be made to accommodate the deck deflections and small amounts of lateral displacement between the two assemblies. Where a shaft connects the two assemblies, flexibility is usually provided by a slip or flexible coupling; if the connection between the two assemblies is hvdraulic piping, the relative movement between assembli~scan be accommodated by designing flexibility into the piping arrangement (i.e., by looping or offsetting the piping, taking care to avoid high regions which would trap air). Motors or gear boxes located in the weather should be provided with watertight housings, while those located below deck or out of the weather may be provided with dripproof protected housings. Bearings may be, in general, of either the sleeve or antifriction type. If of the sleeve type, the bearings should be of a good grade of bearing bronze. Where bearings will be subject to impact loading, as on the wildcat shaft of the windlass, they should be bronze rather than antifriction because of the possibility of "brinelling" occurring. Antifriction bearings should be used in any location where positive alignment, minimum lost motion, or freedom from wear is essential for proper operation. Some locations that fall into these categories are: motor bearings, steering gear differential control assemblies, worm and worm wheel shafts, and tension-sensing assemblies as on constant-tension mooring winches. Reduction gears should be totally enclosed and provided with an oil bath lubrication and with means for filling, draining, and measuring the lubricant level. Properly gasketed inspection and maintenance openings should be provided in all gear enclosures. It is suggested that the interior of gear cases be sand blasted and thoroughly cleaned before final assembly. The design and manufacture of gearing for hull machinery should be in accordance with recognized standards such as those of the American Gem Manufacturers Association. No compromise for strength should be allowed when designing the gearing, but it is considered acceptable practice to make some adjustments in the calculations for durability (wear) because most hull machinery operates on an intermittent basis. The wear rating of gearing for hull machinery can be increased because of its intermittent usage. The range of wear load service factors which may be considered, that is, the factors by which the nominal gear working loads can be reduced to arrive at the design wear loads, are given in Table 1. Spur, helical, and herringbone gearing should be of steel, should have machine-cut teeth, and may be heat treated to increase their ratings. Worms should be made of steel and be integral with their shafts, and worm wheels should be made of bronze. When a worm wheel is large, say over 8 in. in pitch diameter, it may be made with a bronze rim bolted to a steel hub. In order to obtain a high degree of mechanical efficiency, worm reductions should be of an overhauling type. There are, however,
J
Table 1
565
Wear Load Service Factor for Gears
, Fadm Capstans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .0.35-0.5 Crane machinery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 . 7 -1.0 Steering gear drives.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.0 Winches, cargo.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.7 constant-tension mooring.. . . . . . . . . . . . . . . . . 1.0 other.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.35-1.0 Windlasses. . . . . . . . . . . . . . # . . . ..,. . . . . . . . . . . . . . . . . . 0.35-0.5
-Equipmend
instances whereX'anonoverhauling feature is more important than' the efficiency considerations involved. Additionally, it is good practice to avoid the use of worms on overhung shafts because of the misalignment which can result from the worm shaft deflecting. Brake linings should be of an incombustible material that is not adversely affected by either heat or moisture. Woven linings are usually manufactured using a resinous binder. I t is suggested that the linings be baked in accordance with the manufacturer's recommendations to remove any excess amount of binder before the linings are installed. (There is additional discussion regarding lining materials and their installation in Section 2.2, which is devoted to the anchor windlass.) Where flexible couplings are used, as between motors and pumps, they should be of the all-metal type and preferably of the all-steel type. All fasteners exposed to the weather up to and including % in. diameter should be of nonferrous material or of stainless steel. Iiarger fasteners may be cadmium , plated or galvanized steel. Breathers and drain check valves should be of a nonferrous material or of stainless steel. Dowels and taper pins should be of stainless steel. Aluminum should not be used for watertight enclosures. The rope-contact surface of capstan, gypsy, and windlass heads should be hardened to a depth of a t least 4f6 in. to a minimum of 300 BHN to provide an abrasionresistant surface. The main rams for steering gears are hydraulic cylinder rods but, because they are invariably located below deck and out of the weather, they need no surface protection other than that provided by the hydraulic oil that adheres to them as they move in and out of the cylinders. All other cylinder rods, including those located in cargo holds where dust may be present, should be plated. The outer layef of plating should be of chromium because of its fine finish. However, chromium plating frequently contains cracks, invisible to the eye, which will allow seawater or moisture to penetrate and start corrosion. The corrosion spreads under the@er of chqomium and separates the plating from the steel rod. The rusted surface then is free to abrade the packing in the cylinder as the rod moves in and out. This can be avoided by fir& providing a plating of nickel on the rod. The nickel plating provides a completely impervious coating but is rough and would, in itself, tend to wear out the packing and cause leaks. Therefore, a layer of chromium must be added over the layer of nickel. Each coating should be
566
HULL MACHINERY
MARINE ENGINEERING
0.0005 to 0.001 in. thick. It is advisable to provide wipers on the rods to prevent dirt from being drawn into the packing. In some locations it may be advisable to provide boots on the rods to further protect against an accumulation of dirt on the working portion of the rod. If boots are installed, they should be of a clear plastic material so that any leakage of oil into the boot may be obsemed. Wherever possible, the arrangement of the rod and cylinder should be such that the rod is withdrawn into the cylinder when the equipment is secured so aa to protect the rod from surface deterioration due to being exposed to the weather for extended periods. Foundations for hull machinery are usually made of inverted angles that are welded to the deck. Each item of hull machinery is usually provided with a base that has a true, machined mounting surface. In order to ensure that the base of the equipment is not distorted when it is fastened to the foundation, it has been conventional practice to insert steel shims, chocks, or liners between the foundation angle and the base of the equipment. However, steel liners have two major disadvantages: (1) The space between steel liners is subjected to corrosive action and is almost impossible to maintain since the liners are not continuous and are only about % in. thick, and (2) The labor involved in fitting the liners is considerable since they must be inserted, checked for surface con- . tact, removed, filed or ground down, and reinserted, usually several times before an acceptable installation is obtained. As an alternative to steel liners, a chocking system using a poured epoxy resin may be used. Spacers (which may be nothing more than small pieces of $.i-in.-thick steel) are placed on the foundation angles. The unit of hull machinery is lowered onto these spacers, and f o u n h tion bolts are inserted but the nuts are run up only "finger tight." Three sides of each foundation mating surface are then provided with temporary dams and the space between the equipment base and the foundation is poured full of the liquid epoxy material. After about 48 hours (or such time as is required by the particular mixture being used) the temporary dams are removed and the foundation bolts properly tightened. No attempt is made to remove the spacers. The equipment is then properly seated and the epoxy between the equipment base and the foundation provides a degree of corrosion protection. This system has proven satisfactory and is accepted by the regulatory bodies. Considerable care must be taken in pouring the liquid epoxy, and only certain mixtures and methods of application are permitted by the regulatory bodies. Stress analyses should be made during the design of all hull machinery. These should include calculations for gearing, bearings, shafting, structural components, foundation bolts, etc. For merchant ships the normal-duty stresses should not exceed 40 percent of the yield point of the material and maximum stresses should not exceed 75 percent of the yield point; however, other design criteria regarding stresses may be required by the approval authority.
SUPPLY TO OTHER LOADS
A
-
3-POSITION VALVE
ACCUMULATOR 5-POSITION VALVE
- -
-
ACCUMULATOR
J
A
I
FILTER
I.
--
CHECK VALVE VALVE
+I I CHECK VALVE
Fig. 1
DELIVERY, PRESSURE COMPENSATED
Constant-flow hydraulic system
STRAINER
I Where automatic or oil-bath lubrication is not furnished, pressure gun grease lubrication should be supplied. Great care must be exercised in fabricating and installing both lubricating and hydraulic oil piping. In addition, all systems should be cleaned after installation by a method such as that described in reference L11.l Where it is possible for the manufacturer to do so, he should completely assemble, clean, and charge the equipment with oil to protect the equipment until it is placed in service. Equipments such as steering gears, winches, and hatch covers have been successfully handled in this manner. When hydraulic equipment is used, the bedplates for the pumping units should be provided with an oil-retaining rim so that leakage will be confined to a limited area. 1.4 Hydraulic Details. The hydraulic systems used aboard ships vary widely in variety and complexity; however, most shipboard hydraulic systems bear a degree of similarity with either the constant-flow, constantpressure, or demand systems shown by Figs. 1, 2, and 3, respectively. The major components of a constant-flow hydraulic system are the fixed-delivery pump, unloading valve, and accumulator. When the pump is started, fluid is delivered to the accumulator, which is usually of the pneumatic type, until the accumulator pressure equals that of the unloading valve setting. If there is no demand on the system, the unloading valve then opens to bypass the pump discharge back to the tank, and the pump runs continuously. At the same time, the check valve holds the fluid pressure in the system so that the energy stored in 1
Numbers in brackets designate References at end of chapter.
COMMAND FEEDBACK SIGNALS
RELIEF VALVE
i
Fig. 3
CHECK VALVE
,
Demand hydraulic system
7
4
Fig. 2
~onstant/preuurehydraulic system
the accumulator is immediately available upon load demand. When a throttle or control valve to an attached hydraulic actuator is opened, the accumulator immediately provides fluid to this load and the discharge causes a drop in system pressure. I n turn, the unloading valve closes and the pump discharge is returned to the active system. Upon a decrease in the load demand, the system pressure again rises to actuate the unloading valve upon recharging the accumulator, and the load cycle is complete. The sump tank is replenished by returns from the exhaust lines of the attached loads. Constant-pressure hydraulic systems, as shown in Fig. 2, generally use one or more variable-delivery pressurecompensated pumps which supply hydraulic fluid, a t a substantially constant pressure to either a system of multiple loads or to a single load such as a hydraulic elevator hoist. The constant+pTessure hydraulic system pump takes suction from a sump tank and discharges directly into the main supply piping. When the pumping capacity exceeds the load requirements, the system pressure increases to a predetermined value, at which point the pressure compensator acts to take the pump off stroke, thus stopping the flow of hydraulic fluid. A relief valve is provided to protect against overpressurization in the event that the pressure compensator fails to properly reduce the stroke of the pump. When the system demands cause the system pressure to drop to a preset value, the pressure compensator acts to put the pump on stroke, thus restoring the flow of the hydraulic fluid to recharge the accumulator and maintain the system operating pressure. The fluid flow to the individual loads may be controlled by a variety of types of valves, in addition to tlie three-position valve shown for simplicity in Fig. The 2. demand hydraulic system, as shown in Fig. 3, is a closed-loop system particularly adapted to meet the precise demands of varyi,n~,laadssuch as steering gears and automatic weapons handling equipment. - The variable and reversible delivery main pump does not take suction from a sump tank, as a loop of oil is maintained between the main pump and the load. The closed loop eliminates a significant amount of pressure drop which would exist if the load control were achieved by a servo valve located in the main hydraulic piping. A servo pump and a replenishing pump, driven by the same shaft
MARINE ENGINEERING
as the main pump, provide actuating fluid to stroke the main pump and to provide makeup fluid to the closed loop, respectively. Control of the main pump is accomplished by command and feedback signals. The command and feedback signals may be composed of a combination of mechanical, electrical, or hydraulic devices. I n one type of hydraulic steering system (see Fig. 11)) a rudder-position command signal originates at an electric servo located in the wheelhouse while a mechanical feedback signal, which relates the position of the rudder to the helm position, is originated from the steering gear. The input servo command signal is converted to a mechanical command signal at the steering gear control differential. The feedback is the other input to this differential which then measures the "error" between the helm and rudder positions. The "error" is translated into the form of mechanical motion to a rotary servo or other appropriate device which controls the flow of pressurized fluid to stroke the main pump. The response is one of direction as well as magnitude, so that the error is corrected in either the right or left rudder direction. The rate of rudder motion is dependent on the magnitude of the error or differential within the limit of the available pump stroke. For small differentials, as when holding s, course, the pump stroke is moderate and the rudder response is slow. When going hard over in an emergency, the wheel will be well ahead of the rudder, the pump will be forced on maximum stroke, and the rudder rate will be at its maximum. The same principle is advantageous in' hoistingmachinery drives in which a motor runs continuously in one direction to drive the variable-stroke pump, and the hydraulic actuator (motor or ram) is started, stopped, reversed, and driven at infinitely variable speeds within the system capacity by simply varying the pump stroke. Accumulators have a variety of applications in shipboard hydraulic systems. Fundamentally, accumulators are storers of energy. For shipboard use, they are usually pneumatic; that is, the hydraulic fluid is stored in a tank containing gas under pressure. The stored energy is then measured by the energy required to compress the , gas. This is recovered by allowing the fluid to flow into the hydraulic system under accumulator pressure. The pressure/volume characteristics are determined by the fluid/gas volume ratio and the initial charge (i.e. pressure) of the gas. The air and fluid may be in direct contact in the accumulator tank or they may be separated by a piston, sack, or diaphragm. If the hydraulic fluid is inflammable, the charge may be of nitrogen gas rather than air in order to minimize the possibility of an explosion. In typical applications accumulators are used to minimize the pumping capacity that is required in systems which have large oil flow demands followed by significantly lower demands or idle periods. Hydraulic elevators are a typical application of this type. A relatively small pump is used to charge an accumulator, and the elevator can be raised using the oil in the accumulator;
then while the elevator is idle, the pump can recharge the accumulator. A further advantage may be obtained in the recovery of energy by allowing the exhaust fluid from a descending elevator to charge an accumulator at a lower pressure. The pump then merely maintains the Dressure differential between the two accumulators. A second purpose of an accumulator is in applications where an instantaneous source of hydraulic oil must be provided in response to intermittent demands. In this case the alternatives are to either provide a continuously running pump or to provide a pump that is normally idle along with an accumulator. Central hydraulic systems for hatch cover operation contain accumulators for this purpose. Since it is not desirable to drive the pumps continuously, the accumulators are provided to store the energy for a sudden demand such as closing the covers quickly in the event of a shower. Hydraulic systems which operate valves contain accumulators for this same purpose; here the pump charges the accumulator and is normally idle because the accumulator contains sufficient oil to make several valve operations. Another purpose of accumulators is to provide a source of emergency power in the event of a power failure. For example, hydraulic elevators may have accumulators which permit a cycle to be completed in the event of a power failure. I n systems which must be held under pressure indefinitely, accumulators may also be used to make up small increments of possible leakage, thereby eliminating the need of frequent short pumping cycles. Central hydraulic systems are advantageous on some ships, especially large tankers where safety regulations limit the installation of electrically powered equipment and steam is undesirable because of the long steam and condensate lines involved. Hydraulic pumping stations serving two or more equipments have been used in a variety of applications aboard ships. Most difficulties experienced with such systems have been due to deficiencies associated with system basic engineering, system cleanliness, or system venting. The basic engineering of centralized hydraulic systems is of major importance to ensure that the system elements are compatible and are capable of functioning so as to meet all requirements. Cleaning as well as venting the system also starts in the design stage. Not only must access for flushing be provided during the design stage, but also the design must be checked to ensure that the flushing velocities will be sufficient to clean the system. Additionally, pockets and voids, which would tend to collect foreign matter, must be avoided. Three possible arrangements of centralized hydraulic systems which may be considered for merchant ships are as follows: 1. The bower anchor windlass and forward mooring equipment on one system, and the aft mooring equipment on another system. 2. All of the cargo, topping, and vang winches between a pair of hatches on a common system. Adj* cent hatch covers could also be served by these systems.
HULL MACHINERY
I
I
,
3. All hull machinery (except the steering gear) on a single system. In the first two cases, the pumping units would be located close to the driven equipment so that the mains would be comparatively short. In the third case, "the mains would be long but the pumping units would be located in the engine room where they could be serviced readily. The steering gear should be left out of this system since its average demand would be only a small percentage of the installed capacity of the central system and it would be uneconomical to run the large central system pumps continuously at sea. Furthermore, incorporating the steering gear into the central hydraulic system could degrade the availability of the steering gear in that an accident which ruptured a hydraulic line amidships could incapacitate the steering gear. Great care must be taken in sizing the pumps and mains for any of these central systems in order that there be adequate capacity for the maximum simultaneous loads. In the engine room system three or four pumps may be provided to handle the maximum load. These pumps could be cut in or out either manually or automatically as the loads vary. Most hull machinery hydraulic systems are either exposed to the weather or are installed in unheated parts of the ship. For these reasons, during the design stages consideration must be given to the ambient temperatures to be encountered and the viscosity characteristics of the oil selected. If it is not possible to warm up the system prior to actual operation, heating of the oil in the sump by steam or electric heaters may be desirable. Hydraulic piping that will be exposed to the weather should be of a corrosion-resistant material, such as stainless steel, except in the largest sizes used in ship-wide central hydraulic systems. If steel piping is used, its wall thickness should include an allowance for corrosion. Hydraulic systems for steering gears, windlasses, uinches, cranes, etc., should be provided with gage connections fitted with shutoff valves so that system performance may be checked during initial testing and regular overhaul periods. The gages should be valved off and should have gage snubbers in the line so that the gages will not be damaged by system pressure pulsations. Hydraulic sumps, or reservoirs, should generally be large enough to contain 110 percent of the volume of oil in the system so that the system may be drained to the sump for servicing or maintenance. Even when a system cannot be drained+tothe sump, because the units served are below the sump, it is good practice to provide a sump large enough to contain all of the oil in the system; that is, after .the entire system, including all variable-displacement units, has been filled, the pump suction should still be flooded to the operating level. Pumps taking suction from a sump tank should be fitted uith a suction strainer, and the pump discharge should be through a full-flow filter. The filter should be equipped with an indicator that will show the cleanliness of the filtering element.
569
1.S Electrical Details. The characteristics of electrical equipment are discussed in Chapter 17. Particular reference is made to Section 4.3 of Chapter 17 because that section reviews the electrical characteristics of motors which are an important consideration in motor selection. Electrical equipment should be either watertight or dripproof protected as dictdted by its location, in or out of the weather. Motors should have natural ventilation in most instanqe~and may have bearings that are either grease lubricated or permanently lubricated and sealed; the latter is becoming the more common. In the case of grease-lubricated bearings, the bearing housings should be arranged such that the grease will not be forced into the windings. Watertight motors should be equipped with an automatic drainage fitting. Where a brake or gear is to be mounted directly on the motor shaft, the shaft should be tapered and fitted with a key and lock nut. Electric brakes of the spring set, solenoid released, shoe type have proven to provide more satisfactory service than those of the disk type. Nevertheless, where disk brakes are properly designed and applied, they can provide entirely satisfactory service. Brakes should be of sufficient capacity to stop and hold the load under any condition of operation or testing. Brakes are commonly designed to support 200 percent of the normal operating load, and tests should be performed (particularly for cargo gear) to confirm this capability. Watertight brakes are frequently fitted with an external hand release so that, in the event of a poiher failure, the load may be safely lowered without the necessity of removing the brake cover. A hand release also avoids the hazards associated with the alternative of backing the springs off. Direct-current brakes are preferable to alternating-current brakes even though they require the addition of a rectifier to provide the direct current. The direct-current units have a better response when they set, and alternating-current brakes have a tendency to "chatter" when setting. The enclosures for watertight shoe brakes should be of %in.-thick stainless steel or of x-in.-thick mild steel that is treated both on the inside and outside with an inorganic zinc coating, and the hardware should be corrosion resistant. The slight extra cost for these designs will ensure that the enclosure will last for the life of the ship. The watertight enclosure should be fitted with an automatic drainage fitting. Where possible, control panels should be arranged so that they may be serviced from the front. This will allow panels to be mounted back&+back or against a bulkhead. Where the panels are to be mounted in the weather, the enclosures should be of stainless steel with corrosionresistant hardware. Limit switches are a constant source of trouble, particularly when they are exposed to the weather. Their reliability is further impaired if they am subject to being coated with paint when the equipment is painted. Great care should be taken in their selection to ensure
'
MARINE ENGINEERING
that they are of rugged construction. Shafts in limit switches should be of bronze, monel, or stainless steel. Master switches for variable- or multiple-speed equipments should be of the cam type with vertical handles having integral knobs. They should be permanently marked to indicate operating directions and, if the control is of the multiple-speed type, there should be detents so that the operator can sense the speed points. The enclosures should be of either ductile iron or bronze and should be raised to a convenient working height. Many owners prefer that even cam-type master switches be of the "spring return to OFF" type to guard against the possibility of the operator leaving his position with the equipment running. If this type of switch is provided, the detents must be very slight so that there will be no chance of the handle "hanging up" on one of the running points. All hull machinery control equipment must incorporate undervoltage protection due to regulatory body requirements. If the master switch is not of the "spring return to OFF" type, the control circuitry should be arranged
HULL MACHINERY
so that it is necessary to return the master switch to the OFF position to reset the undervoltage relay. If the controller overload relay is of the manual reset type at the controller, no extra undervoltage control relay need be provided. If the controller overload relay is of the automatic reset type or if an automatic reset temperature-sensing element is built into the motor, undervoltage protection should be provided by an undervoltage relay which will not reset until the master switch is returned to the OFF position. Heaters should be provided for all watertight motors and brakes and, where space permits, in watertight master switches; heaters should also be provided for all motor controllers. The heaters should be connected to an electric circuit that is continuously energized and independent of the power feeders. One method of providing heating for alternating-current motors is to provide reduced voltage in one of the three phases. All heater circuits should be p a n g e d so that they will be interrupted when the main power circuit for the motor is energized.
HAND PUMP
DIFFERENTIAL ASSEMBLY TRICKWHEEL /
Fig. 4
Rapson-slide type steering gear
Section 2 Hul Machinery Installations 2.1
i
Steering Gear
a. Introduction. Steering gear arrangements in great variety have been used to control the position of ships' rudders. The various types of steering gears which have been used are described in detail and illustrated thoroughly in references [2], [3], and [4]. As may be noted from a review of the early steering gear designs, the advent of the Hele-Shaw and Vickers variable-stroke pumps made possible the design of hydraulic steering gears of a torque capacity adequate for the largest vessels. These forerunners of many commercial vumvs of similar vrinciple are driven co~tinuously,while in operation, in one direction of rotation. The pump discharge is varied in rate and direction by controlling the stroke of the pump pistons (see Chapter 12 for discussion regarding variablestroke hydraulic pumps). Right and left rudder motion or the holding of rudder position is accomplished very smoothly by the use of these pumps. These characteristics have made the electrohydraulic steering gear, driven by electric motors and variable-stroke pumps, the favored selection for installation on oceangoing vessels. To simply state that a steering gear is of the electrohydraulic type is not definitive because there are several types of electrohydraulic steering gears which offer advantages in particular applications. A common type of electrohydraulic steering gear is the Rapson-slide type. This type is well suited for very high rudder torque ratings in that the Rapson-slide arrangement provides an increasing mechanical advantage at larger rudder angles (see Section 2.1 c); however, its use is not confined to
high-torque applications. Rapson-slide steering gears may be of either the single ram group type (as is common for low-torque applications) or the double ram graup type as il]us$rated by Figs. 4 and 5. The steering gears in Figs. 4 and 5 ayp f p t h e r classified as being of a "unitized" construction. A uqitized construction is one in which the entire steering gear is assembled as a unit at the manufacturer's plant and is installed as a unit by the shipbuilder (as opposed to constructi~lgthe ram groups and power plants as separate entities, each of which must be vrovided a foundation by the shipbuilder and each of khich must be connected togethe; by the shipbuilder). A unitized construction is not only economical in terms of shipboard installation costs, but it alsq reduces opportunities for foreign matter to be introduced into the hy: draulic system because the unitized gear is completely assembled by the manufacturer and shipped full of oil to ensure preservation. A link type of steering gear, typically illustrated by Fig. 6, also employs a ram group; hoyever, the principle of operation is considerably different, as discussed in Section 2.1 c. Link-type steering gears generally have a decreasing, rat,her than increasing, mechanical advantage at larger rudder angles; however, there are two cases for which a link type of steering gear is ideally suited. One is a case where there is insufficient space around the rudder stock to permit the installation of rams (by comparing Figs. 4 and 6 it will be seen that the space required in way of rudder stock is considerably different for the two types). Another situation well suited for a link-type
Fig. 5
Four-cylinder, unitized, Rapson-slide steering gear
steering gear is a twin-rudder ship that can be arranged such that one link-type steering gear can serve two rudders, as shown by Fig. 6. Yet another type of electrohydraulic steering gear is the rotary vane type which is illustrated by Figs. 7 and 8. Rotary vane steering gears have no rams; with these units the rudder torque is developed by the differential pressure that acts across the vanes which can be seen in Fig. 8. Rotary vane steering gears are not suitable for the higher torque ratings, but for rudder torque ratings of about 15,000,000 in.-lb and less they offer the advantages of low space requirements, low installation costs, and a low weight. b. Rudder Torque Rating. One of the more nebulous aspects of the ship design process is the establishment of the maximum rudder torque for which the steer-
ing gear should be designed. 8hip specifications rarely stipulate the maximum design rudder torque; instead, the specifications usually state that the steering gear shall be capablh of moving the rudder a t a prescribed rate when the ship is proceeding at maximum speed. With the ship specifications written in this manner, the shipbuilder has the responsibility of establishing the maximum design .,+rudder torque. Some of the analytical procedures which have been employed to estimate the maximum design rudder torque are discussed in detail in references [5] and [6]. It may be noted, however, that there can be a considerable difference between the rudder torque values predicted by the various procedures. The situation becomes further clouded when attempts are made to correlate analytical predictions with test results.
-
572
---
HULL MACHINERY
MARINE ENGINEERING HAND PUMP (FILL AND DRUN)
k O F LINKS
.-1
HYDRAWIC PIPING NOT SHOWN
-L & -
.-,
ELECTRIC M O T O R
CHANGE OVER /VALVE CONTROL
HYDRAULIC PUMPSTROKE CONTROL LINKAGE
\
Fig. 9
Link motion diagram
DIFFERENTIAL CONTROL 'ILLER (RUDDER HARDOVER
-+a TILLER (RUDDER AMIDSHIP)
Fig. 6
U
/
'RAM
CYLINDER
Link type hydraulic steering gear for a twin-rudder ship
Fig. 7 Vane type steering gear with dual power units
In many cases the data resulting from steering gear tests consist of only the time required to complete a maneuver and the maximum pressure that was observed during the mafieuver, with both bits of data observed and manually recorded by a data taker. The value of such data is s&erely limited because the rate of rudder movement during the maneuver is not known with certainty (e.g., the power limiter may have reduced the pump stroke during part of the maneuver); and the "maximum pressure" recorded could be an "average" maximum pressure, the maximum pressure sustained for a "significant" period of time, or the peak pressure ob-
Fig. 8
Vane type steering gear with cover removed
served. Additionally, with ram-type steering gears the rudder angle at which the pressure occurred must be known in order to relate the pressure to a torque value. When the data are taken manually, the rudder angle is
generally not known, but the maximum pressure is often assumed to have occurred at the maximum rudder angle. In view of the above, it should not be surprising that the "maximum" rudder torques reported from sea trials of sister ships have varied as much as 50 percent. More sophisticated test instrumentation has been used to continuously record the rudder angle, ram pressure, ship's heading, etc., as a means of defining the test conditions. While this procedure gives definitive rudder rate data, problems are nevertheless encountered in establishing the "maximum" rudder torque observed due to the errlttic characteristics of the data. For example, spikes in the recorded pressure data may reach extraordinarily high values and be of such a short duration that the relief valves do not lift. The maximum pressure observed should not be taken to be the peak value, but some lesser value, e.g., the maximum value sustained for a 0.2-sec time period. The most reliable method of estimating the maximum rudder torque required for a given ship is to first select a similar ship for which trial data are available. Next, using one or more computational procedures, such as described in references [fi] and [6], estimate the rudder torque required for the similar ship and then compare the estimate with the trial results. If there is a significant disparity between the calculated torque and the torque established by the trial results, as is usually the case, determine a correction factor that can be applied to the estimated torque to male it agree with the trial results. After the correction factor has been established, use the same computational procedure and apply the established correction factor to estimate the rudder torque required for the ship in question. Even after completing this procedure, the proper maximum design rudder torque can seldom be stated with certainty; however, in the final analysis, a single design value must be selected. c. Ram Group. Hydraulic steering gears employing rams are generally of either the Rapson-slide type, shown by Fig. 4, or the link type, shown by Fig. 6, with the former being the more common. A force diagram for a link arrangement is shown by Fig. 9. As can be seen, when the rudder is hard over, the effective crosshead arm (i.e., the torque arm) is very nearly equal to the crosshead radius multiplied by the cosine of the hardover angle with
Fig. 1 0
Rapson-slide force diagram ~ :-
the link arrangement. The torque arm therefore is usually least when the torque to be developed is a t its maximum; as a result, the forces required with this arrangement are relatively large. The torque developed by a single-ram link arrangement is
Q
=
FR cos a
(1)
where
Q = torque developed, in.-lb F = ram force, lb R = crosshead radius, in. a = rudder angle, deg
As may be seen from Fig. 10, an increasing mechanical advantage is obtained at larger rudder angles with the Rapson-slide type of mechanism. The torque developed by a single-ram Rapson-slide arrangement is
where the terms are as defined for equation (1) and shown by Fig. 10. For the same ram force and crosshead radius, the torque which can be developed by a Rapson-slide arrangement is greater than the capability of a link arrangement by a factor of cosa a. At a rudder angle of 35 deg, this factor is 55 percent. In view of the mechanical
HULL MACHINERY
MARINE ENGINEERING
I
i
I
advantage offered by a Rapson-slide arrangement, it may appear that a link arrangement need not be considered; however, such is not the case as discussed in the foregoing. Ordinarily, ram groups of the link type are provided with two links for each ram, one above and the other below the ram; likewise, Rapson-slide mechanisms are usually built with crosshead arms above and below the ram. One or more tie rods should be provided betwecn the two cylinders of each ram group. The tie rods are usually shouldered at the cylinder bosses in which they are mounted, and set up with double nuts. Being of exact length, they aid in aligning the cylinders both while the steering gear is in the course of fabrication and installation and while in service. The tie rods are not necessary for the transmittal and distribution of axial loads but they are convenient guides for bushings or shoes carried in extensions of, or bracketed from, the rams in order to prevent rotation of the rams about their longitudinal axes. Bronze bushings are usually installed in the cylinders to provide a suitable bearing surface for the ram; the bushings are made to form the inner ends of the stuffing boxes. With the link type of construction the bushings carry little load other than the weight of the ram itself; the lateral load introduced by the link is slight, even at the maximum obliquity. However, with a Rapson-slide mechanism, the ram is subjected to a considerable lateral force, and since the forces are usually the largest at a hard-over rudder angle, when the crosshead is near one cylinder or the other, one of the bushings will carry by far the greater part of the load. When a ram is made of high-strength steel, which permits high stresses with no improvement in the modulus of elasticity, it is conceivable that the deflection of the ram may be great enough to cause binding in the cylinder bushings. Consequently, the bushings for Rapson-slide mechanisms should be given careful study. Since the maximum axial load on a ram is limited by the hydraulic pressure, which is only a small fraction of the permissive stress in the ram metal, column or even strut strength calculations are unnecessary. However, ram bending calculations can be important for Rapsonslide designs when, as is sometimes done for weight reduction, the rams are bored out from both ends or when the gear is designed for an abnormally large rudder movement (e.g,, Great Lakes ships which are commonly designed for a 45-deg to 45-deg rudder movement); in such cases flexural stress may also become a critical consideration. , Carefully designed ram groups have a crosshead radius (or with a Ra~son-slidemechanism, the distance from the
ram stroke, with the ram diameter being correspondingly increased. The maximum design hydraulic pressures for steering gears generally fall in the range of 1500 to 2500 p i with naval ships in the upper end of the range and conserva-. tively designed merchant ships occupying the lower end of the range. ~ e r c h a i tship specifications commonly require that the steering gear be capable of moving the rudder from hard over (35 deg rudder) on one side to within 5 deg of the hard-over position (30 deg) on the opposite side within 28 sec when proceeding ahead a t the maximum rated propeller rpm. The timing cycle is normally terminated 5 deg before the hard-ovcr rudder position in order to avoid timing the rudder movement as the follow-up mechanism is reducing the pump stroke. For merchant ships, the rudder rate when going astern is commonly not specified, although it has been stipulated to be one half of the value specified for ahead steering. The rudder rate specified for naval ships varies from one ship to the next. Generally, it ranges upward from 29 deg per sec for ahead steering, with no rate specified for astern steering. In order to avoid having astern considerations control the capacity of steering gears for naval ships, it is common practice to conduct special astern steering tests for the purpose of establishing the maximum astern speed a t which the rudder can be moved without exceeding the ahead rating. A sign warning that there is a hazard of overloading the steering gear at astern speeds greater than that established is then made and installed in the helmsman's view. Table 2 illustrates the considerations which enter into the preliminary design of a steering gear and gives a procedure for computing the approximate characteristics of a steering gear. If only an approximation of the maximum motor horsepower requirements is desired, the following expression may be used for a steering gear of the Rapson-slide type:
proximately three times the ram diameter. It is convenient to use this ratio for preliminary design purposes bearing in mind that arrangement considerations may necessitate shortening the crosshead radius, and therefore the
to the maximum pressure observed during the timing cycle commonly occurs at the 30 deg position before reaching hard over, and with the exception of Q , all other terms in equations (3) and (4) are considered to have con-
H =
&(tan 4
torque at maximum pressure during timing cycle, in.-lb a = Ndder angle at =
and all other terms are as defined in Table 2. The corresponding relationship for a link-type steering gear is
H =
Type of Linkage
Rapson
I
I
Number of rams.. . . . . . . . . . Maximum permissible operating pressure, psi. . . . . . . hlaximum rudder angle, deg . Rudder augle at end of timing cycle, deg. . . . . . . . . . . Rudder torqne at 0, in.-lb.. . ~Iaximumdesign torque ahead, in.-lb. . . . . . . . . . . . Rudder angle at Q1, deg. . . . Maximum design torqne astern, in.-Ib.. . . . . . . . . . . Rudder an le at Q2, deg. . . . am-to-rujder efficiency. . . Approximate crosshead radiu~,~ in.. . . . . .. . . . . . . . Approximate ram diameter = R1/3, in.. . . . . . . . . . . . . Crosshead radius used, in.. . . Ram diameter used. in.. . . . . Maximum ram pressure aheadlbpsi.. . . . . . . . . . . . . . . blaximum ram pressure a ~ t e r npsi.. ,~ .. . .. . . . . . . .. . Ram stroke during timing cycleldin. . . . . . . . . . . . . . . . Oil volume during cycle = a DeZS, cu in.. . . . . . . . . . Duration of timing cycle ahead,*sec.. . . . . . . . . . . . . Oil flow requirements, 0.2597V!t1 gpm.. . . . . . . . . Pump effimency.. . . . . . . . . . Maximum ram pressure ahead during timing cycle,f psi . . . . . . . . . . . . . . . . . . . . . Maximum motor horsepower requirements =
paC/liY4e,rhp. . . . . . . . . .
Motor horsepower used? hp .
2
1700
Link
2
1700
35
35
27
27
D' R D PI pa
S V
t
124
e
0.85
183 0.85
pa
1075
940
H HI
91.5 75
100
C
118.1
Nofes for Table 2 -
.
+
tan 0) cos2 a 66OOEte
where
&
Table 2 Preliminary Design Calculations for an Eledrohydraulic Steering Gear
+
&(sin 4 sin 0) 6600Ete cos a
I
QI and 4 as well as Qa p.nd cue are used in the above equations and the largest value of R' is used. 9 1 = 4Q1 cose cul/rZREDefor Rapson p1 = ~ Q I / ~ Z R E cos Dul~ for link Cheeks should also be made to ensure that higher pressures do not occur at smaller rudder torque values. " See note b using Qe and cua vice Q1 and ul d S = R (tan + tan 0) for Rapson S = R (sin + sin 0) for link a Duration of the ahead timing cycle is 28 ~econdsminus one second for stroking the pump. I The maximum ram pressure during the ahead timing cycle often occurs at the 30 deg position before reaching hard over; however, the ram pressures at other angles should be checked, using an equation of the form given in note b, to ensure that none are larger. 0 A similar calciilation may be required to confirm that the astern horsepower requirements are not a controlling consideration. h The amount of motor overload permitted for merchant ships is typically 25 percent during ahead steering and 50 percent during astern steering [7]; naval ships me permitted ta be averloaded 50 percent during both ahead and astern operations [8].
++
575
stant values in many cases. As a result, equations (3) and (4) can each be reduced to a constant times Q. d. Steering Gear Power Units. Steering gear power plants are required by the regulatory bodies to be installed in duplicate, with each power plant having the capability of moving the rudder a t the specified rate under all conditions of ,full-speed maneuvering. Thus, a completely redundant power/plant is installed. Each power plant consists essentially of an electric motor driving a"wariab1e-stroke pump. Motors for the steering gear , pumps are generally of the 440-volt, 3-phase, 60-cycles per second, squirrel-cage induction type and are dripproof protected. When the steering gear is powered by continuously operated, reversibleflow, variable-delivery pumps, the motor should be rated for continuous duty a t 15 percent of rated load, followed by full load for oiie hour. Since the motor is required to start while under load, a class C motor is used (see Section 4.3 of Chapter 17). As a means of reducing the motor horsepower rating required, horsepou7erlimiters are often provided. Horsepower limiters (secondary stroke-control devices) operate on the principle of reducing the stroke of the main hydraulic pump in the event that the hydraulic pressure exceeds a preset value (that pressure corresponding t o the motor rating plus the allowable overload). Since the power required to drive a pump is proportional to the product of the discharge flow and pressure, the rating of the installed motor horsepower can be minimized by reducing the pump flow during the_periods of peak pressure requirements. Horsepower liinlters are usually designed such that the stroke of the hydraulic pump is decreased with increasing hydraulic pressures so that the motor horsepower required is nominally held constant a t the preset value. When the horsepower limiter acts to reduce the pump stroke, the rudder rate is also reduced; however, if sufficiently high rudder rates can be attained during other regions of the timing cycle, where the torque is lower, and the overall average rate during the timing cycle exceeds the minimum rate specified, the horsepower rating of the driving motor may be reduced. Usually, the replenishing and servo pumps are driven by the same motors that drive the main pumpa. If, however, these auxiliary pumps are separately driven, their motors should be rated for continuous duty. If separate motors are provided for these pumps, each should be interlocked with its main pump motor so that the main pump cannot be started without first starting the auxiliary pumps. Steering gear motor controllers should be provided with under-voltage protection. It is customary to shift o y r from onsgower unit to the other a t regular intervals, in order that both power units may operate approximately equal lengths of time. Main pump selector valves in the hydraulic piping are used to effect the shift hydraulically. For manual control, these may include two common stop valves for each pump, or dual 4-way or &way plug or piston-type valves. Transfer valves for dual power unit systems are usually arranged in one of the following modes:
576
MARINE ENGINEERING
The control system may be of the "non-follow-up" type, "on and off" type, or the "full follow-up" type. I n the first, the helmsman applies left or right rudder to hold a course. As long as a wheel, lever, or pushbutton is held to energize the steering gear for rudder response in one direction, the rudder moves in that direction until the steering gear is deenergized automatically in the hardover position. Of course, the rudder may be moved in small increments, left or right, by holding the control off neutral for small intervals of time. The helmsman's experience and judgement are important factors in minimizing steering gear activity and oversteering. With the second type, the rudder returns automatically to amidships upon the helmsman's release or centering of the control. The full follow-up system, the third type of control, Another system includes automatic transfer valves senses any difference between helm and rudder angles and which can be controlled in the steering gear room or from the direction of the difference; it moves the rudder autothe remote steering stations. These valves are operated matically in the proper direction to eliminate this difby main pump servo pressure and are controlled by soleference and holds the rudder at the angle signalled by the noid-operated pilot valves which are connected elechelm until a difference is reestablished by moving the trically in the main pump motor control circuits in such a helm or by drifting of the rudder due to hydrodynamic manner that the pump driven by the first motor started forces. "Feedback" and "differential" as well as "followis connected automatically into the hydraulic circuit and up" are terms often used in describing this control since becomes the active pump for the operation of the steering gear. If the second pump motor is then started, this the correction is derived from "feedback" of the rudder motion; the control senses or in some cases measures, pump is bypassed hydraulically. By stopping the first within limits, the difference or differential between the motor, the second pump becomes the active pump autohelm and rudder angles. Representative types of followmatically. This system ensures maximum safety when a up control systems are described and illustrated in refervessel is under way in restricted waters in that with both ences [2], [3], and [4]. units operating, the idling bypassed pump becomes the Remote control systems in common use include active pump automatically upon electrical or servo presmechanical, hydraulic, and electrical types. Shafts, wire sure failure of the other unit. rope, sprockets and chain, push-pull flexible control Valves may also be included in a four-cylinder ram cables, and their combinations are used to transfer mogroup piping system which cut out and bypass any two oppositely acting cylinders, thereby providing for two- tion proportional to that of the helm from remote steering stations to the local control input at the steering gear. cylinder "emergency" operation. These mechanical means are simple and reliable for use When a hand "fill-and-drain" pump is specified, it is in smaller vessels. usual to stipulate that it be possible to use this pump as a "rudder positioning" device. Such a pump is convenient , The hydraulic type, represented by the hydraulic if a ship is docked and no power is available to operate t,elemotor, came into common use as a remote control dethe steering gear, or to provide rudder trim to aid in tow- vice with the earliest applications of power steering. I n ing a dead ship. The fill-and-drain pump should not be older designs, the motion of the steering wheel displaced considered as an "emergency steering" pump unless the a piston or rams to force fluid to and from a receiving unit system has been designed to suit an emergency steering consisting of a cylinder and double-acting piston assemrequirement such as that outlined in the Rules of the bly in the steering gear compartment [2, 3, 41. The principal types of electrical remote controls inAmerican Bureau of Shipping [9]. Also, suitable check * valves should be included i n the fill-and-drain pump p i p ing to prevent the pump from being overhauled and 1. Self-synchronous a-c "rotary transformers" with a driven as a hydraulic motor. transmitter driven through step-up gearing by the The hydraulic systems are protected by suitable relief steering wheel and receiver providing proportional valves. When hydraulic braking is stipulated, a seconinput motion to the local control. dary set of rehef valves of the cross-blow type may be in2. Balanced-bridge circuitry in which turning the stalled between the selector valves and the ram cylinders. wheel moves a potentiometer to unbalance the elece. Steering Gear Controls. Steering gear control trical system. Unbalancing the electrical system systems usually provide the capability of controlling the starts an intermediate powered servo which prosteering gear remotely from a steering console located in vides input to the local control and moves a followthe pilot house and locally from the steering gear space by up potentiometer which rebalances the circuit and means of a trickwheel. A schematic diagram of a control deenergizes the servo when the helm and relative system typical of naval ships is shown by Fig. 11. servo angular positions coincide.
( a ) Both pumps de-energized; rams blocked. (b) Pump A energized and "on the line" with the ram group; pump B blocked to the ram, and bypassing if energized. (c) Pump B energized and "on the line" with the ram group; pump A blocked to the ram, and bypassing if energized. I1 ( a ) Both pumps de-energized; rams blocked. (b) One pump energized and "on the line" with the ram group; opposite pump de-energized. (c) Both pumps energked and "on the line" with the ram group, thereby driving the rudder at almost double the rate specified for singlepump operation.
J
.
HULL MACHINERY
LEGEND
- - --- MECHANICAL ELECTRICAL
z8"" HELM
RUDDER ANGLE lNDlcATO~60
o
c3
HELM ANGLE
I
PILOT HOUSE
HYDRAULIC MAIN SYSTEM PRESSURE HYDRAULIC CON1ROL PRESSURE
I
DISCONNECT SWITCHES
Fig. 1 1 Schematic diagram of the controls for an electrohydraulic steering gear
9
578
The powered servo is usually an electrohydraulic unit which is essentially a miniature steering gear and is designed for application to any type of steering gear as one link in a chain of servomechanisms from the steering wheel to the rudder. It is also used as an automatic pilot when the input is taken from a suitable compass. Arrangements for this purpose are proprietary and include other control features providing for adjustments to suit the response of the vessel, the sea state, and the course followed. It is completely feasible to omit the power unit as a separate servo if the balanced bridge circuit is used to control the main steering gear directly. Using the electrohydraulic steering gear as an example, the usual mechanical control differential is omitted. The power unit is used to put the main pump(s) on stroke, when the bridge is unbalanced, in the hydraulic flow direction for right or left rudder depending on the polarization and resulting , flow direction of electrical current through the bridge circuit. The balancing potentiometer is then moved in direct proportion to the motion of the rudder. The rudder is held at the selected new angle, as the pump stroke is returned to neutral when the circuit becomes balanced. The first Of such arrangements is discussed in reference [lo]. Steering gear power units which are designed for naval ships in accordance with reference [8] are required to contain features which limit the speed of the hydraulic pump and electric motor in the event that electrical power is interrupted and t.he rudder takes charge of the hydraulic ram group, thereby driving the hydraulic pump and electric motor. The method of limiting the overspeed of the electric motor may be either an arrangement of springloaded valves operated by servo pressure, or a brake on the electrie motor. The method usually employed is to incorporate a hydraulic lock feature in the blocking valve and associated solenoid valves which operate the bloeking valve, as shown in Fig. 11. As may be seen from Fig. 11, in the event of an interruption of electrical power, the bloeking valve is shifted so as to place a hydraulic lock on the ram group and isolate the ram group from the pump. It should be recognized that other valve arrangements may be used to provide the same feature. During sea trials of naval ships, steering gear power failure tests are conducted with the ship proceeding at full speed in the ahead direction and at the maximum astern speed permitted for unrestricted rudder operation. The power supply is interrupted at a series of rudder angles and during the circumstances in which a power failure would have the most adverse effects. Before sea trials, it is good practice to ascertain the closing time of the blocking valve by interrupting the electrical power when operating the steering gear power units with the ship at the dock. The anticipated motor overspeed during power failure tests at sea can then be calculated by using the measured closing time of the blocking valve and the
579
HULL MACHINERY
MARINE ENGINEERING
rudder design torque values. Experience has shown that a maximum closing time of approximately 0.25 sec for the blocking valve will normally assure that the speed of the electric motor will be limited to 150 percent of the normal operating speed. The detailed specifications for a vessel usually include very complete descriptions of the steering gear and its control syetems. These may follow arrangements on similar ships which have given satisfactory service, or they may reflect the owners' or naval architects' preferences. I t is desirable to consult steering gear manufacturers early in the design stage of a vessel in order to facilitate the selection and arrangement of the steering gear. 2.2
Anchor Windlass
a. Introduction. The ship specifications usually require that a windlass be capable of hoisting the anchor at an average speed of not less than 5 or 6 fathoms per minute (30 or 36 feet per minute, respectively) from a depth of 30, 60, or more fathoms. The required chain pull thus is dependent not only on the weight of the anchor but also on the weight of the chain to the specified depth, from which total the water buoyancy may be deducted. Even though the windlass may be constructed as a sophisticated machine (some windlasses are fitted with ingenious mechanisms), it must be built to perform the crudest task on shipboard. The anchor chain is heaved in through a hawsepipe in which the friction loss runs from 35 to 40 percent in good designs and sometimes exceeds 55 percent; the chain is operated by a wildcat which is hardly comparable in efficiency to the sprockets used in refined chain drives. The wildcat is usually made with five whelps (comparable to a 5-tooth sprocket), causing the chain to move with a jerkiness which is aggravated by its tendency to turn over or "slap" in the hawsepipe and to slip on the wildcat. Windlasses therefore require more ruggedness of construction than any other machine on board ship. The Rules of the Classification Societies [91 contain tables of required equipment consisting of anchors, chain cable, towlines, and hawsers. These items are identified by, and sized in accordance with, an "equipment number" or "equipment tonnage" included in the tables. The number or tonnage figure is calculated by substituting in' empirical formulas certain dimensional and displacement measurements of the vessel for which the equipment is desired. This has become a standard mode of equipment selection for oceangoing merchant vessels. The basic dimensions and scantlings of an anchor windlass depend on the anchor weight and chain size. These follow, therefore, from the equipment selection to suit a given case. The size of the vessel, the nature of the service, and the desired anchor handling and stowage arrangements are also contributive in the choice of the anchor windlass. In many cases, the windlass is used only in emergencies. However, in some cases the windlass may be
440V POWER CONTROLLER
I
II
L
H A N DLE
v
I I
LEGEND ,= 1
ELECTRICAL MECHANICAL
----
HYDRAULIC
Fig. 12
Schematic diagram ef a horizontal electrohydraulic windlass
used regularly (e.g., some operators set an anchor in a "flying moor," approaching a dock, far enough offshore to warp the bow into the stream when preparing to get under way, thereby reducing or eliminating the need of a tugboat). Also, it is usual to install capstans or warping heads, driven by the windlass power plant off the windlass gear train; therefore, the windlass may be used for normal warping duty. As with the other items of hull machinery, anchor windlasses are available from specialty manufacturers who can provide machinery to suit a variety of requirements. Their recommendations are very helpful in the shi~ design stage. 6. wrndlass Types. There are two fundamental configurations of anchor windlasses which have evolved from the "winding log" and capstan of earlier times, namely the horizontal arrangement and the vertical arrangement. The horizontal windlass is a specialized winch which is powered by a steam or internal combustion engine or an electric motor. The prime mover is connected either directly or indirectly (through a hydraulic transmission) to a train of gearing which drives one or more chain
-
Fig. 13
Horizontal-shaft, double-Wrjt windlassdriven by a d-+ electric motor
sprockets, called "wildcats," through sliding-block "locking heads" or comparable jaw clutches. Figure 12 is a schematic diagram of a horizontal electrohydraulic windlass. A photograph of a horizontal windlass that is driven directly by a d-c motor is shown by Fig. 13. In its
HULL MACHINERY
MARINE ENGINEERING
/
COMBINATION MOORING WINCH P/S ANCHOR WINDLASS P/S
.
Fig. 15
I MAIN DECK Fig. 14
Forecastle of a cargo ship with combination mooring winch/anchor windlass units
simplest form, the entire machine is self-contained on a common bedplate. Each wildcat is provided with a brake of the band type to restrain the chain when the anchor is let go under a controlled drop, for veering chain to the desired scope, and for holding the chain while the chain stoppers are being attached. One or more warping heads are usually keyed to overhangs of intermediate shafting. The gear train through which the warping head is driven usually affords a line pull in the order of one quarter of the available wildcat chain pull, at four times the normal chain speed. The self-contained, horizontal type of windlass is the least expensive in terms of installed cost. However, it requires more maintenance than does the vertical type because, aside from the protection provided by tarpaulins, etc., the windlass machinery is completely exposed to the weather and to the spray and waves which break over the bow during storm conditions. With the continuing increase in the size of vessels and with more general adoption of wire rope mooring lines, which are stowed on the drums of mooring winches, a useful combination is that of a horizontal mooring winch with a clutched drum driving a chain wildcat through an auxiliary gear reduction and sliding pinion or jaw clutch. A forecastle arrangement employing combination mooring winch/anchor windlass units is shown in Fig. 14. In the preliminary design stage of a vessel, it is good practice to develop the anchor handling arrangement to the extent that the chain leads are confirmed to be satisfactory. In the case of ships with large bulbous bows,
the anchors must be located farther aft so that the anchors will not hit the bulb when they are dropped. This usually requires that two separate windlasses be provided with each set at an angle to the s h i ~centerline in order to obtain proper lead; to the ha\;sepipes. For further discussion relative to windlass deck arrangements see reference [ l l 1. A vertical windlass consists essentially of a wildcat mounted on a vertical shaft which is carried in a rugged set of main bearings in a casting or weldment which is bolted or welded to the deck, as typically illustrated by Figs. 15 and 16. The strengthening of the deck and supporting ship structure in way of this assembly is ubually made adequate to sustain all anticipated loads due to the chain pull, independently of the main shaft extension to the deck below. The wildcat is brought as close to the deck as possible in order to minimize bending moments due to the chain pull. The chain wraps approximately 180 degrees around the wildcat. and then enters a chain deckpiie leading to the chain locker. The shafts from vertical wildcats and associated capstans are extended to one or more dkcks below where they are coupled to main and intermediate shafts, respectively, of transmission gearing. The gears are usually completely enclosed, in American practice. The shaft couplings are of a type allowing limited relative vertical motions between the decks. The wildcat locking head may be above the weather deck. In the simplest arrangement, a loose block key may be used. In these cases the brake drum will be
I
I
I
I
I I
I
, I
i
Schematic diagram of an electrically driven vertical windlass
above deck also. Chain troughs are then desirable between the inboard bolster of the hawsepipes and wildcats, if the span of chain is of appreciable length. The span may be divided by a common tongue-type chain stopper, which is also helpful as a guide to minimize vertical vibration of the span and slapping of the chain on the deck when the windlass is in use. A preferred design locates the brake drum and locking head below the weather deck where they are protected from the weather. This arrangement also permits the wildcat to be located as closely as possible to the weather deck. For handling warping lines, a capstan head may be keyed to the main shaft, above the wildcat. However, unless there is a speed change in the gearing of about 4 to 1, the rate of hauling light line will be unsatisfactorily slow. A preferred arrangement is one in which a capstan is located adjacent to the wildcat and is driven from the windlass gear train by a separate shaft revolving about 4 times as fast as the wildcat; the equipment for such an arrangement is illustrated by Fig. 16. As a simplification, the capstan head is keyed fast and is allowed to rotate or idle when the windlass is used to h a d chain. The vertical windlass affords extreme flexibility in the development of anchor handling and mooring arrangements, from the simplest single unit to as many independent systems as there are anchors installed. Some arrangements feature one power unit driving a single variable-stroke pump. The hydraulic system includes selector valves enabling the single pump to drive either of two hydraulic motors serving two wildcat/capstan systems. For utmost reliability, two power plants are installed and electrical, mechanical, and hydraulic cross connections or their combinations, depending on the basic features of the system, are included in order that both anchors p a y be recovered despite a derangement in one system; see Fig. 16.
UNIT
'
Fig. 16
Vertical electmhydraulic windlass
c. Special Design Considerations. The practice recommended for the design and testing of anchor windlasses is outlined in refere&% [12]. There are, however, some detailed design considerations which warrant special emphasis as they can have a large impact on the performance of a windlass. The links of stud l i d anchor chains are almost invariably made with an inside length equal to four times the nominal chain size (i.e., four times "the wire diameter"). Since 10 links will wrap around a &whelp wildcat, the wildcat circumference in inches = 10 X 4 X the nominal
582
HULL MACHINERY
MARINE EClGlNEERlNG
chain size. From this the mean pitch radius of the wildcat is readily found. Detailed information regarding the design of wildcats is given by one of the Navy Department's standard plans [13]. Wildcats should be made of a reasonably hard grade of cast steel. The fitting of wildcats and chains is important. Usually, the final dimensions of a chain are attained as a result of stretching in a proof test [9]. A new chain may be within the minus allowance of the tolerance on length (measured over six links in American practice) in order to allow for stretching and wear in service. However, such a chain will not run properly on a wildcat made to fit the nominal chain size. The "pitch diameter" of the wildcat is a function of the depth of the pockets between the whelps, in which alternate links lie substantially flat. If the links must be hauled off by the chain stripper when hoisting anchor, or if when paying out under power the links ride higher and higher on succeedingwhelps until one crashes over the top of a whelp, the pitch diameter should be reduced by chipping the pockets deeper. With a proper fit, but one sprocket tooth or set of whelps holds a t any one time. This subject is discussed in detail in references [14] and [15] and a recommended procedure is given for checking the wildcat and the chain with which it is to run. The number of links in a full circumference is twice the number of whelps (10 links for a 5-sided wildcat, for example). To check the fit of a chain on a wildcat, a length of chain containing one link less than that which will fit in a full circumference is lashed snugly around the circumference of the wildcat (a longer length of chain may be used if the standing part is hauled aside). The gap for the missing link is then measured. If the length is excessive, the pockets must be chipped deeper. Measured between the outside of the link ends, the gap should not exceed 2 chain sizes in length. Calipered inside the ends, the measured length should not exceed 4 chain sizes. The chain deck pipe, which leads to the chain locker, for a horizontal windlass should be located well under the wildcat. The vertical centerline of the deck pipe should project upward through the axis of the wildcat, or, preferably, slightly forward of the axis. The chain pipe lip must be flared to assure that the chain will be hauled into the locker by gravity. This will aid in arresting the rising of chain from the locker and will minimize "jumping" of links over the wildcat, due to kinetic energy, if the brake is applied suddenly when an anchor is being dropped. Chain stoppers are normally furnished by windlass and chain manufacturers. The common tongue type is favored in American merchant practice as a means of securing the chain when riding at anchor. Pelican hooks, modified turnbuckles or "devils' claws," and turnbuckles on lengths of chain passed through the anchor shackles at the inboard ends of hawsepipes are used as stowing stoppers. These should align as closely as possible with the run of chain. The wildcat brake should not be used to hold the anchor in the stowed position, because if the brake should slip, the anchor will back out and pound in a seaway.
The pelican hook is often used on naval vessels as a riding and stowing stopper. It is best suited for use with a vertical windlass, i.e., with the chain close to the deck. On a large horizontal windlass the chain may rise 5 feet or more from the inboard bolster of the hawsepipe to the wildcat tangent point. The pelican hooks in sizes large enough for application with such a windlass are very heavy and difficult to manage. Their height above deck also creates a hazard to the operator who must release them with a maul. For a detailed discussion of anchor, mooring, and towing arrangements see reference [I11. The wildcat band brake design and the selection and treatment of brake lining material are critical in windlasses which must be subjected to the free drop tests required by the regulatory bodies. The Maritime Administration stipulates a single drop from 45 to 60 fathoms, '(under control of the brake" [7]. These tests should be conducted in a clear depth of a t least 65 fathoms and preferably no greater. The falling anchor is permitted to be under the control of the brake to the extent that an operator senses his chances of arresting the run of chain and can do so by applying very little additional braking effort. No mishaps have been reported from Maritime Administration drop tests. The specifications for naval windlasses require that the hand brake performance be demonstrated during four successive average drops of 15 fathoms each, stopping with 15, 30,45, and 60 fathoms of chain out, respectively [161. Anchor windlass brake tests must be conducted with the utmost respect for the magnitudes of the masses, velocities, and forces involved. These tests can be extremely hazardous, especially in the event of a runaway. Test observers should stand well clear of the windlass with only those responsible for operating the brake and recording data allowed to be near. A satisfactory anchor windlass brake must. stop the anchor and chain within a period of about two secorids after the brake is set. Due to the short time available for the brake to absorb most of the kinetic energy possessed by the anchor and chain, the surface of the brake lining usually reaches a high temperature. Elevated temperatures can result in a degradation of the brake lining frictional characteristics such that the brake is no longer capable of arresting the continuing run of the anchor and chain. Lf the situation should get out of control, the problems compound; that is, as the anchor continues to run after the brake is set, the quantity of energy that the brake must absorb increases and more heat is generated. For this reason, it is essential that the brake promptly arrest the anchor and chain after being set. Anchor windlass brakes are usually of the lined band type. For maximum effectiveness, these brakes should wrap around the drum as near 360 degrees as possible. The features of a typical band type of brake are shown by Fig. 17; auxiliary power-assist mechanisms for setting the brake have also been used to advantage on very large windlasses. If molded linings are used, they are best secured to the band in a manner permitting lateral ex-
Table 3
I
-1-
POWER SCREW
Anchor Windlass Brake Calculations
Size of anchor chain, in.. . . . . . . . . . Outside length of one link, in.. . . . . Pitch of link = G - 2C, in.. . . . . . Number whelps on wildcat.. . . . . . Wildcat pitch diameter = 2ap/rl ip. Brake drum diameter, in.. . . . . . . . . Brake band width. in.. . . . . . . . . . . . Brake band thickneps, anchor end, ~ n .. .. . . . . . . . . . ,,. . . . . . . . . . . . . . Brake band thickness, slack end, in. Handwheel diameter, in.. . . . . . . . . . Screw diameter. in.. . . . . . . . . . . . . . Lead of screw, in.. . . . . . . . . . . . . . . . Screw efficiency. . . . . . . . . . . . . . . . . Bell crank ratio. . . . . . . . . . C . . . . Gear ratio. . . . . . . . . . . . . . . . . . . . . . Gear efficienc . . . . . . .. . Uechanical a & & k k e ' = a ~ ~ ~11,. ~ Fat.hom of -- chain ol;t,. ...~ > - fat,h.. - ~. . ~ .~. Weight of chain, lb. . . . . . . . . . . . . . Weight of anchor, lb. . . . . . . . . . . . . Tot,d weight = WI Wa,Ib. . . . . Hawsepipe efficiencyn.. . ... . . . . . . . Buoyancy factor.. . . . . . . . . . . . . . . . Static load at wildcat = e3bWa,lb. . Angle of brake wrap around drum, deg. . . . . . . . . . . .I . . . . . . . . . . . . . Angle of wrap = lr61/180, red. . . . . Brake lining coeff. of frictionb.. . . . Ratio factor = e w . . . . . . . . . . . . . . . Assumed velocity of fall, fps. . . . . . Specified stopping distance, ft. . . . . Deceleration force = e3W3V1/2h lb Force at wildcat = f i W4, fb!. . Force at brake drum = Fsd/D, lb. . Slack end r l l ->/(K - 1), lb. . . Anchor en pull - F3 Fa, l b . . . ? Handwheel effort = F4/2Lfelest lb. . Anchor end band stress = Fs/wtl, ps1 . . . . . . . . . . . . . . . . . . . . . . . . . . . Slack end band stress = F4/wts, psi. Mean band stress = (81 S2)/2, ps1 . . . . . . . . . . . . . . . . . . . . . . . . . . . Band modulus of elasticit , psi . . . Band stretch = ~ 3 ~ a / 2 l i n . .. . Handwheel turns to set braked = yBG1/L. . . . . . . . . . . . . . . . . . . . . . . Maximum 2Fs/wD, brake psi.. . .band . . . . .pressurea . . . . . . .. = .. ~-
~~
~
-
:'.
~ ~ - - -
Symbol
C G
P
a d D w ts
+
Fig. 17 Schematic of a band type of anchor windlass brake
1)
pansion as the binder breaks down and swells with heat. Molded linings are most effective under the conditions of the drop test if deeply scarified or if vented with lateral grooves about 3i6 in. wide by %6 in. deep, pitched about 23 in. around the circumference. A new brake should be "run in" by operating the windlass under power with the chain held clear of the wildcat. The band should be examined periodically and the operation continued until the area of the brake lining in contact with the drum surface is at least 75 percent of the total area of the brake lining. The drum and lining surfaces should be free of exudation; if need be, they should be cleaned with a solvent. Care should be taken that overheating does not occur during the run-in operation. These precautions should be accomplished at the shipyard rather than at the windlass manufacturer's plait. The new linings will then be in the best condition possible, arid preservatives and rust preventive media will be eliminated as prospective reasons for a reduction of the band lining coefficient of friction. The design of an anchor windlass brake is heavily influenced b y t h e experience gained with previous designs, as may be noted from the typical anchor windlass brake calculations shown by Table 3. The brake lining coefficient of friction used in the design calculations is generally somewhat less than the value suggested by the brake lining manufacturer and may a t first appear to be unreasonably low. However, the lining coefficient of friction suggested by the manufacturer is usually based on ideal laboratory conditions which hardly simulate anchor windlass brake service. Special instrumentation was provided during the sea trials for the aircraft carrier USS America (CVA 66) in order to confirm that the design criteria employed in the design of the-anchor windlass brake were adequate. Before the sea trial, every reasonable precaution was taken
+ +
1.0 0.5 H 20 2.0 do L 0.25 0.165 2 1 GI e1 0.92 M 503 f- - . 60 19,840 W Z 12,693 Wa 32,533 0.80 0.87 W4 22,600
tl
-
Type of Ship C4 CVA 66 2jis 4% 1498 28.5 19.0 9% 5 31 60.48 48 90 2 X 10 10 1.0 0.75 26 2.5 0.5 0.238 2.76 6 0.903 2705 60 80,000 60,000 140,000 0.80 0.87 97,440
"7
+
:
Notes for Table 3
An optimistically high hawsepipe efficiency should be used in this calculation. b. In general the brake lining coefficient of friction quoted by manufactur6rs should not be used, as that is the value ?btained?n a laboratory. Anchor windlass tests show that a coeffiaent of fr!ction of 0.225 and 0.30 can be expected with molded and woven lmings, respectively. c The braking force required at the handwheel should be approximately 100 lb or less. -. d Efforts to red~uce the ha&wheel effort by inaxeasing the mechanical advant.age may result in an excessive amount of elasticity in the system. The maximum pressure on the brake lining should be as recommended by the manufacturer and proven by experience. f In order to prompt1 initiate braking action, a hydraulic-assist mechanism was installeion the CVA 66. The mechanism wae des i r e d s~uchthat the brake was set hydraulically when the handw eel was turned about one-half turn in the direction to set the brake. Turning the brake handwheel also set and secured the brake mechanically as in conventional practice. ,,&+
HULL MACHINERY
MARINE ENGINEERING
COEFFICIENT OF FRICTION '
1
COEFFICIENT OF FRICTIW
d ,/CHAIN
VELOCITY
200-
160-
EN0 PULL
ANCHOR END PULL 0
0 '1.
U) W
100-
I
0 J 0
PULL
i
J
2
so-
TIME, SECONOS
(a) Drop to 15 fathoms 0TIME, .SECONDS
(c) 3 0 to 4 5 fathom drop OF FRICTION Z
Y
P
1
b0.4-
L L
0
COEFFICIENT OF FRICTION CHAIN VELOCITY
20.
0
; L
10.
0.2,
I*
t
Y
14. ANCHOR END PULL
SLACKENDPULL
10 TINE, SECONDS
(b) 15 to 3 0 fathom drop Fig. 18
USS America, CVA 66, anchor windlass brake test data
to ensure that the brake lining material was properly prepared and "run in." Figure 18 shows the results obtained during the CVA 66 tests, and the relatively low brake lining coefficient of friction anticipated (see Table 3) is seen to be confirmed.
A chain counter is a very useful tool that may be installed on a windlass. A chain counter provides a mechanical readout at the windlass of the number of feet or fathoms of chain that have been payed out, and provides an electrical readout in the wheelhouse. The officer
TIME, SECOND8
(d) 4 5 to 6 0 fathom drop Fig. 18 (continued)
586
HULL MA
MARINE EN(
on dutv then knows the amount of chain in use without sending a man to thc forecastle to check the markings on the chain. The indication of the amount oT chain out is also helpful (when the depth of water is known) in paying out enough chain to ensure that the anchor will hold. A mechanism may also be installed to provide a means for releasing the anchor or anchors from the wheelhouse. To set the windlass up for this operation, the hand brake on the wildcat is tightened, the wildcat clutch disengaged, and any devices used to secure the chain at sea, such as the devil's claw or tongue-type stopper, are removed from the chain. A hydraulic cylinder, powered by an accumulator, is arranged mechanically to override the brake screw mechanism. Then. when a solenoid valve is energized from the wheelhouse, the wildcat brake is released and the anchor will drop. Speed governing is built into the hydraulic system to limit the rate of fall of the anchor. This type of arrangement has been successfully used between a tug and an unmanned barge. I n this instance the control was accomplished by a radio signal, with the small amount of electric power required on the barge supplied by batteries. Remote control of the anchor would-be particularly advantageous in a long river passage where it is customary to have at least one man standing by at the windlass in case an emergency drop is necessary. d. Windlass Power Units. Since the late 1960's., verv " few steam-driven windlasses have been manufactured. However, steam-driven windlasses were common before then, particularly on tankers which carried inflammable cargoes. Steam-driven windlasses are usually of the horizontal type with all of the components located above deck; such a windlass would look much like the one illustrated by Fig. 13 if a steam engine were substituted for the electric motor. The steam engine is commonly a horizontal reversible type with two cylinders. Steamdriven windlasses are designed to operate with a steam pressure of about 100 psig at the throttle. Steam-driven anchor windlasses are inherently rugged, simple, and reliable; however, these considerations are seldom criteria for selection as other types of windlasses can be designed to be equally dependable. The major advantage associated with steam-driven windlasses is that they entail no fire hazards when used on tankers that carry inflammable cargoes. On the other hand, the long runs of piping from the engine room pose two problems: one of actually getting steam to the windlass, and the other of maintaining insulation on the pipes when they are run above the weather deck, as is the usual arrangement. These two problems are, obviously, closely related. In cold weather, if the windlass is steam-powered, it is usually necessary to turn the steam on well before the windlass will be needed so that steam and not condensate gets to the unit when it is needed for anchor handling. The two commonly used powering systems for windlasses are direct-connected electric motors and electrohydraulic systems. Electrohydraulic systems permit complete control over the hoisting speed and also provide
protection (by relief-valve action) against shock loadings in the t~.ansmissionshafting and gearing in the event that the anchor is inadvertently housed too abruptly. When an electric motor is directly connected to the windlass, it may be either a squirrel-cage or wound-rotor alternating-current motor or a direct-current motor. A d-c motor provides sufficient speed control to house the anchors safely. If a squirrel-cage motor is used, it should be of either the two- or three-speed type with the slowest speed usually one quarter of the full-load speed and slow enough to house the anchor satisfactorily. Even if a multi-speed a-c motor is used, the anchor should be driven through a slip-type clutch coupling so as to limit inertial loadings in the event that the anchor is housed too abruptly. If variable speeds are necessary or desired, then either a wound-rotor a-c motor or a d-c motor may be used. If a wound-rotor motor is used, it should be separately ventilated with the ventilation air taken from below deck so that water will not enter the air intake. The discharge air may be directed back to a protected space, or the air may be discharged directly on the open deck. In the latter case, the discharge outlet should be protected by a solenoid valve arranged so that the solenoid will open the air duct when the blower motor is energized. Interlocks must be provided so that the main motor cannot be energized unless both the blower motor and resistor fan motor are running; and additional interlocks should be provided to shut the entire system down in the event that the temperature of either the main motor discharge air or the resistor bank discharge air exceeds a predetermined safe limit,. If the variable speed for a motor-driven windlass is to be obtained from a d-c motor, the direct current may be provided by the adjustable voltage output from an a-c/ d-c motor-generator set. This drive is advantageous if the windlass is part of a combination unit, the other part of which is a constant-tension mooring winch. Windlass speeds and loads are controlled as described later under Cargo Winches and Constant-Tension R'Iooring Winches. In the electrohydraulic windlass, the pump, or A-end, is usually located below deck and driven by an a-c motor; and the hydraulic motor, or B-end, is mounted on the input shaft of the windlass gear reducer. When the windlass is of the horizontal-shaft type, the B-end ismounted in the weather. When the windlass is of the verticalshaft type, the B-end is mounted below deck out of the weather. Some vertical-shaft windlasses, partlicularly those on naval vessels, have two completely separate power plants, one for each wildcat or wildcat and capstan combination. These are then arranged so that, in the event of a casualty to one power unit, the other unit may be engaged (usually hydraulically) so as to operate both vertical shafts; see Fig. 16. A more common arrangement 011 merchant ships is one with a single, doubleended, electric motor driving two pumps with each pump discharging to a hydraulic motor. In each of the hydraulic arrangements described above, the hydraulic transmission would consist of a positivedisplacement, reversible-flow; variable-stroke pump
piped in a closed circuit to a fixed-stroke hydraulic motor. The first pinion in the gear reduction should be coupled to, rather than mounted on, the B-end output shaft. The hydraulic circuit should include an auxiliary, positive-displacement, replenishing pump. The pumping unit bedplate is usually built as a storage tank, and it should be large enough to contain 110 percent of the oil in the system so that all oil may be drained to the tank for servicing or maintenance. Some windlasses are provided with a horsepower-limiting device which is responsive to the pressure in the system. The horsepower limiter should be designed so that when the pressure reaches a predetermined value, the pump stroke is reduced with increasing pressures so that the electric motor horsepower remains constant. As the pressure in the system reduces (i.e., the anchor chain is hauled in), the horsepower limiter will return the pump stroke to the setting called for by the operator. Each side of the hydraulic circuit should be provided with a relief valve of the "cross-blow" type, returning the oil to the suction side of the pump. Alternatively, in some cases the discharge from the relief valves may be led to the sump tank. Direct electric-driven windlasses should be provided with an electric brake on the motor shaft. ,Hydraulic windlasses may be provided with either an electric brake or with a hydraulic brake which should be mounted on the B-end. Electric brakes should set upon loss of electric power, and hydraulic brakes should set upon loss of either electric power or hydraulic pressure. Speed and directional control for direct electricrdriven units should be effected by a master switch located aft and at a safe distance from the windlass. The master switch is usually equipped with a vertical handle, as described in Section 1. Stroke control for a hydraulic windlass may be effected either manually or by servo control. The control wheel stand should be raised to a convenient height above the deck and should be located aft and at a,safe distance from the windlass. If manual control is used, the shaft should be provided with a deck stuffing box through which it is led to the A-end stroking mechanism. The run of control shafting should be made with great care in order to avoid excessive friction, lost motion, and derangement that may be caused by deck deflection. Universal joints and slip couplings should be used to assist in eliminating the deleterious effects of misalignment. The handwheel for the hydraulic stroking device is usually arranged in a horizontal plane. An auxiliary handwheel for the stroking device should be located adjacent to the A-end to assist in servicing or warming up the unit. The stroke-control mechanism should be provided with a spring detent for the neutral position of the A-end stroking spiadle. Limit switches should be arranged to prevent the pump from being started if the pump and servo control are not in the neutral position. This is necessary to ensure that the anchor does not start to move when the pump is energized. Electric master switches should be provided with de-
Table 4 Power Calculations for an Electrohydraulic Windlass Type of Ship.. . . . . . . . . . . . . . . . . . . . . . . . . . . Number of ar~ehorshoisted. . . . . . . . . . . . . . . Anchor depth at beginning of hoist, fath.. . . Anchor weight, lb . . . . . . . . . . . . . . . . . . . . . . . . Anchor chain size, in.. . :. . . . . . . . . . . . . . . Chain weight (each anchor), lb. . . . . . . . . . . . Buoyancy factor.. . . . . . . . . . . . . . . . . . . . . . . . Weight per wildcat = b(W1 WZ),l b . . . . . Hawsep~peefficiedcya . . . .' . . . . . . . . . . . . . . . . Pull at each wildcat = WJ/e?, lb . . . . . . . . . . Outside length of one cham link, in.. . . . . . . . Pitch of links = G - 2C, in.. . . . . . . . . . . . . . Number of whelps on wildcat. . . . . . . . . . . . . Wildcat pitch radius = ap/?r, in.. . . . . . . . . . Torque at each wildcat = Plr, in.-lb. . . . . . . Electric motor speed, rpm. . . . . . . . . . . . . . . . Hydraulic pump speed, rpm.. . . . . . . . . . . . . . Hydraulic motor speed, rpm. . . . . . . . . . . . . . Gearing ratio: first reduction.. . . . . . . . . . . . second reduction. . . . . . . . . . third reduction.. . . . . . . . . . . Chain hoisting speed = 2?rrNa/l2R1RzRa,fpm Specified hoisting speed, fpm.. . . . . . . . . . . . . Hydraulic ump and motor efficiency. . . . . . Gearing efkciency: first reduction. . . . . . . . . second reduction. . . . . . third reduction.. . . . . . . Efficiency of wildcat. . . . . . . . . . . . . . . . . . . . . Torque per hydraulic motor = Tl/R1R&aezeae4eal in.-lb. . . . . . . . . . . . . . . . . Hydraulic pressure required,b psi. . . . . . . . . . . Total electric motor hp requiredc = nPIS/ 33,000 ele2ese4e5,hp. . . . . . . . . . . . . . . . . . . . . Electric motor hp provided,d hp. . . . .. . . . . . Capstan diameter, in.. . . . . . . . . . . . . . . . . . . . Capstan rope diameter, in.. . . . . . . ., . . . . . . .
,
+
Gearing, efficiency B-end = ezea. to capstan: . . . . . . . . . . . . . . . . . . .. . reduction = RlRz.. . . . . . . . . . . . . . . . . . . . Capstan rpm = Na/R4, rpm. . . . . . . . . . . . . . d)N4/12, fpm. . . . . . . Rope speed = r ( D Permissible rope pull = 33,000 ele$lz/S,, lb. .
+
Symbol n h W1 C WZ b Wa e. PI
G
P a r TI Nl NZ NJ Rl Rz Ra S Sl el eJ e4 e5 Tz
C4 2 30 12,693 0.87 19,673 0.60 32,788 14% 9% 5 15.5 508,200 1150 1150 1100 7.83 7.40 4.55 33.9 30 0.75 0.97 0.97 0.97 0.96
P,
2200 735
HI HZ D d
102.6 100 24 2.55
e6 R4 N4 S, F
0.94 57.9 19.0 132 17,625
Notes: a The hawsepipe efficiency used should be the lowest anticipated. b A check must also be made to confirm that there is sufficient capacity to hoist one anchor and the full scope of chain at no specified speed. In this arrangement, the electric motor has shafts on both ends, each of which drive a windlass through a Dension 60 hydraulic pump and motor. d Specifications often permit the electric motor to be overloaded 25 percent at the beginning of a hoist.
tents so that the operator can sense the speed position selected. Handwheels for controlling hydraulic windlasses should be equipped with a speed indicator, marked ''gll' ' ' ~ , '"~/4," ' and "Full," to each side of "Stop" for the neutral position. atio ion of %he handwheel in a cloc/kwise direction should start the windlass in the hoisting direction. The procedure follotved in determining the powering requirements of an electrohydraulic windlass with an attached capstan is shown by Table 4; the procedure for a direct-electric windlass is similar. Motors for direcb connected windlasses should have a 30-minute shorttime rating; whereas, for a hydraulically driven wind-
r,
4
1
II
588
MARINE ENGINEERING
HULL MACHINERY
jU SWIVEL FITTING FOR TOPPING LIFT
,--_ /
J(_SWIVEL
As shown by Fig. 19, the upper vang pendants, which are secured to the boom head, are used to swing the boom. Vang lines are required on both sides of the boom head. Lastly, the rigging arrangement must be capable of hoisting and lowering the load. The load would be secured to the cargo hook shown in Fig. 19 and would be hoisted and lowered by means of the cargo hauling part that goes to the cargo (hoist) winch drum. There are many variations of the basic rigging arrangement shown by Fig. 19. Reference [17] contains a discussion of some of the rigging arrangements which have been used and also relates the considerations which the naval architect must entertain when desighing cargo handling systems. Consequently, this chapter will concentrate on the design of the winches. Cargo winches are discussed in this section, and topping and vang winches are the subject of the following section. b. Mechanical. Ratings or duties of cargo winches have become standardized over the last 30 years. Most cargo winches are driven by 50-hp direct-current electric motors. Where hydraulic or steam winches have been installed, they also have approximately the same fullload speeds as the d-c motor-driven winches, but the light-line speeds may be different, being somewhat slower for the hydraulic units and somewhat faster for steam units. Cargo winches have different ratings at different line pulls; typical ratings for a 50-hp d-c motor-driven dlectric cargo winch are as follows:
PA0
Drum Line Pull, lb Drum Line Speed, fpm 0 500 3720 290 7450 220 14,500 105" 19,000 70' " "Two-speed" winch in low gear. Separate auxiliary drum.
OBLONG SWIVEL EYE
'
SHACKLE VANG PAD
Fig. 19
Nomenclature for cargo gear rigged for swinging or slewing
lass, the motor rating should be 30 minutes at 15 percent load followed by 30 minutes at full load. If a separate replenishing pump is provided, it should be rated for continuous duty. The main pump motor should be interlocked in such a way that it will not start until the replenishing pump is running. A start-stop pushbutton arrangement should be provided on or near the weather deck control station in addition to the buttons on the controller cabinet. A low-surface-temperature electric heater may be provided in the windlass sump tank to maintain the oil temperature at about 60 F in cold weather. This heater should be supplied from a circuit
independent of the main power feeders and arranged to be disconnected when the main pump motor is energized.
2.3 Cargo Winches a. General. The simple rigging arrangement shown in Fig. 19 has the capability of performing the three basic functions required of boom and winch cargo handling gear. One of the three functions required is the ability to top the boom, i.e., raise the boom head to the proper elevation. A second function required of the rigging arrangement is the ability to swing (or slew) the boom so as to control the transverse location of the boom head.
Most cargo winches are provided with a double gear reduction. The first reduction is frequently of the herringbone type while the second reduction is usually of the spur type. In some few instances, when the winches ate to be mounted on or adjacent to living quarters and very quiet winches are required, they are fitted with a single worm-gear reduction. Moat winches are of the singledrum type. If a gypsy head is desired on the winch, it is mounted on an extension of the drum shaft and the drum is then fitted with a mechanical brake and a clutch so that the drum may be secured when the gypsy is used. Occasionally a double-drum winch may be installed in place of two single-drum winches, thus saving the cost of the second motor and electric brake. On such a winch, both drums must be provided with clutches that are interlocked so that only one drum can be engaged at a time, but permitting both to be declutched at the same time. Also, each drum must be provided with a mechanical brake. A variation of the double-drum winch is a single-drum winch which drives an "auxiliary" or "derrick" drum through a third gear reduction. The shaft extension,
Fig. 2 0
50-hp, two-speed, d-c electric cargo winch without gypsy head
that might have had a gypsy head mounted on it, is fitted with a flexible coupling through which the third-reduction pinion is driven. The auxiliary drum will thus produce a line pull that is greater than that produced by the drum of the basic winch. Of course, the line speed on the auxiliary drum is reduced (see the foregoing winch rating data). Two of these winches are frequently used where a pair of burtoning booms are mounted on a pair of king posts and a heavy-lift boom is installed on the centerline. The hoist lines for the burtoning booms are led to the high-speed drums of the winches, and the hoist and topping lines for the heavy-li,ft=boomare led to the two auxiliary drums. Again, the clutches for the two drums must be arranged so that both can not be engaged at the same time. Still another type of cargo winch is the so-called "twospeed" winch such as shown by Fig. 20. In this winch, two pairs of second-reduction gears are provided, either of which may be clutched in to suit the load being handled. The output speeds of the alternative second reductions are usually in the ratio of about 2: 1. The output load and speed ratios may be varied to suit the rigging arrangement. The mechanical band brakes provided on cargo winches are intended to be holding devices rather than stopping devices. In an emergency, such as that caused by a power failure with a load suspended, the mechanical band brakes may be used to lower the load to the deck. For heavy-lift winches, the brake mechanism should be of the screw compressor type. On these brakes, a shaft with a handwheel has an acme thread which provides the force necessary to tighten the brake in a manner similar to the anchor windlass brake . a w n in Fig. 17. The brake bands are usually lined with a woven brake lhing similar to that used for the anchor windlass brake. In addition to the clutch interlocks mentioned above, it is advisable to interlock the brake and the drum clutch in such a way that the clutch cannot be moved out of engagement until the brake is set. As a substitute for the band brakes, it is frequently possible to use a simple locking device such as a bar or
MARiNE ENGINEERING
Fig. 21
Winch showing "double-diamond" type of spooling device
pin that can be inserted in a sleeve in the winch pedestal and pushed through to engage a hole in the flange of the winch drum. Drum clutches should be of the positive-engagement type; the most commonly used type is the jaw clutch. If the clutch is of the dry type, the jaws should be relieved a few degrees so as t o facilitate engagement. If the clutch is to be lubricated, then the jaw faces should be parallel, but a few degrees of backlash should be provided to facilitate the engagement. Brake and clutch linkages should be arranged and designed so that they will be extremely rigid in order that they will not be bent or otherwise deranged by being "forced." This is particularly true of clutch linkages. On many winches, there m a y b e no reason to declutch the drum for long periods of time. Due to disuse, it may then be necessary to use a large force to move the clutch and linkage, particularly when engaging the clutch. If the clutch linkage does not have sufficient stiffness, the position of theclutch shifter may not show the true position of the clutch jaws so that there may be only partial engagement of the clutch jaws. Insufficient stiffness in the clutch linkage has also been known to result in the clutch jaws working themselves completely out of engagement with the result that the load was dropped. (That is, even though the clutch handle was secured in the engaged position, the clutch linkage was so flexible that the clutch jaws separated and disengaged due t o internal clutch forces.) All clutches should be provided with a device that will secure the clutch handle in either the engaged or the disengaged position. The drums should be fitted with removable rope guards designed to prevent a slack rope from being wound over the flange and being wrapped around the drum shaft.
When a large amount of wire rope is to be stored on a winch drum, i t is advisable to use a spooling device. These devices ensure that the wire is distributed evenly across the drum and does not pile up in one place. Such a pileup may occur a t the center of the drum or against either flange depending on the lead of line from the heel block to the drum. The most elementary form of a spooling device is a grooved drum. The groove is a long spiral from end to end of the druni. However, a grooved drum will only control the storage of one layer of wire and only if the fleet angle is very small and if the wire is kept under tension continuously. Mechanical spooling devices may take any one of several forms, the most popular of which consists of a pair of rollers that are arranged with their axes a t right angles to the line lead and parallel to the drum flanges. The rollers are mounted on a trolley or carrier which is moved back and forth across the width of the drum to ensure that the wire is wound evenly across the full width. Translation of the carrier is provided by a doublediamond threaded shaft which is rotated by gearing that is powered from the drum shaft. A winch fitted with this type of spooling arrangement is shown in Fig. 21. The design of the gearing and the double-diamond must be such that the carrier moves one wire diameter for each turn of the drum. Another mechanical device is that known as the LeBus Fleet Angle Compensator, which relies on special drum grooving and a specially mounted fair lead sheave. There is additional information on this spooling problem in reference [17]. Openings should be provided under the-motors to permit inspection and maintenance of the underside of the motor and access to the studs that secure the field poles. c. Electrical. There are a number of ways of obtaining variable or multiple speeds when a winch is driven by an electric motor. One alternative is a single-speed squirrel-cage motor driving the winch through a variable &splacement hydraulic pump. Three- and four-speed squirrel-cage motors and wound-rotor slipring motors are also possibilities. The most popular type of drive on cargo ships is that which uses an a-c/d-c motor-generator set to produce adjustable-voltage direct current t o control the speed and direction of the winch. In the cases of the hydraulic and motor-generator set drives, the prime mover is started once during an entire cargo handling day; whereas in all other cases the prime mover must be started each time the winch is required to haul in or pay out rope. These frequent starts and stops will cause voltage fluctuations in the ship's system. The most popular system of control for cargo winches is the modified "Ward-LeonardJ' system, which employs an a-c/d-c motor-generator set to provide an adjustable voltage output to vary the speed of the d-c cargo winch motor. Usually a squirrel-cage a-c motor drives two d-c generators, each of which supplies a controlled voltage to one of a pair of motors driving two winches that are arranged for burtoning cargo. The motor-generator sets
HULL MACHINERY
Fig. 2 2
Vertical motor-generator set for cargo winches with rotor removed i
Fig. 23
Vertical motor-generator set to sene a pair of cargo winches
are located in deckhouses so that they may be of a dripproof construction. The control is usually of the solid- The lower part of the rotor in Fig. 22 is the a-c motor and state type and provides infinitely variable speeds both the two d-c generators are above. hoisting and lowering. The a-c motor should have a conA horigontal motor-generator set is illustrated by Fig. tinuous rating, but the generators and the d-c motors 24. The a-c motor in the center drives two d-c generamay have a 30-minute short time duty rating. The ch- tors, each of which powers one cargo winch motor. cuitry should provide safe lowering speeds if either a Multi-speed squirrel-cage motors with three or four brake failure or power failure, or both, occurs. The d-c speeds have been used very successfully for cargo winch motors may be shunt, stabilized-shunt, or compound drives. One such three-wed motm has 4, 8; and 28 wound, and of totally enclosed, nonventilated, water- poles; another uses 4, 8, and 36 poles; and% third has tight construction and should be provided with an auto- 2, 4, and 40 poles. I n each case, the high-speed (4 or 2 matic drainage fitting. pole) winding can handle loads up to about 3 tons drum Vertical motor-generator sets for cargo winches are pull a t about 300 fpm, and this is also the maximum shown in Figs. 22 and 23. The units are force ventilated light-hook speed. The second speed (8 or 4 pole) proby a fan on the rotor. Air is drawn in through the con- vides a line pull up t o about 6 tons a t about 150 fpm. troller door, passes through the base, and is discharged .An automatic step-back relay is provided so that if the vertically from the top of the motor-generator enclosure. line pull on the drum exceeds the nominal 3-ton rating,
592
MARINE ENGINEERING
Fig. 24
Horizontal motor-generator set
the motor will not accelerate to the high-speed winding even when the master switch is advanced to the highspeed position. The low-speed windings are used to land loads and to take up on a slack line. Although only three speeds are available, the transitions between speeds are quite smooth, the speed changes being absorbed by the inertia of the winch and the inherent elasticity of the entire rigging system. The four-speed motor offers 4, 8, 16, and 32-pole arrangements, with the 16-pole connection serving principally as a cushion between the slowest running speed and the third and full-load speeds. In each case, acceleration from one speed to the next is provided by time delays so that the motor will attain its speed before moving on to the next speed. An advantage that is obtained if squirrel-cage motors are used is that they are essentially trouble-free since there are no commutators, slip-rings, or brushes to be maintained. The control of the hoisting speed of wound-rotor slip ring motors is generally accomplished by inserting varying amounts of resistance in the rotor circuit. By inserting a large amount of resistance in the circuit, the starting torque of the motor is increased and the inrush (or starting) current is reduced. The smaller inrush current is beneficial in that there is less disruption of the ship's electric system each time a winch is started. As successive amounts of resistance are removed from the rotor circuit, the motor speed increases until, when all external resistance is removed, the motor will, in effect, be a squirrel-cage motor and will run at approximately synchronous speed. Although a rheostat could be used to provide the external resistance, it is more usual to remove the resistance in four or five steps, thus giving either five or six motor speeds. This method of control is essentially the. same as that used for the control of constant-potential
direct-current motor@supplied by a direct-current generator in the engine room. Both control systems are also similar in that the resistors produce large amounts of heat that must be~emovedfrom the winch equipment room by a forced-ventilation system. The control circuitry for the winch should be arranged so that the winch motor cannot be operated unless the ventilation fan is operating and also so that the entire winch system will shut down if the temperature in the equipment room rises above a safe limit. Special consideration must be given to the control of speed when lowering an overhauling load with a woundrotor motor. There are two common methods of doing this. One is by plugging or countertorque lowering, and the other is the unbalanced-stator method. In the countertorque method, although the master switch is moved in the lowering direction, relays sensitive to current caused by an overhauling load actually supply torque to the motor that opposes the load. Varying the amount of this counter torque changes the speed of lowering. When there is no load on the hook, the winch must be driven to pay out line. Since there is no load to produce current, the current-sensitive relays do not function and the motor drives the winch to pay out l i e at the maximum motor speed. In the unbalanced-stator method for controlling lowering speeds, unequal or unbalanced voltages are applied to the stator. Usually, an autotransformer is connected across one phase of the power supply. A series of taps from the transformer lead to contactors which furnish voltages both above and below normal. This results in applying two rotating fields to the motor, with one field rotating clockwise and the other counterclockwise. By changing the amount of unbalance, the relative strengths of the two fields change and the speeds are varied. There are several other methods of controlling both the hoisting and lowering speeds of wound-rotor motors, but these usually require elaborate circuitry and additional rotating equipment [18]. d. Hydraulic. Variable-delivery hydraulic pumps have been used to drive winches on a few cargo ships, and are also used on tankers where they replace steam winches which have their pipe insulation problems. These hydraulic units have been furnished in two ways: with the complete winch including electric motor, pump, and hf" draulic motor and brake on a common bedplate; or alternatively, with the electric motor and pump located in a deckhouse and connected by pipe to the hydraulic motor and brake mounted on the winch bedplate. With the latter arrangement, the electric motor usually drives two pumps, each of which serves one winch of a pair assigned to a pair of burtoning booms. With this arrangement, the capacity of the electric motor need only be about 150 percent of the capacity of the hydraulic motor to drive a single winch, since it is impossible to draw full load from both winches at the same time. Thus, the total installed electric horsepower is reduced. Because the motors and pumps are out of the weather, the motors can be dripproof.
I
593
HULL MACHINERY
a clean hydraulic system. Where the pumps are located in the deckhouse, the piping between the units must be fabricated and installed by the shipbuilders. Extreme care must be used to ensure the removal of dirt and other contaminants from the system by the use of a cleaning procedure such as is described in reference [I]. Special attention must also be paid to the provision of vent fittings at the high points in the hydraulic piping system for the removal of entrapped air.
Fig. 27
Worm-geared topping or vang winch fltted with disk-type electric brake
4
lowering the boom to the deck for servicing. In addition, topping winches are used to hoist the boom to the highest working position or, if the boom is to be stowed vertically, to the boom rest on the king post or crosstree. b. Mechanical. Topping and vaag winches are usually designed so that the lubrication and the fill-and-drain 2.4 Topping and Vang Winches connections will function properly whether the winches a. General. Topping and vang winches are used to are mounted horizontally or vertically. This allows the move the cargo boom vertically or laterally, respectively, shipbuilder latitude in arranging the winches so as to in order to position the boom head to handle the hook provide the best possible cargo handling arrangement. load (see Fig. 19). They are usually of simple mechanical The winches should be capable of handling the working and electrical arrangement, capable of positioning the and test loads without the aid of a drum ratchet and pawl boom when it is fully loaded, and are also capable of or other securing device. The drive arrangement may be
594
of either the spur gear (double or triple reduction) type as in Figs. 25 and 26 or of the worm-gear type as in Fig. 27. The 7.5-hp vang winch in Fig. 26 is designed such that the motor and brake are mounted within a vang post; only the drum is exposed to the weather. c. Electrical. Tcpping and vang winch motors are usually of the single-speed, reversible, squirrel-cage type, and are equipped with brakes. The motors should be rated for 30 minutes, short time, full-load duty. Master switches should be of the "spring-return-toOFF" type. Frequently the master switches for the topping and vang winches for a boom are arranged so that they are operated by a single lever called a "joy stick." As an example of the use of a joy stick, movement of the lever away from the operator will cause the topping winch to pay out wire and lower the boom; and movement of the lever to the right will cause the boom to swing to the right. A movement of the lever in any direction in between these two directions will both lower the boom and swing it to the right. Providing the control of the two winches in a single lever permits the operator to position the boom quickly and land the, load with great accuracy. 2.5
HULL M
MARINE EN(
1 DECK
Fig. 29
Capstans
a. General. There are three mechanical arrangements of capstans that are in general use. In one arrangement the motor, electric brake, gear reducer, and capstan head are mounted on a common bedplate on the weather deck. I n a second arrangement only the capstan head is mounted in the weather with the motor, electric brake, and gear reducer hung from the underside of the weather deck. I n a third arrangement the capstan head is on the weather deck with the motor, brake, and gear reducer on the deck below, as illustrated by Fig. 28. I n all three cases the master switch is located near the capstan head on the open deck, and the controller is located in a protected location such as a deckhouse or the capstan machinery space. The first arrangement mentioned has the advantage that the xomplete unit can be assembled by the capstan manufacturer for bolting in place by the shipbuilder. However, it has two disadvantages in that the motor and brake h u s t be of watertight construction and that the capstan head is elevated to an inconvenient height above the weather deck. The second arrangement has the advantage of having the motor and brake out of the weather so that they may be of dripproof construction. However, it introduces a deck penetration which must be made watertight, and the capstan head must be mounted by the shipbuilder. Neither of these problems is particularly difficult to handle; but the fact that the power unit is suspended from the overhead makes it more difficult to inspect and maintain. The third arrangement, which is common, entails a problem of alignment between the driving and driven units and requires the installation of a flexible coupling that can accommodate a small amount of misalignment
Warping winch with extended shaft
in Fig. 28 may be of either the roller type or of the sleeve type with bronze bushings. For merchant ships it is commonly specified that when the capstan is handling the specified load, the stresses should not exceed 40 per cent of the yield point of the materials. The capstan head, main shaft, bearings, and capstan base should be designed to withstand the breaking strength of the hawser applied tangentially a t midheight of the head without exceeding 75 percent of the yield point of the materials. c. Electrical. Capstan motors should be reversible constantand are usually of the two-speed (full and horsepower, squirrel-cage type. They should be rated for 30 minute short-time, full-load duty on either winding. A brake should be provided on the motor shaft. The motor and controls should provide adequate control of overhauling loads. Stepback protection from high to low speed should be provided so that, when retrieving a light line on the high-speed point, the motor speed will automatically step back to low speed if the rated horsepower is exceeded in high speed. Automatic return to high speed should also be provided should the line pull be reduced. Capstans are usually ,designed for line speeds of about 30 to 35 fpm in low speed since this is about as fast as a man can handle the line and keep it tight around the head for friction purposes. If the motor is full and quarter speed as suggested above, this will result in a light-line speed of from 120 to 140 fpm. 2.6 Warping Winches. A warping winch is typically used to warp a ship alongside a pier or to move a ship from one place to another, by means of hawsers, without other assistance. The warping head on a warping winch is similar to the head on a capstan, except that the warping head or heads are mounted on a horizontal shaft. I n some instances, the heads are mounted on extensions of the main shaft so that they may be at a considerable distance from the power unit, as in Fig. 29; if the shafts are extended in order to spread the heads apart, these extensions are provided with outboard bearing pedestals close to the heads. Occasionally a drum is installed, as in Fig. 30, for handling a wire rope for a stream anchor. If a drum is installed, it is provided with a clutch and a band brake of the screw compressor type so that the drum may
x),
n Fig. 28
DECK
Capstan w i h machinery on deck below
and variation in the vertical distance between decks which may be caused by either temperature changes or deck loadings. This type of capstan offers the advantage of having the power unit on the deck, rather than overhead, so that it is more readily serviced. b. Mechanical. The capstan head is usually of the smooth-barrel type (without whelps) and its configuration should approximate that given in reference [19]. The gear reducer generally consists of a worm-and-wheel reduction and a spur, helical, or herringbone reduction. It is preferred that the worm-and-wheel reduction (if used) precede the other reduction in order to take advantage of the slightly higher efficiency of the worm at the higher rubbing speed. All bearings in the reducer should be of the ball or roller type because of the necessiby of accurate alignment of the worm gearing. The deck bearings for the capstan arrangement shown
Fig. 30
Electrohydraulic mooring winch
be secured when the heads are being used. The electric motor, brake, and gear reducer are all mounted on a common bedplate. The warping heads should be of the smooth-barrel type (without whelps) and located a t a height that will ensure a suitable lead of line from the bulwark chocks. The gear reducer is usually of the double-reduction type with spur, helical, or herringbone g e a r i n ~but occasionally a single worm reduction is used. A stress analysis should be made for warping winches employing the limitations as described above for capstans; in addition, if a drum is provided, the unit should be capable of withstanding the stall torque of the motor without exceeding 75 percent of the yield point of the materials. The electrical equipment described for the capstans would be suitable for the warping winches. 2.7
Constant-Tension Mooring Winches
a. General. A constant-tension mooring winch is a device used to maintain a preset tension in a mooring line after a ship has been tied up a t a pier. The mooring lines holding the ship to the pier will maintain a constant tension even with changes in tide or ship's draft, without the necessity of manually adjusting the lines. When constant-tension mooring winches are used, two or more are installed forward and two or more are installed aft. As many as ten constant-tension mooring winches may be installed on ships such as&he largest$ankers. . Figure 31 illustrates a pair of constant-tension mooring winches that are rigged to lead the line from each over the same side of the ship through univer~alchocks. Alternatively, the lines may be taken to the other side of the ship. Occasionally each of two of the forward winches is combined with a wildcat to combine the windlassjconstanttension functions into a single unit. The constant-ten-
MARINE ENGINEERING
Fig. 31
Constant-tension mooring winches
sion winches are also arranged for manually hauling in and paying out the wire rope while the ship is being tied UP. b. Mechanical. The essential mechanical difference between a cargo winch and a constant-tension mooring winch is usually the interposition of a planetary gear set between the two usual gear reductions and the drum. In addition, a band clutch is provided with the clutch friction surface on the outer diameter of the ring gear on most constant-tension winches. When the band clutch is set, the ring gear is locked and the sun gear is then able to rotate the planetary gears, thereby rotating the spider which is directly connected to the winch drum. The band clutch is not primarily intended for "free spooling," which is a desirable capability if the mooring line has been secured to a bollard when there is a small amount of way on the ship. Instead, most control circuits contain a "drift" position which energizes only the brake, thus allowing the winch to overhaul, and thereby permits "free spooling" without the necessity of making manual adjustments at the winch. All bearings should be of the ball or roller type to minimize friction so that the tension-sensing mechanism is adequately sensitive to changes in line tension. A typical constant-tension winch has the ability to maintain, within a reasonable tolerance, a tension of from 8000 to 20,000 Ib. There are usually three or four intermediate settings between these two extremes. Tension is usually sensed by either a coil spring or a torsion bar. If the tension sensed exceeds the preset tolerance, the motor is energized to pay out or haul in line until the original tension has been reestablished. The band clutch is usually capable of holding from 125 to 150 percent of the maximum tension setting. For the winch cited above, the band clutch would be capable of resisting a pull of from 25,000 to 30,000 Ib. If a sudden surge were to occur, the clutch would slip thus preventing excessive forces from being developed within the winch. The tension-sensing mechanism and the band clutch and their linkages should be totally enclosed. Where
.
-
this is not possible, nonferrous or stainless steel parts should be provided. If a gypsy or warping head is required, it must bk arranged so that it will function with the drum declutched and secured. If the head is to be used for warping duty, then the head, shaft, bearings, etc., should be designed to withstand the breaking strength of the hawser applied tangentially at mid-length without exceeding 75 percent of the yield point of the materials. c. Electrical. Most constant-tension winches are powered by an adjustable-voltage d-c drive similar to that described under Cargo Winches. Other satisfactory drives include wound-rotor slip-ring motors and variable-delivery reversible-flow hydraulic transmissions, both of which are also described under Cargo Winches. Details of motors and generators are about the same as for cargo winches except that the motor should have a duty rating capable of sustaining continuous cyclical operation in a five-second total cycle. Thermal protection should be provided in the motor, and the circuitry should be arranged so as to shut down the motor and sound an alarm if the motor temperature exceeds allowable limits. One master switch should be provided for each winch at each side of the ship. Generally the two or three switches required on each side at the bow or stern are combined as duplex or triplex units and may be arranged for either bulkhead or pedestal mounting. The master switch to be used at any time may be selected by a selector switch mounted on each winch. Alternatively, each port and starboard pair may be arranged electrically so that if one is moved away from the OFF position, the other switch is disconnected. Each master switch should provide the following operating positions: fast heave, slow heave, automatic, off, drift, medium payout, and fast payout. When the master switch is set at the "automatic" position, the line pull will be dictated by the setting of the tension switch on the winch. d. Variations. There are three major variations or adaptations of the constant-tension mooring winch. These are: (1) the winch which has a warping head and can be used as a warping winch; (2) the winch that is combined with a wildcat so that it can handle a bower anchor and its chain; and (3) the winch that is designed to handle a synthetic mooring line on the drum. Constant-tension mooring winches are often fitted with "gypsy" heads of small diameter for handling miscellaneous lines. These are not practical for handling mooring lines of the usual 8 to 9 in., circumference. If it is desired that the winch be capable of warping the vessel, using such lines, heads approximating the sizes quoted in reference [19] should be fitted. These must be operable separately from the drum. A secondary shaft and gear reduction may also be required. The second variation is the constant-tension winch that, through an additional gear reduction, drives a wildcat that handles the bower anchor. This is an advantangeous arrangement since it is unlikely that it will be
HULL M
Fig. 33
Bow thruster
below deck, out of the weather, provided adequate fairleads are installed and a spooling device is mounted on the storage drum. 2.8 Fig. 32
I
I
I
1
Ship with bow L ~ S t e r
necessary to handle both the mooring line from the drum and the anchor chain at the same time. Some flexibility is permitted in the arrangement of winch leads and chain leads; one arrangement is such that the wildcat which handles the starboard anchor is on the port constanttension winch and vice versa. This is illustrated by Fig. 14. The possibility of locating the wildcats and the upper ends of the hawsepipes almost at will permits proper chain leads while still ensuring that the lead from the drum is adequate for all mooring requirements. The third variation mentioned above permits the use of a synthetic line as a constant-tension mooring line. There is considerable advantage to be gained from this arrangement since the line itself has much greater elas ticity than a steel line and therefore can absorb most minor adjustments by stretching or relaxing without the winch being energized. A new factor is introduced, however, in that synthetic lines under tension must not be stowed on a drum unless the drum is properly reinforced. In at least one known case, a synthetic line was stored overnight on a drum after being hauled in under "normal" (say 20,000 lb) line pull; the result was that the line exerted enough compressive force on the winch drum to crush it. In order to avoid such a situation, the winch drum may be arranged to act as a capstan, with only enough t u n s of line on the drum to ensure adequate friction. The line is then pulled off the drum and is wound on a second drum for storage purposes. This second drum may be located close to the winch or may be located
Special Thrust Devices
a. General. S~ecialthrust devices are ~rovidedon ships primarily to ifnprove their maneuverinicapabilities at zero, or substantially zero, ship speeds when the rudder is relatively ineffective. Theke are three general types of these special thrust devices. The most common type consists of a propeller that is installed in an athwartship tunnel; the propellers in these units may be of either the fixed pitch or reversible pitch type. Trainable thrusters, a second type, are commonly designed such that they can be lowered through the bottom of the ship and trained through 360 degrees so that thrust can be developed in any direction; however, there are other types of trainable thrusters. One type of trainable thruster is similar to a large outboard motor, and the active-rudder type is designed to be incorporated into the rudder. The activerudder type consists of a small shrouded propeller that is driven by a submersible electric motor built into the aft edge of the rudder. In a third type of thruster, a pump is qrranged to take suction from beneath the keel and discharge to either side, to develop thrust port or starboard as desired. There is a considerable amount of literature that deals with the subject of special thrust devices; however, the most comprehensive treatpent of thesubject is contained in-reference [20]. b. Tunnel Type. The tunnel type of thruster, which is widely known as a "bow thruster" when installed a t the bow of a ship, is illustrated by Figs. 32 and 33. "Stern thrusters," that is, similar devices installed at the stern of a ship, have been used but they are not common. The principal performance features of a typical series of bow thrusters are shown in Table 5. The performance
MARINE ENGINEERING
598
Table 5 Performance Characteristics of Typical Reversible-Pitch Bow Thrusters
Horsepower
Dheter
RPM
150
3'-7" 4'-3C 5'-5" 6'-7' 7'-11" 9'-2"
450 420
300
500
800
1200
1800
340 290 240 210
Thrust, Ib 4,500
7,900 13,200 20,400 30,200
44,100
HULL MACHINERY
20
Thrustlhp
30
26 26
16
25.5 25 24.5
0 0
el2
.I.
cn
of any particular unit may vary from a comparable one shown in Table 5. In general, an improvement in efficiency is obtained with a propulsor having a larger diameter and lower rpm. Bow thrusters are often designed such that the thrust developed is both variable and reversible. This is generally accomplished by using a constant-speed electric motor to drive a controllable- and reversible-pitch propeller. The unit is started with the propeller set at zero pitch; then, as the need arises, the pitch is adjusted so as to provide the desired thrust to either port or starboard. For the smaller bow thrusters, however, a fixedpitch propeller is sometimes used with a variable- and reversible-speed driver. When locating the tunnel in the ship, it is desirable to have the tunnel well forward so as to obtain the maximum turning moment from the thrust developed. The depth of the tunnel must also be considered, particularly "or vessels which will operate at light draft, because the thrust developed breaks down at shallow submergence depths. A test was conducted at the Philadelphia Naval Shipyard to investigate the effect of submergence depth on the thrust developed. The bow thruster tested was powered by a vertical 800-hp electric motor which drove a reversible-pitch propeller through right-angle bevel gears. The thruster tunnel diameter was 6 f t 7 in. The ship draft was varied and the bollard pull was measured when thrusting to port and to starboard. The wind was steady at only a few knots, and there was ample clearance both below the keel and to adjacent piers. Figure 34 shows the data obtained when thrusting to starboard; values when thrusting to port were similar. I t may be noted from Fig. 34 that when the top of the tunnel is submerged less than about 1.5 f t , there is a marked loss of thrust. The design of the junction of the tunnel and hull is another factor which "requires study. The tunnel openings definitely affect the resistance of the hull; however, the effect is difficult to accurately quantify because it is of a relatively small magnitude. With a well designed tunnel arrangement, it would be reasonable to expect an increase in the ship's resistance of about 1 percent. When it is desirable to minimize the resistance added by the tunnel, flow studies should be used as means of engineering the shape of the tunnel fairing and also the orientation of protective bars, if used. Protective bars are usually mounted along the flow lines in the area of the tunnel openings. Usually they
-
m A
6
z
I
4
0
Fig. 34
3
1 0 -1 2 8UBUERGENCE OF TOP OF TUISmL, FT
Effect of tunnel submergence on thrust developed by a bow fhruster
have a slightly downward slope when going aft, as may be seen from Figs. 32 and 33. Protective bars of several different designs have been used. In some designs several of the bars are bolted in place so as to be portable and permit access to the mechanism and also to permit a removable blade to be unshipped. However, this practice has not been uniformly satisfactory because the bars that are bolted on have a tendency to come adrift. Consequently, a preferable design is one with the bars simply welded in place. If access is required at a later date, the bars are burned off and rewelded; it is often necessary to burn off the bars to accomplish major work irrespective of the design due to rigging problems. Access problems are alleviated with some designs by the provision of a propeller blade removal hatch in the tunnel immediately above the propeller. This hatch is large enough for a man to get through, and no access is necessary fiom the tunnel openings. I n this case, to remove a propeller blade, the ship is ballasted until the tunnel is out of the water and the blade is lifted vertically up into the ship through the hatch. The tunnel is generally made of mild steel and is welded into the hull. To minimize tip leakage losses, the clearance between the blade tips and the tunnel should preferably be no more than 0.25 in. However, in practice it is difficult to maintain the propeller and the tunnel concentric. The diiculty involves the welding practices during installation at the shipyard. After installation, it is not uncommon to find the tunnel to be no longer circular, thereby requiring the blade tips to be ground off. For seawater service all bolts, studs, nuts, and other fastenings should be of monel. Propeller blades are usu-
I
i
I
ally of stainless steel or nickel aluminum bronze, with the propeller hub of bronze and the pod struts made of steel. The propeller is driven through a right-angle gear drive (usually of the spiral-bevel type) that is contained within the pod assembly. The pinion shaft or inputghaft of the right-angle drive extends out of the pod and through the tunnel assembly. The arrangement can be designed such that the pinion shaft penetrates the tunnel assembly at any angle desired; however, the shaft is normally either vertical or horizontal. With a horizontal arrangement, the shaft normally goes directly to the prime mover without involving another right-angle drive. In the case of a vertical drive, the prime mover can be located many decks above. In a common arrangement a vertical electric motor drives the thruster input shaft through auxiliary shafting. If the prime mover is a type which cannot be oriented vertically (e.g., a diesel engine), it is then necessary to use a second rightangle gearbox. The second right-angle drive preferably should be of the spiral-bevel type supported by oil-lubricated, heavy-duty, antifriction bearings. Flexible couplings should be provided between the prime mover and the shafting. The hydraulic power unit is often mounted at least 10 ft above the load waterline so as to avoid the need for a separate gravity tank. This head pressure is necessary to ensure that an adequate pressure is maintained on the oil seals in the propeller hub so that seawater will not enter the pod if the seals should leak slightly. The unit should have a motor-driven pump mounted on a reservoir, complete with necessary piping, suction filter, pressure gages, relief valves, etc. The reservoir should have a capacity in gallons of at least 2.5 times the capacity of the pump in gpm and should be fitted with direct-reading level gages or sight glasses, fill, drain, and vent connections, and access covers large enough to permit the reservoir to be cleaned. The master control stand is located in the wheelhouse and is sometimes made a part of the wheelhouse console. When a reversible-pitch propeller is used, this stand should contain a single lever that will pneumatically, electrically, or electrohydraulically control the hydraulic blade positioning system so as to provide stepless pitch control from zero to maximum either to port or starboard. For fixed-pitch propeller installations, the propeller speed and direction of rotation should be controllable. Frequently, auxiliary control stands are located on each bridge wing and are connected electrically, hydraulically, or mechanically to the master control stand. In addition to the control stand or stands, a control and indicating panel should be provided in the wheelhouse. This panel should contain pushbuttons for starting and stopping the prime mover, the hydraulic system for pitch control, etc. It should also indicate the alignment of the system (e.g., propeller pitch indicator) and contain lights apd alarms for critical pressures and temperatures. If the bow thruster is to be used only for docking, the main motor need only have a one to two hour full-power rating. If, however, it will be used for extended periods
\
Fig. 35
Thrust device that can b q IoQered and rotated 360 d&reer to develop thrust in any direction
of time, e.g., as an assist in steering the ship at slow speeds during a long passage, then it should be rated for continuous full-load duty. All other motors for the unit should be rated for continuous full-load duty. The controller should incorporate low-voltage protection, and interlocks should be provided to ensure that the main motor cannot be started unless the pitch control system is in neutral, that lubricating oil and hydraulic control pressures are available, and that the thruster room ventilation fan is in operation (as applicable). c. Trainable Type. The trainable type of bow thruster is the most versatile of all since, when it is lowered'from the bottom of the ship, it may be trained in any direction to move the ship to port, starboard, ahead, or astern. It is rather unusable as a docking device, however, since it extends several feet below the keel and therefore is quite suscep5ile to d a ~ a g e . In this arrangement t?ik propeller and drive'shaft are usually lowered and raised as a unit, as generally indicated by Fig. 35. The mechanism which trains the propeller also moves up and down with the rest of the machinery. The propeller is mounted in a protective ring or Kort nozzle which is designed to improve the thrust from the propeller. A closure plate is welded to the bottom of the protective ring; therefore, when the
600
HULL MACHINERY
MARINE ENGINEERING
9 = the maximum design wave slope (in terms of a wave height h and a wave length A, 9 =
sin-' ~rh/A),deg
I
1
C FIN SHAFT
The stabilizing moment developed by the fin stabilizers
M, = NaL
-
WTER LEVEL
(6)
where
M, = fin stabilizing moment, ft-lb N = number of fins a = lever arm between the resultant lift force and the ship's center of gravity, ft L = lift force per fin, lb
( a ) CROSS SECTION OF A YON-ARTICULATE0
FIN STABILIZER (ALSO KWWN AS A SPAM TYPE OR SINQLY ALL-MOVABLE STABILIZER)
The lift force developed by each fin can be expressed in the conventional manner as: Fig. 36
Active fin stabilization sketch
where unit is retracted it must be aligned fore and aft. The closure plate provides a nominally smooth hull. One use for this type of thruster is for station keeping on a survey ship when it may be necessary for the ship to remain in position over a particular spot while bottom sainples or cores are being taken. One such installation has a fixed-pitch propeller driven by a constant-speed squirrel-cage motor through an eddy-current coupling. Varying the excitation produces a propeller speed of up to about 95 percent of the motor speed. The difference in speed between the motor and propeller shaft, or slip, results in the creation of heat which must be removed either by a ventilation system or by a seawater-cooled heat exchanger. Various interlocks are required for this type of unit so as to ensure that the space ventilation fan, the eddy-current ventilating fan or cooling water pump, etc., are operating before the unit is lowered and the main motor is started. An additional interlock is required to ensure that the unit is in exact fore-and-aft alignment before being housed so that the closure will fit into the hull opening. d. Jet Type. In this third type of thruster, a pump with large clearances takes suction from the bottom of the ship near the centerline and discharges it to either a port or starboard discharge pipe to develop lateral thrust. The flow may be directed by a "splitter" which can proportion the flow to provide varying amounts of flow to each side. Mechanically and electrically this is the simplest, but least efficient, of the three types described. 2.9
Active Fin Stabilizers
a. General. Active fin stabilizers are fin-type control surfaces which are usually located just above the turn of the bilge near amidships, port and starboard. In a seaway, the angle of attack of the fins is varied continuously by automatically controlled tilting gear so as to
produce heeling moments that reduce the ship's tendency to roll. A compilation of the characteristics of many active fin installations is contained in reference [21]. Active fin stabilizers require ship forward motion in order to develop lift, which increases as the ship speed squared. In practical terms, this speed dependency limits the application of active fin stabilizers to ship speeds above approximately 12 knots. Below that speed, the fin size tends to get too large, and other devices (e.g., anti-roll tanks) find application, as described in references [22] and [23]. b. Simplified Calculation of Stabilizing Moment and Number of Units. The determination of percentage of stabilization and the design of automatic controls require a complex analysis. However, the fin size and tiltinggear machinery characteristics and location can be determined adequately using a simplified approach. With the simplified approach, it is assumed that a regular beam seaway having a small surface wave slope and a wave period approximately equal to the ship's natural period can build up large roll angles. By designing fin stabilizers to counteract the wave slope heeling moments, it is possible to reduce the large roll angles. Figure 36 illustrates the concept for a ship with two symmetrical fins. The rolling moment induced by the seaway is expressed as follows: Mo = 2240 AGM sin 9
CL = a nondimensional lift coefficient (Typical values of CL would be approximately 1.1 for nonarticulated fins, and 1.4 for flapped fins. Figure 37 defines the two types of fin surfaces. Section 8 of reference [5] thoroughly discusses the procedure for computing CL.) p = water mass density, lb-seca/ft4 A = area of one fin, sq ft V = water speed used-in fin design, fps (using VK as the ship speed in knots, V = 1.69 VK)
f
#
(5) R
where
I
Mg = roll-induced moment, ft-lb A = ship displacement, tons
GM
= metacentric height (the distance between the
ship's center of gravity, G, and the ship's metacenter, M), f t
.
/
k FIN
LFLAP
I
t b ) CROSS SECTION OF A FUPPIO FIN STABILIZER (ALSO KNOWN AS AN hRTICULATE0 OR DOUBLY n u - MOVIBU (CT~BILIZER)
If the stabilizing moment computed from equation (6) Fig. 37 Common types of fin stabilizen is equal to or greater than the induced rolling moment calculated from equation (5), an effective stabilization system is considered to be achieved. The key to this simplified method is the somewhat arbitrary selection of the seaway wave slope capacity 9 and the associated ship 9 = 5 deg speed VK. Experience has shown that stabilizers should CL = 1.2 be designed for wave slopes of about 4 or 5 deg. Lower a = 56 percent of B values are reasonable for very large ships, and higher values for small ships, since small ships are more likely The values of CL can later be refined (e.g., Fig. 31 of reference [24] indicates that the CL for a flapped fin to be subjected to roll excitation in a given seaway. varies between 1.6 and 1.3 from speeds of 5 to 25 knots, The lowest ship speed, VK, at which a significant roll reduction is desired, is usually taken to be approximately and Sectjon 8 of reference [5] gives data for nonarticu50 percent to 80 percent of full-power speed. This im- lated control surfaces as a function of aspect ratio, plicitly recognizes that ships in a heavy seaway are sweep, and taper ratio.) The usual practice is to try N = 2 and N = 4, and then make a final decision as to the generally operated well below full power. number of fins. Next, combinations of nonarticulated I n order to obtain a first approximation of the fin area versus articulated and nq;stsetractabb versus retractable required, the following expression may be used: fins are investigated in order to establish the preferred arrangement. Some of the considerations involved are discussed below. c. Location and Type of Units. The simplest mawhere B is the ship's beam in feet and all other terms are chinery and most compact arrqgement are obtained as previously defined. Equation (8) can be derived by with nonretractable units. On the other hand, the smallcombining equations (5), (6), and (7) with the tentative est fin area is obtained with high-aspect-ratio fins (e.g., assumptions that fins with an outreach of at least twice the fore-and-aft
MARINE ENGINEERING
HULL MACHINERY
tially the control system measures the instantaneous heel coypling) ; or (2) accepting only limited roll angle residual ship roll motion, assumes a simplified equation stabilization in order to reduce high-frequency roll anguof roll, and orders fin angles at the proper phase relation- lar acceleration. ship to achieve fih lift and corrective righting moments. Most automatic fin controls are not designed to correct Robb [26] provides an introduction to the theory, Lewis steady heel from ship's loading or wind, but rather to [27] presents more advanced theory, and Bell [28], stabilize around the mean heeled angle. This is done to Chadwick 1221, and Conolly [25] provide specialized de- conserve the limited fin stabilizing moments for correctsign procedures. Thus, one set of simplified control ing the oscillatory roll variations. equations is given in reference [22] as: Another feature in many designs is the automatic angle or fin lift-limiter. This is a consequence of having 4 21,-4 4 = 7' - 7 fins which develop full rated lift several knots below fullShip: W2 W power ship speed. At the higher ship speeds, the lift and 4 4 (10) torque, which are proportional to the ship's speed Control: a = K14 K2 - K3 squared, might be excessive and not really needed. Lift W u2 is limited to safe values by a fin angle limiter which is Fin: 7 = K,U2a (11) either in the fin lift control circuit, if such is installed, or in the fin angle servo circuit, using as input the ship spied where the symbols have the following physical signifi- deduced from the main propulsion shaft rpm. cance :
+
+
+
Fig. 38
Nonretractable articulated fin unit, port side, looking forward
+
4
length) that are located in the ship a t positions of maximum beam and located at angles which slope downward going outboard such that the fins have maximum leverage. The solution for the least fin area practically always involves fins projecting beyond the hull block dimensions, thus requiring retractability or a compromise design with low-aspect-ratio fins. Additional considerations to be entertained when developingthe arrangements are: (1) in the rigged-out position, the fins should be a t least a few feet below the design heeled waterlines (so as to minimize cavitation), and not immediately upstream of important sea chests (such as for main condensers); and (2) the fin tilting gear should be in a compartment suitable for routine maintenance (e.g., an auxiliary machinery space), and preferably not adjoining prime sleeping quarters. Clearly these contain built-in codicts, and require judgement of the naval architect and marine engineer to reconcile. I n commercial practice, the compromises generally result in one or two retractable, articulated fins per side. The retractable feature permits locating the fins in the ship at positions of maximum beam, where there is favorable leverage. The articulated (doubly all-movable) fins require about 30 percent less planform area than nonarticulated fins and are therefore easier to retract. On destroyer escorts, the U. S. Navy has used nonretractable units, some with articulated and some with nonarticulated fins. In this application, the hull form sections are much more rounded than cargo ship hulls, and also sonar domes extend several feet below the keel. Each fin axis is thus sloped well downward going outboard, and the fins do not protrude beyond the maximum beam or the sonar dome navigational draft. In this type, illustrated by Fig. 38, the tilting gear, fin, and hull insert plate can be readily furnished as one unit with a factorysealed hydraulic system. The fin unit shown is installed with the fin sloped 55 deg downward going outboard, and it has a span of 8 ft and chord of 4 f t (main body 36 in.,
.
tail flap 12 in.). Each unit has a total weight of 19,000 lb. The hydraulic tilting gear has a capacity of 226,000 in.-lb and is powered by a 20-hp motor. d. Fin Tilting Gear. The fin shafts of virtually all successful installations are tilted by hydraulic power, which is well suited to the requirements for rapid reversal. The tilting gear is generally similar to an electrohydraulic steering gear, such as described earlier in this chapter. Both Rapson-slide and vane-type rotary actuator installations are used; they are actuated by variable-delivery pumps whose flow direction is ordered by the amplified automatic control signal. Some of the significant differences from steering gear practice are: (1) There is usually only one power unit per fin shaft, since stabilization is not as critical a function as steering, and even if one unit fails there is another on the other side of the ship; (2) Whereas steering gears are not required to perform high-speed hard-over to hard-over rudder maneuvers for extended periods, the tilting gear may be heavily loaded for many hours of continuous operation, hence requiring appropriate motor ratings and system coolers; (3) Where retractable fins are used, the stowing and rigging mechanisms, interlocks, and indicators have to be provided; (4) Fin angular rates are about ten times greater than rudder rates (in Appendix 111 of reference [25], the minimum time for "hard-over to hard-over" fin angles is derived as 0.22 times the natural period of roll) ; (5) the rapid reversal of fin angles requires added torque to overcome the mass inertia of the mechanism (some allusions to this calculation are contained in Appendix B of reference [21] and in the author's closure of reference [241). e. Automatic Controls for Fin Stabilizers. Unlike the situation for rudders, human control of fin angles is ineffective. This is because the roll response of a ship in a seaway is at such a high frequency that the fin angles must be varied continuously and rapidly. Automatic controls have proved to be reliable and efficient. Essen-
603
roll angle of ship natural frequency of roll Ts = a damping ratio, involving GM,appendages, and hull form 7 = effective wave slope 7' = 7 2f , = ~ disturbing moment 7 = stabilizing moment of fins, expressed in terms of equivalent wave slope U = ship speed Kl9K2,K3 = constants of the particular system a = fin angle of attack = w =
+
i-
Usually a computer study is made to assure the desired stabilizer performance throughout the spectrum of sea conditions. Some coefficients can be calculated; others are estimated based on specific model tests or specialized hydrodynamic literature, such as reference [27]. In a computer study, the gains and sensitivities can be varied to suit the particular installation. Almost all successful controls use the ship roll velocity as the principal input, plus some roll angle. A case can be made for roll acceleration input and sway input in some applications. Many of the servo designs order fin angle, and measure the actual fin angle to get the closed-loop feedback. In some designs, e.g., reference [24], fin lift is used rather than fin angle. In typical shipboard designs 4 is measured by a gyro or athwartship accelerometer device, B, is measured by a rate gyro or by differentiating the roll angle, and 4 is measured by angular accelerometers. Some studies have been made of feedahead devices utilizing pressure taps port ancl starboard to get wave slope inputs. These devices have not been practical because the real seaway is not the simple wave shown in Fig. 36. Instead, the wave form is the sum of very many waves of varying height, period, and direction, as described in reference [27]. In most cases, the automatic controls are designed simply to reduce the roll angle. Several variations, however, may be considered, such as: (1) stabilizing to the apparent vertical (for improved passenger comfort in sway-
References
1 ('Cleaning and Protecting Integral Piping for Oil Hydraulic Power Transmission Equipment ," Military Standard MIL-STD-419. 2 John Flodin, "Hull Machinery" in Marine Engineering, vol. 11,edited by Herbert Lee Seward, SNAME, 1944. 3 Frank C. Messaros, "Steering Gear and Deck Machinery" in Modern Marine Engineer's Manual, vol. 11,edited by Alan Osborne, Cornell Maritime Press, New York, 1943. 4 A. M. Nickerson, ~ r . , ' a n dV. A. Olson, "Steering Gears and Their Selection," Tram. SNAME, 1952. 5 Philip Mandel, "Ship Maneuvering and Control" in Principles of Naval Architecture, edited by John P. Comstock, SNAME, 1967. 6 Karl E. Schoenherr, "A Program for the Investigation of the Rudder-Torque Problem," Marine Technology, SNAME, July 1965. 7 "Standard Specification for Cargo Ship Construction," Maritime Administration. 8 "Steering Gears, Electro-Hydraulic, Marine," Military Specification MIL-S-17803. 9 Rules for the ClassiJication and Construction of Steel Vessels, American Bureau of Shipping, New York. 10 R. M. Luke and F. P. West, Jr., "An Integrated Steering System," New England Section, SNAME, 1960. 11 William H. Hunley, "Anchor, Mooring and Towing Arrangements; Steering" in Ship Design and Construction, edited by A. M. DIArcangelo,SNAME, 1969. 12 "Guide to the Design and Testing of Anchor Windlasses for Merchant C%ips," Technical and Research Bulletin No. 3-15, SNAME, 1964. 13 "Dimensions of Wildcats," Navy Department Drawing 52601-860304. 14 Catalog, Baldt Corporation, Chester, Pa. 15 A. M. Nickerson, Jr., and C. H. Russell, "Anchor Handling and the Anchor Windlass," SNAME, New England Section, 1957. 16 "Windlasses, Anchor, Electric-Hydraulic, Verti-
MARINE ENGINEERING
cal and Navy Type, Naval Shipboard," Militajr Specification MIL-W-19623. 17 D. F. MacNaught, "Cargo Handling" in Ship Dasigrl and Construction, edited by A. M. D'Arcangelo, SNAME, 1969. 18 P. B. Harwood, Control of Electric Motors, John Wlley & Sons, Inc., New York, 1952. 1g "Capstans and Gypsy Heads," Navy Department Drawing 82601-860303. 20 S. Hawkins, "The Use of Maneuvering Propulsion Devices," Robert Taggart, lnc., Report RT-8518, prepared under contract MA-3293 for the Office of Research and Development, Maritime Administration, January 21, 1965. 21 Hector T. McVey, "Anti-Roll Fin Stabilizers," Philadelphia Section, SNAME, 1959. 22 J. H. Chadwick, Jr., "On the Stabilization of Roll," Trans. SNAME, 1955.
1
23 A. J. Giddings and R. Wermter, ('A Survey of Ship Motion Stabilization," 5th Symposium on Naval Hydrodynamics, Office of Naval Research, ACR-112, 1964. 24 J. E. Flipse, "Stabilizer Performance on the SS Mariposa and SS Monterey," Trans. SNAME, 1957. 25 J. E. Conolly, "Rolling and Its Stabilisation by Active Fins," Quarterly Trans. Institution of Naval Architects, London, vol. 111, no. 1, January 1969. 26 A. M. Robb, Theory of Naval Architecture, Jlondon, 1952. 27 E. V. Lewis, "The Motion of Ships in Waves" in Principles of Naval Architecture, edited by John P. Comstock, SNAME, 1967. 28 J. Bell, "Ship Stabilization Controls and Computation," Trans. Institution of Naval Architects, London, vol. 99, 1957.
Burr
I Electric Plants Section 1 Introduction
1.1 Nature and Scope of Electric Plants. A- complete shipboard electric plant is similar to the electric power generating, distribution, and utilization system of a self-contained shore-based industrial installation. Electric power is required for motors driving propulsion plant auxiliaries and deck machinery, interior and exterior illumination, navigation lights, ventilation and air conditioning, stores and cargo refrigeration, electric heating, galley equipment, drinking water and sanitary systems, and casualty control machinery such as fire and bilge pumps. Power must also be supplied for interior communication systems, announcing and alarm systems, radio communication, radar, and other electronic aids to navigation. For passenger vessels, the electric power requirements extend to hotel and recreation loads, theater and dance floor lighting, restaurant and swimming pool equipment, motion picture projection, public address systems, and stewards call systems. For passenger and prew safety, the electric installation includes automatic fire detecting and alarm systems, power-operated watertight doors, and electrically released, spring-closed fire screen doors. Electric power is vital to all shipboard operations and to the safety and comfort of the passengers and crew. For this reason, shipboard electric plants must contain equipment necessary to maintain continuity'of service, since a vessel at sea is isolated from external sources of electrical energy. Therefore, standby ship service generating capacity, usually equal to the rating of one of the ship service generators, is provided. In addition, one or more sources of emergency power, designed to automatically assume load upon loss of ship service power, are required to supply those loads that are necessary for the of the passengers and crew; the source of power should also have additional capacity adequate to supply those loads vital to getting the pmpulsion plant and generators back in service. Quick-starting diesel generators are usually provided for emergency power; however, storage batteries or gas turbine driven generators are satisfactory for this service. Emergency - - storage batteries combined with d-c/a-c motor generator sets are required on passenger vessels to provide temporary emergency power to certain vital loads until the emergency generator can start and assume the entire emergency load.
To avoid prolonged shutdown at sea, adequate spare parts should be stowed aboard ship to replace vital parh which are subject to wear and breakdown. It follows that adequate detail drawings and manuals containing instructions for operation, repair, and adjustment also should be placed aboard ship. For greater dependability at sea, electric equipment necessary for the operation of the vessel is required to have certain marine features such as dependable operation during rolling and pitching of the vessel, mechanical parts resistant to shipboard vibration, and windings and hardware resistant to moisture and corrosion. A shipboard electric plant includes: generating equipment; switchgear for control of the generators and distribution of power; and distribution panels, transformers, motor generators, and bus transfer equipment as necessary to provide the proper type of power to electrical loads. 1.2 Rules and Regulations. Merchant marine electrical installations must comply with a number of laws promulgated by the United States Government. The following is a listing of such rules and regulations. 1 United States Coast Guard, Code of Federal Regulations, Title 46-Shipping. 2 United States Coast Guard-Rules of the Road, International-Inland. 3 United States Department of Health, Education and Welfare-Standards of Sanitation and Ratproofing for the Construction of Vessels. 4 Federal Communication Commission Rules and Regulations. I
It should be noted that the requirements of the Intemationd Conference of Safety of Life at seaof 1960 are incorporated in item above. addition to the rules and reguhtions imposed by law,
11
d
compliance with other rulesfG usually required by the specifications for building particular ships. The most generally included in ship,s specifications am as follows: 1 American Bureau of Shipping Rules for Building and Classing Steel Vessels, Lloyd's Register of Shipping Rules and Regulations for the Construction and Classing of Steel Vessels, or similar
I I
11
606
MARINE ENGINEERING
classification society rules depending 'on which society's rules are selected as a prerequisite for marine insurance coverage. 2 IEEE Standard No. 45-Recommended Practice for Electric Installations on Shipboard, published by the Institute of Electrical and Electronics Engineers. 3 Suez Canal Authority Rules of Navigation. 4 Illuminating Engineering Society-Marine Lighting. 5 Marine-Type Electric Lighting Fixtures UL 595 published by Underwriters Laboratories, Inc. 1.3 Alternating-Current Electric Plants. Alternatingcurrent plants are the standard for most marine installations. Selection of a-c plants as standard over d-c plants -provides many significant advantages, e.g., savings realized in first cost, weight, and space requirements, reduction in maintenance effort, better availability of equipment, and increased reliability. Many of these advantages are realized through the use of squirrel cage motors in lieu of d-c motors having commutators and associated brush rigging. A frequency of 60 cps is recognized as the standard for a-c plants. Alternating-current plants may be any one of the following types: 120 volt, 3 phase, 3 wire 230 volt, 3 phase, 3 wire 450 volt, 3 phase, 3 (or 4) wire The 120-volt, 3-phase, 3-wire a-c generating plant with 115-volt, 3-phase light and power distribution is suitable for only small vessels having few motor-driven auxiliaries. The 230-volt, 3-phase, 3-wire acc generating plant with 220-volt, 3-phase power distribution and 115-volt, 3-phase lighting distribution through transformers is acceptable to the regulatory agencies. As an alternative, a 120/20%volt, 3-phase, 4-wire system may be used for both power and lighting on small ships without the necessity of using transformers. However, these plants are seldom selected since there are no advantages to be obtained when' compared with the standard 450-volt, 3-phase, 3-wire plant. The 450-volt, 3-phase, &wire a-c generating plant with 440-volt, 3-phase power distribution and 115-volt, 3-phase lighting distribution through transformers is the standard for most marine installations. However, the 450-volt, 3-phase, 4-wire, a c generating plant provides advantages and should be considered for' vessels having extensive 208- or 220-volt container refrigeration requirements. On vessels with'a-c electric propulsion machinery, the high-voltage main propulsion generator may be used to power large motors driving cargo handling machinery when the vessel is in port. Cargo pumps on electrically propelled tankers are typical high-voltage large-motor applications. Usually, 2500 kw is considered the maximum feasible capacity for 450-volt, 3-phase shipboard generators.
Larger capacity generators a t this voltage will not generally be used because of generator design limitations. As the demand for electric power aboard ships increases, particularly for nuclear propulsion plants and vessels with unusually large cargo handling power requirements, generators having a capacity larger than 2500 kw will be required. To provide this increased capacity, it is feasible to install prime generating plants designed for 2400- or 4160-volt, 3-phase output. In this type of installation, large motors would be powered from the high-voltage system with normal ship service loads of 450 volts supplied from the high-voltage system through transformers. 1.4 Direct-Current Electric Plants. Direckurrent plants are economically feasible on relatively small vessels and on those intermediate-size vessels on which a preponderance of deck machinery requiring d-c wide speed range motors and controls is installed. Directcurrent generating plants may be of either of the following types: 120 volt, 2 wire 240/120 volt, 3 wire The 120-volt, 2-wire d-c generating plant with 115volt light and power distribution is suitable for small vessels having few motor-driven auxiliaries. The generators are usually rated a t 75 kw or less. The 240/120-volt, 3-wire d-c generating plant is the most usual arrangement used on d-c vessels. This arrangement provides for 230-volt, 2-wire distribution to power loads and 230/115-volt, 3-wire distribution to lighting panels. The neutral of this 3-wire dual voltage system should be solidly grounded. Lighting fixtures, appliances, bracket fans, etc., supplied from lighting panels are connected between positive and neutral lines and neutral and negative lines in as nearly as practicable equal loads. 1.5 Drawings. The following is a listing of the types of electrical system drawings usually prepared by the shipbuilder. For a detailed listing of plans required for a specific ship, reference may be made to the detailed specifications for building a similar ship. 1 One-line diagram for power distribution system. 2 Isometric wiring diagrams for power system feeders and mains. 3 Isometric wiring diagrams for lighting system feeders. 4 Elementary and isometric w i h g diagrams for interior communication and electronic systems. 5 Deck arrangement plans for power, lighting, interior communication, and electronic systems. 6 Power system list of feeders and mains. 7 Lighting system list of feeders and mains. 8 List of motors and controllers. 9 Wireway locations and details. 10 Electric load analysis of generating plant. 11 Fault-current analysis and voltage dip calculations. 12 Application and coordiaation of protective devices. 13 Switchboard schematic and instrumentation.
ELECTRIC PLANTS
Generating Plants I
I
1:
.,
.
2.1 Load Analysis. To determine the correct aggregate rating for a generating plant, it is necessary to establish the probable peak loads under the various operating conditions of the ship. This is accomplished by the preparation of an "Electric Load Analysis," an example of which is shown in Table 1. This analysis is a detailed tabulation of the total connected load and the operating loads at sea, during maneuvering and a t port. Operating loads are determined by applying a service factor to the expected connected load for each application for each operating condition. The service factor assigned to each application is a combined load factor and diversity factor representing the percent of its own possible maximum that is contributed to the load on the generator plant over a %-hour period. Ocoasional loads such as fire pumps, anchor windlass, capstan, and boat winches are assumed to have a zero factor under all operating conditions. 2.2 Aggregate Generating Capacity. The aggregate generating capacity, exclusive of any, emergency and propulsion generating equipment, will always be greater than the peak load determined by the load analysis. Ship generating plants must consist of at least two ship's service generating sets so rated that, with one set not in operation, the remaining set or sets can carry the necessary sea load under normal operating conditions without exceeding the normal rating. The probability of installing additional electrical loads a t some future time should also be considered when determining the aggregate generating capacity. 2.3 Number and Rating. After determination of the peak load, the next step is the selection of the most desirable number and individual rating of the generating sets. The factors to be considered are: (a) the first cost, (b) operating cost, (c) size and weight, and (d) desirability of using generators of a standard size. Figure 1 (for turbine-driven generators) and Fig. 2 (for diesel-driven generators) show the variation of factors (a), (b), and (c) plotted against generator rating. No precise mathematical evaluation of the optimum number and rating of generator sets is possible; therefore the final selection must be based on experience, consideration of data delineated by Figs. 1 and 2, and available standard generator ratings. For peak loads up to about 2200 kw, two generators are usually provided, each capable of carrying the total sea load. There is a trend with steamships to fit one turbine-driven and one diesel-driven generator usually of equal capacity. This permits the full sea load to be handled in the event of a boiler failure. Suitable arrangemerits can be made for parallel operation of the turbogenerator and the diesel generator; such an arrangement does not preclude the fitting of an emergency generator. For loads greater than 2200 kw, three or more generators are usually installed with the total number of
generators selected providing maximum flexibility, reliability, and continuity of service for the electrical distribution system. 2.4 Location and Installation. On most vessels the ship service generators are located in the main engine room. This reduces the number of watch standers and provides the most economical piping arrangement since piping for generator sets and associated auxiliaries and for propelling machinery are comporlent parts of the same piping systems. On large vessels with two or more engine rooms, there is usually ope ship. service generating plant complete with auxiliaries and switchboard in each of the engine rooms. In general, ship service generators and associated switchboards are normally arranged so that the generators are in view of the switchboard attendant and so that the run of electric cables from each generator to the switchboard is as short as possible. Emergency generator sets must be located ~ b o v ethe freeboard deck, aft of the collision bulkhead, and outside the machinery casing to satisfy regulatory body requirements. All generating sets should be installed with the shaft fore-and-aft so that rolling of the vessel will not cause undue loads on the bearings due to gyroscopic effects,or cause oil to spill from the bearing bousings. 2.5 Generator Set Characteristics. The considerations governing the choice of a d-c or a-c electric plant and the selection of a distribution system and the voltage a t which it is to operate are discussed in Section 1. These basic features having bee^ determined and the number and rating of generating sets determined, the characteristics to be specified by the ship design engineers in the request for bids from manufacturers aye discussed in the following. The prime mover for ship service generators may be either a steam turbine, gas turbine, diesel engine, or a combination thereof, Emergency generators are usually diesel driven. A discussion of the different types of turbines and engines may be found in Chapters 5, 6, and 7. The generating set is selected after an evaluation of bids has been made, taking into consideration the price, weight, dimensions, fuel consumption, reputation of engine and generator builders, and the quantity of similar sets already in service. Important factors in the selection are simplicity, rdiability, and maintainability. a. Steam Turbine Prime Movers. Turbine-generator sets designed especially for installation, operation, and servicing on shipboard are available. They are designed for single or parallel operation for long periods without shutdown for maintenance or repair. Marine generator turbines are horizontal, multistage, and of the axial-flow impulse type. Each marine turbine generator set should be compact and complete with turbine, speed reduction gear,
, -"
1
I
I
i
I, I
1 1
I1
11
li I,
1
B 1
I II 1 I
I
I
610
ELECTRIC PLANTS
MARINE ENGINEERING
STEAM RATE AT FULL LOAD 3
9
,
8
0 I
23
C I-
a
2
3 I-
Y
3
w
-
ul
z . 5 9 2.0. k.40 P
240
7
6 5
2
5.10 z3 1 0 3 2 0- 9- 0",
0 -
W
u
Fig. 1 Approximate unit cost, weight, and deck area required and steam rate of geared-steam turbine generator sets
4
h.30 z 3 0 5 k.20 :20 2
0
3
Fig. 2 Approximate unit cost, weight, and deck area required and fuel consumption of diesel-driven generator sets
200 300 400 500 R A T E D KILOWATT OUTPUT
l
I
0
500 1000 1500 2000 R A T E D KILOWATT OUTPUT
NOTE: DATA IS B A S E D ON 1800 RPM, DIRECT- CONNECTED
2500
FOLLOWING CHARACTERISTICS: 1200 R P M , 450 VOLT, 3 P H A S E , 60 C Y C L E A C G E N E R A T O R ; S T E A M CONDITIONS 850 PSIG 950 D E G R E E S F. T O T A L T E M P E R A T U R E . A N D 28.5 INCHES OF M E R C U R Y '
generator, rotating exciter (if used), condenser, condenser air ejector, gage board, and a self-contained lubricating oil system consisting of a cooler, reservoir, strainer, attached pump, hand-operated pump, and electric motor driven pump (when specified), all supported on a common base with provision for lowering into place as a unit. For some units the shipbuilder may prefer to procure or build the condenser separately and combine it with the set a t the building yard. The set should be complete with all attached piping and accessories supported in place and ready for connecting to shipboard piping. Pipe connections should be either flanged, brazed, or welded. The design should preclude leakage of vapor, water, and oil to the outside and accumulation of liquid in pockets. The turbine should be complete with the following equipment. 1 Steam governing valve assembly. 3 Oil relay type constant-speed governor or electrohydraulic load-sensing speed governor when small speed variations and quick response are desired. 3 Speed adjusting (synchronizing) device with local manual adjustment and electric motor or potentiometer for remote adjustment from the switchboard when synchronizing generators. 4 Combined trip and throttle valve held open by lubricating oil pressure and automatically shutting off steam to the turbine upon loss of oil pressure; also for admitting steam gradually by hand when starting up a set. 5 Interlock switch for energizing the generator circuit-breaker tripping device, to disconnect the generator from the switchboard in case the throttle,trips closed. 6 Overspeed governor independent of the constant-
speed governor, which upon overspeed of not over 15 percent will relieve oil pressure from the throttle trip valve to shut down the turbine. 7 Steam sealing manifold for pressurizing the turbine shaft packing to prevent entrance of air into the turbine, also piping and valves for drainage of steam leakage through the packing. 8 Automatic atmospheric relief valve for exhausting to the atmosphere in case a high exhaust back pressure occurs in the turbine casing due to a condenser malfunction. 9 Sentinel valve for sounding an alarm before the automatic atmospheric relief valve functions. 10 High exhaust back pressure trip device which will relieve oil pressure from the throttle trip valve to shut down the turbine. 11 Manually operated rotor-turning device for use during maintenance. 12 Reduction gear, generally single reduction and single helical, with pinion and low-speed gear shafts supported by two bearings. The low-speed shaft is generally flexibly coupled to the generatq shaft. 13 Steam supply strainer for protecting the turbine. The strainer may be integral with the combined trip and throttle valve. Marine turbine-generator sets are provided with a self-contained lubricating oil system. A gear-driven oil pump supplies oil to the turbine, reduction gear, and generator bearings, and also to the constant-speed governor, trip throttle valve, high exhaust back pressure trip device, and overspeed governor. The pump takes suction from the reservoir and discharges through a magnetic duplex-type strainer and a tube-type cooler. Oil coolers are generally designed to use 85 F (maximum)
450 VOLT, 3 P H A S E , 60 C Y C L E A.C. GENERATORS. S E T S A R E SELF,- C O N T A I N E D W l T H B A T T E R Y START!NG EQUIPMENT: C U R V E S A R E FOR S E T S W l T H 2 C Y C L E ENGINES.. F U E L CONSUMPTION A N D P R I C E OF 4 CYCLE E N G I N E S W l L L BE APPROXIMATELY T H E SAME BUT W E I G H T A N D DECK A R E A W l L L BE S L I G H T L Y HIGHER. PRICE, WEIGHT, A N D DECK A R E A O F LOWER S P E E D E N G I N E S W I L L B E A P P R E C I A B L Y H I G H E R A N D FUEL CONSUMPTION W l L L BE SLIGHTLY H I G H E R .
seawater for cooling and are provided with zinc anodes exhaust from auxiliaries when the ship is a t anchor. on the seawater side to minimize corrosion. The coolingDue to the trend toward automated vessels, additional water pressure should be maintained lower than the oil monitoring and control devices such as those devices pressure to avoid seawater contamination of the oil in necessary for remote start-up, operation, and shutdown the event of a cooler tube failure. A hand-operated of the turbine-generator set may also be incorporated in lubricating oil pump is provided for use during start-up the design of the set. and maintenance. Generally, the lubricating system b. Diesel Engine Prime Movers. Diesel engine also includes an electric motor driven pump which is driven ship service generators should, like turbine started automatically by a pressure switch upon failure generator sets, be designed for continuous operation for of the turbinedriven pump; this pump may also be used long periods of time, singly or in parallel. They are during start-up and shutdown operations. When spec- usually located in the engine room. Because of their ified, the lubricating gystem contains switches for low relatively slow normal speed, diesel-generator sets are oil pressure and high oil temperature alprms. The oil appreciably larger and heavier than turbine-generator level in the reservoir is measured by a dipstick and a sets. Sets --rated rqt_ 150-kw and below are available a t visible level indicator, when specified. S J X up ~ ~tcJ800 rpm. Above 150 kw, most sets are Gageboards should include, as a minimum, pressure d e s i g n e n r either 900 or 1200 rpm. Also because of gages for the inlet steam, gland sealing steam, bearing the low speed, the generator is coupled directly to the oil, and oil pump and thermometers for oil to and from engine. The generator may be of the two-bearing type the cooler. with the rotor flexibly coupled to the engine crankshaft The gageboard and control devices such as speed or may have, a front bearing only with the rear shaft end adjustment knob, throttle trip valve handwheel, hand oil coupled rigidly to, and supported by, the diesel engine pump handle, gland seal steam control valve, hand crankshaft. Generator bearings may be integral with shutdown trip lever (or button), tachometer, and valve the generator end brackets or supported separately by for regulating the oil-cooler water should be located so pedestals. If a rotating exciter is used, the armature that one man can start, operate, and secure the set. will be overhung on the gen-tor front- haft extension. A motor-driven condensate pump and a condenser Two- and four-cycle engines are available for marine circulating water pump are generally provided for each electric plants. Four-cycle engines tend to be heavier turbine exhaust condenser. These pumps are normally and costlier but more efficient than two-cycle engines. provided by the shipbuilder but may be furnished with Most marine engines above 1000 hp are of the two-cycle the turbine generator. Pipe connections may be made type. Two-cycle engines are generally equipped with so that during emergencies the generator turbine exhaust attaclyxl positive-displacement blowers to supply scavcan flow to the main propulsion turbine condenser. The enging air to expel gases from the cylinder a t the end of generator turbine condenser may be arranged to receive the exhaust stroke. I n addition, both two- and four-
612
MARINE ENGINEERING
cycle engines may be equipped with turbochargers (driven by gears, exhaust-gas turbine,, or a combination of both) to increase the engine capacity and provide improved fuel economy. Most marine engines of 2000 hp and over are turbocharged to reduce their size and weight. It is important that provisions be made for sufficient combustion air, either by running air ducts from the outside directly to the engine air intake, or by means of the room ventilation system. Additional room ventilating air also must be provided for removing engine waste heat not removed by the lubricating oil and jacket water systems. The ductwork must be checked for pressure drop from the outside atmosphere to the engine; usually not over 6 in. of water is allowable to obtain guaranteed engine performance. For the same reason the exhaust-gas pipeline from the muffler to the atmosphere should be sized to obtain a back pressure at the muffler outlet not exceeding about 16 in. of water. The complete marine engine includes additional attached and unattached equipment as follows: 1 Fuel control system consisting of an engine-driven fuel pump, duplex filter, suction strainer, and injector control lever for manual starting, stopping, and emergency speed control. 2 Lubricating oil system consisting of an engine driven oil pump, full-flow filter with by-pass relief valve, strainer with relief valve, and cooler with by-pass relief valve. When dry-sump engines are used, the lubricating system includes a scavenging pump with suction strainer and by-pass relief valve. 3 Piston cooling oil pump. 4 Freshwater systems for diesel-generator sets larger than 350 kw consisting of an expansion tank, enginedriven water pump (some engines may have two pumps) cooler, and automatic water temperature regulator. Smaller sets generally have a radiator and fan for cooling fresh water. 5 Exhaust system consisting of a dry, spark-arresting type muffler and watercooled or insulated exhaust header. For 350-kw generator sets and larger, the engine is usually equipped with an exhaust temperature indicating system consisting of a a t of thermocouples, a selector switch, and galvanometer. One thermocouple is installed in the exhaust of each cylinder. The selector switch and galvanometer may be mounted on the gageboard. 6 Starting air system consisting of a starting motor (two motors may be necessary to accelerate large engines) or air distributor for sequential admission of air to the engine cylinders, strainer, and air control valve. A solenoid-operated valve may be provided for remote engine starting from the switchboard. Starting air systems are generally designed to operate a t pressures of 125 to 250 psi. Starting air tanks and the air compressors are normally furnished by the shipbuilder. Diesel generators rated at 500 kw or less generally use either an electric or hydraulic starting system. Capacity for a t least ten successive starts, beginning with a cold engine, is provided with either method. Either a centrifugal or
pressure-operated device is provided to prevent inadvertent attempts to initiate cranking after the engine has started. These devices or a solenoid-perated valve are used to automatically stop cranking after the engine has started. Also, means are provided to prevent unintended starting during maintenance of the set; for electrically started engines, a disconnect switch is provided in the starting clcuit; for hydraulically and pneumatically started engines, a cutoff valve is provided in the associated piping system. 7 Air intake filter-silencer. 8 Oil relay type constant-speed governor or electrohydraulic load-sensing speed governor when small speed variations and quick response are desired. 9 Speed adjusting (synchronizing) device designed for local manual adjustment and an electric motor or potentiometer for remote adjustment from the switchboard when synchronizing generators. 10 Overspeed trip device which upon overspeed of not over 15 percent will close the fuel racks or combustion air supply to shut down the diesel. 11 For ship service diesel generators, an interlock switch for actuating the generator circuit-breaker tripping device to disconnect the generator from the switchboard when the engine is shut down due to overspeed. 12 Gageboard including, as a minimum,. pressure gages for the freshwater and seawater pump discharges, fuel oil and lubricating oil flter inlets and outlets, lubricating oil strainer inlet and outlet, scavenging air, starting air, and also thermometers for the fresh water and lubricating oil from the engine. 13 Manual engine-turning gear for turning small engines or an air-motor type turning gear, with an interlock, for turning large engines. 14 I n addition to the aforementioned equipment, turbocharged engines are equipped with a scavenging turbocharger and turbocharger air intake cooler. Tubetype coolers are used for cooling the lubricating oil and fresh water for diesel generators larger than 350 kw. Seawater is used as the cooling medium in the freshwater cooler and fresh water is used as the cooling medium in the lubricating oil cooler. For sets rated 350 kw or less, fresh water is generally cooled by a radiator and engine-driven fan. If diesel generator sets are orderedior paralleling with existing generators or with new sets procured from another manufacturer, the required speed regulations and electrical characteristics must be specified. c. Gas Turbine Prime Movers. Gas turbine driven generator sets are available for ship service or emergency power and a limited number have been installed. Gas turbines are smaller and Lighter than comparable steam: turbines or diesel engines. Since gas turbines operate a t very high speeds, they are usually noisier than steam turbines. The starting time for gas turbines is normally 30 to 40 sec, but it can be reduced below 10 sec when used for driving emergency generators. Batteries or compressed air may be used for starting.
I
ELECTRIC PLANTS
d. Electrical Characteristics. Except for small loads supplied from batteries, all direct-current loads are supplied by a-c/d-c motor generator sets or by rectifiers. Rotating field type a-c generators are used for ship service power. The generators are rated at 450 volts, 3 phase, 60 cycles. Stator windings may be either delta- or wye-connected but usually are the latter; only three main terminals are required per generator. The inherent voltage regulation of a-c generators is comparatively wide due to the high synchronous reactance of the windings. This is an advantage inasmuch as the synchronous reactance limits short-circuit current, but in order t o maintain the required voltage regulation an automatic voltage regulator must be used with each machine and the degree of regulation depends on the sensitivity of the regulator. Generators may be dripproof protected or totally enclosed. If totally enclosed, the generator will be equipped with a double-tube air cooler using seawater as a cooling medium. Silicone insulation should not be used for totally enclosed generators unless the slip rings are located outside the generator enclosure; this is to prevent abnormal brush wear and increased slip ring maintenance. If the design and arrangement of the generator are such that circulating currents may be expected in the rotor shaft, means (such as the use of insulated bearings) should be provided to prevent circulating currents from passing between the journals and the bearings, as the babbitted bearing surfaces may otherwise be destroyed. Ship senrice generators that weigh more than 1000 lb, excluding the shaft, and all emergency generators should be provided with electric space heaters to prevent moisture condensation during shutdown. Generators rated a t 500 kva and above should be provided with resistance-type temperature detectors embedded in the stator windings, The temperature indicating instrument should be located conveniently, .preferably on the switchboard generator control panel. An excitation system is procured with each a-c generator. Two types of rotating exciters are the d-c exciter and the a-c brushless excitkr. Both of these exciters are coupled to the generator shaft and are similar in outward appearance. In place of- the commutator on the d-c exciter, the a-c brushless exciter has a solid-state three-phase rectifier mounted on the generator shaft to provide the d-c excitation for the generator field. The a-c brushless exciter responds faster than the d-c exciter. A third type of exciter in general use is the static excitation system. This system eliminates the necessity of rotating components and has a faster response than either type of rotating exciter. ,Selection of the exciter and voltage regulator should be coordinated to obtain the desired recovery time. The a-c. brushless exciter and the static excitation system have in general replaced the rotary amplifier type of exciter. e. Emergency Generators. Emergency generators provide a power source independent of any other equip-
61 3
ment on the vessel and are usually diesel driven. The requirements outlined in the preceding paragraphs for diesel engine prime movers apply equally to emergency sets, except as noted in the following. Each engine is equipped with a self-contained cooling system requiring a radiator and fan. Tf ventilation ducts are installed to and from the engine radiitor, the radiator fan must develop enough head to force the cooling air through the ducts. Motor7:;dperatedlouvers are generally installed in supply and exhaust duct terminals which are exposed to the weather. The vent motors are energized from the generator side of the generator circuit breaker to insure that the louvers are open when the diesel engine is in operation. Emergency generator units should be arranged to shut down automatically upon loss of lubricating oil pressure, dangerous overspeeding, or release of carbon dioxide in the emergency generator room. An audible alarm device should be provided that will sound in the event of low oil pressure or a high cooling-water temperature. Engines are generally arranged to start automatically upon failure of ship service power. If battery-started, a voltage-sensitive relay with contacts that close when the ship service power fails is used to energize the control circuit of the starting motor; if hydraulically started, a loss of ship service power will de-energize a solenoidoperated valve to initiate the starting process. In either case, means must be provided for automatically rendering the starting devices inoperative after the engine has attained firing speed. Devices for manual starting control are required with each type of starting equipment for test purposes. Emergency generators are not required to operate in parallel with the ship service generators, except where closed transition transfer with proper synchronization with ship service generators is desired (to prevent power outage when transferring power sources). An automatic voltage regulator should be provided with each emergency generator. The emergency generator should have the same voltage rating as the ship service generators and should be of sufficient capacity to permit "cold" starting of the ship's main power plant. f. Special Generator Arrangements. A variable generator frequency is obtainable by using special control equipment. Variable-frequency generators may be desirable where speed control of several large motors is required. Generally, it is more economical to use d-c motors supplied by an a-c/d-c motor generator set for most variable-speed shipboard applications. Multiple unit generator sets are available, eack consisting of two generators driven by one steam turbine. Diesel generator sets consisting of one generator driven by two diesel engines have been installed on some vessels. In case one engine fails, it may be declutched and the generator can continue to deliver approximately one half of its rated output. g. Voltage Regulation. A direct-acting type of voltage regulator employing a mechanical regulator element may be used for controlling the field of d-c
614
MARINE ENGINEERING
rotating exciters or very small auxiliary generators. The regulator element may be a torque motor, solenoid, or electro-dynamometer energized by the generator voltage and restrained in motion by a spring. Rheostatic elements are automatically operated field rheostats of which there are two types. One consists of a motoroperated face plate rheostat controlled by means of contacts on the regulator element and the other consists of a resistor with numerous steps cut in and out by mechanical linkage with the regulator element. Both types have antihunting or damping devices to prevent fluttering with small changes in load and overregulation with large sudden changes in load. Either of two types of static voltage regulators may be used depending on the type of excitation system. When a static excitation system is used, the voltage regulator senses the generator output voltage which is rectified and applied to the control winding of a magnetic amplifier. The output of the amplifiers is impressed across the control winding of three saturablecurrent potential transformers, one per phase. The outputs of the transformer secondaries are rectified and impressed across the generator field. If an &c brushless exciter is used, the exciter field current is supplied by a static voltagr: regulator which senses generator output voltage. An error voltage is impressed across a reactor which becomes saturated, conducts, and fires a silicon-controlled rectifier to provide current to the exciter field. The average exciter field current is determined by the point a t which the rectifier fires during each positive half-cycle. Provisions are made in static regulators to provide sufficient excitation that will result in fault currents large enough to ensure selective tripping of overcurrent devices during short-circuit conditions when the generator voltage is zero. The regulators described in the foregoing are suitable for use with generators operating in parallel, each generator requiring an individual regulator. For parallel operation, a compensating effect to reduce wattless
ELECTRIC PLANTS
current is obtained by making use of the potential generated in the secondary of a current transformer, the primary of which is in series with one of the main leads. Each regulator, except for some emergency generator applications, ingludes a transfer switch for cutting out automatic operation and provides for manual control of the generator voltage by means of a rheostat. h. Generator Terminals. The preferred run of cables between the generator and switchboard is downward, under the generator platform, and up behind the switchboard. The preferred location of the terminals is therefore at the bottom of the generator, unless the generator is of such small size that connections may he made inside a standard terminal box mounted on the side of the generator frame. All generator terminals should be protected against accidental contact and mechanical damage. If terminals are located on the top or side of the generator frame, they should be protected by a watertight enclosure with removable covers furnished by the manufacturer. Where cables enter a terminal enclosure on the top or side of the generator frame, the enclosure should be provided with terminal tubes. For terminals located on the bottom of the generator frame, the necessary protection, usually expanded metal, is furnished by the shipbuilder. Additional terminals are required for exciter and imbedded temperature detectors and generator space heaters, if used. Any such miscellaneous terminals should be located and protected in a manner similar to the main terminals. All generator power connections should be silver-plated. The number, type, and size of terminal lugs to accommodate the conductors should be specified and should be clearly marked by the generator manufacturer for identification. Particular care should be exercised when terminals are brought out and marked to assure that the phase rotation with respect to the terminal marking will be the same on all generators, including emergency generators.
Section 3 Switchboards and Panels %
The following does not cover in detail the design and construction of switchboards since these details are the responsibility of the switchboard manufacturer. The procurement of marine switchboards does, however, require that the shipbuilder preparefor the guidance of the switchboard manufacturer-detailed descriptive specifications giving complete information on the required type and arrangement of the switchboard, the number, rating, and wire size of feeder circuits including bus ties, and any limits to the height, length or depth of the switchboard imposed by its location in the vessel. An elementary diagram of the electrical distribution system should also be given. 3.1
General.
The first step in the preparation of switchboard descriptive specifications is a thorough study of the ship's requirements. The ship's requirements determine the number, rating, and type of generators to be controlled and usually impose specific requirements regarding the switchboards. A review of the requirements of all applicable rules is also advisable. 3.2 Special Requirements. The following features should be considered in connection with switchboards for installation on shipboard. Switchboards should be located in dry areas and should be accessible from the front and rear. Also, switchboards should be located as far inboard from the ship's sides as practicable. The
usually restricted space on shipboard requires careful study of the switchboard assembly drawings, hull structural drawings, and machinery arrangement plans to ensure space for installation without interference frdm girders, beams, stanchions, bulkhead stiffeners, and major equipment. The space in front and rear of switchboards should be unobstructed and adequate for operation and mainte nance; minimum clearances are 36 in. in front and 30 to 36 in. a t the rear (see regulatory rules for specific requirements). Since the space behind switchboards should not be accessible to unauthorized personnel, an enclosure is usually provided which extends from the ends of the switchboard to the ship's structure with doors arranged for locking. When this arrangement is not feasible, protective enclosures are mounted on the switchboard rear framework. Switchboards should be located as close to their associated generators as is practicable so as to keep the length of cables to a minimum. Space permitting, neither steam, water, nor oil lines should be located over or dose to the switchboards. Under all conditions, pipe joints should be located away from switchboards. Drip shields should be provided a t the top of all switchboards to protect against dripping from pipes and also against falling objects. Ventilation ducts should not discharge air directly on the suritchboard. Insulating mats or gratings should be provided on the deck in front and rear of switchboards to insulate personnel from ground and to prevent slipping. These mats or gratings should extend the entire length of the switchboard and be of sufficient width to suit the operating space. Nonconducting handrails attached to the front of switchboards are provided; usually these handrails are horizontal. Nonconducting guardrails are also provided a t the rear of switchboards. The switchboard and its component parts as finally installed must be capable of withstanding shipboard vibration without damage or faulty operation and should operate successfully when inclined a t an angle of 30 deg in any direction from the vertical. Special consideration should be given to adequately supportingtthe bus bars. If required, switchboards may be braced to a bulkhead or the deck over. However, overhead bracing should be flexible to allow deflection of decks without causing the switchboard structure to buckle. 3.3 Rating and Characteristics. The rating and characteristics of the ship service generator switchboards and emergency generator switchboards are usually fixed by the rating, type, and arrangement of the electric plant. Three-phase, 450-volt, 60-cps generation and distribution systems are standard for both passenger and nonpassenger vessels. The quantity and kilowatt rating of the generators are dependent upon the total connected loads and demands of the electrical system. Direcb
current power supplies 'are often provided for cargo winches, and other loads requiring precise speed control. This power source will normally consist of motor generator sets with individual control and distribution facilities separate and apart from the main switchboards. Switchboards, both' maia and emergency, usually contain a 3-phase, 120-volt, 60-cps distribution section, energized via transformers, having suitable kva ratings, from the local en-volt, Bphase, 60-cps bus. 3.4 Types bf Switchboards. 'There are two general, types of switchboards, dead-front and live-front. The applicable rules require that dead-front type switchboards be used for all a-c applications where the voltage to ground or between poles is in excess of 55 volts and for all d-c switchboards where the voltage to ground or between poles is in excess of 250 volts. A dead-front design provides for all energized parts to be enclosed within the switchboard structure, whereas a livefront design permits surfacemounted fused lever switches, circuit breakers, and instruments. A deadfront design offers advantages involving personnel safety and equipment protection, and is usually a specific requirement of ship specifications for all marine switchboards. On dead-front switchboards, open-frame circuit breakers, rheostats, and other heavy equipment requiring front-of-board o~eration are mounted on suitable supports fasteneb to interior angles with only the operating handles or knobs projecting through the front hinged panel. Distribution circuit breakers are mounted on suitable support plates fastened to interior angles with only the operating handles projecting through the front panel. Instruments, indicating lights, and control switches are semi-flush mounted on hinged front panels. 3.5 Lighting of Switchboards. Switchboard illumination may be provided by the compartment lighting system. However, a preferred arrangement is the use of lighting fixtures mounted on and under the overhang of the switchboard drip shield. The lights, of which there may be several, are connected in part to the emergency supply so as to provide adequate illumination upon loss of normal supply. 3.6 Arrangement of Switchboards. The usual practice is to provide a panel for the control of each generator and additional panels in number as required for circuit breakers and switches controlling power distribution. With switchboards of small and medium size, the generator panels may be a t one end of the switchboard and all feeder panels to the right or left. For large switchboards, a considerabId9avings in'cost of the main bus may be effected by locating the generator panels in the center with feeder panels of approximately equal working load on each side. With this arrangement no part of the main bus carries more than approximately one half of the total working load. 3.7 Switchboard Applications. Switchboards are normally used in the following shipboard applications: a. Main Generator and Distribution Switchboard. The main switchboard provides for the control, protec-
ELECTR1IC PLANTS
61 6
tion, and paralleling of the local ship service generators and control and protection of the ship service power distribution system. Usually only one main switchboard is required on all but large passenger vessels and special-purpose vessels. For large and special applications two or more main switchboards, connected together through bus ties, may be installed depending on the number and location of main generators. ~ e a n sof connecting the main ship service bus to shore power and also to the emergency switchboard are provided through a shore-power circuit breaker and a bus tie circuit breaker respectively, each located on the main switchboard. Normally main switchboards consist of a generator panel for each connected ship service generator, a shore power and bus tie panel, and distribution panels in size and number as required. Figure 3 is a typical main switchboard one-line diagram. Figure 4 is a typical main switchboard generator control panel one-line diagram. Each generator control panel should include, but not be restricted to, the following : 1 A trip-free circuit breaker with separate overcurrent trips for each pole except that trips are not required for the neutral of dual-voltage systems. The circuit breaker should be arranged to open all three poles with a predetermined load existing on any one or combination of poles. When two generators are installed, each generator breaker should have inverse-time overcurrent trips; for three or more generators arranged for parallel operation, each generator breaker should have inversetime and instantaneous overcurrent trips. (Inverse time tripping means that the opening time of overcurrent devices decreases as the magnitude of the current increases.) For large generators the circuit breaker should be electrically operated to provide personnel protection and for quick closing during paralleling. 2 An unfused disconnect device which will completely disconnect the generator and its circuit breaker from the bus. This device is not required when the generator circuit breaker is of the drawout type. 3 An ammeter, with a selector switch to read the current of each phase. 4 A voltmeter, with a selector switch to read each phase of the generator and one phase of the bus. Also, one of the generator voltmeter switches should provide for reading each ~ h a s eof the shore connection. 5 A sGchro&ope and synchronizing lamps with a selector switch to provide for paralleling ship service generators including the emergency generator when closed transition transfer between the main and emereencv " switchboards is desired. 6 A control for prime mover speed for paralleling generators and frequency adjustments. (The speed control may be located on the engine for emergency switchboards.) 7 An indicating wattmeter. 8 A frequency meter with a selector switch to connect
-
-
:
MARINE ENGINEERING BUS
to any generator. This may be common to all generators. (This is required for the local generator only on the emergency switchboard.) 9 Space and mounting for generator field and exciter rheostats. 10 A switch and indicator light for the generator space heater supply. 11 A double-pole field switch with discharge clips and resistor. For generators with variable-voltage exciters or rotary-amplifier exciters, each controlled by a voltage regulator unit acting on the exciter field, and for generators equipped with static excitation and regulating systems, the field switch, the discharge resistor, and the generator field rheostat may be omitted. 12 A voltage regulator complete with accessories. 13 A switch for cutting out the voltage regulator. 14 A white light to indicate that the generator is running. 15 Adequate means for ground detection. 16 The necessary current and potential transformers. 17 Current and potential test receptacles for use with portable instruments (only on ship service switchboards). 18 One temperature indicator and selector switch, when the generators are rated a t 500 kva and above. 19 A reverse-power relay for each generator, when two or more generators are to operate in parallel. b. Emergency Switchboards. The emergency switchboard provides for the control and protection of the emergency generator and the emergency power, lighting, and interior communication systems. Usually only one emergency switchboard is installed; however, for safety reasons large passenger vessels may have two emergency sources of power, thus requiring two emergency switchboards. Emergency switchboards normally consist of a generator control panel, a bus tie and distribution panel and 24-volt, 120-volt, and 450-volt distribution panels in size and number as required. Emergency generator switchboards include those devices noted in the foregoing for main switchboard generator control panels plus the following: 1 A white light to indicate that the normal supply is available. 2 A green light to indicate that all devices are "set up" for automatic operatio& (This indication should be extended to the ship service switchboard to alert the operator when the emergency generator controls are not properly positioned for automatic startup.) 3 A normal supply circuit breaker, unless the automatic bus transfer is of the contactor type. 4 A feedback switch when the automatic bus transfer is of the contactor type. Except under emergency conditions, the emergency switchboard serves as an extension of the ship service power distribution system with its main bus supplied from the ship service main switchboard through a bus tie and an automatic bus transfer device. This bus transfer device usually consists of two circuit breakers, the emergency generator breaker, and the bus tie breaker,
i
I
I
t TOSHORE PWR TERMINALS
I
t
TOEMER SWITCHBOARD
LEGEND NO. I CIRCUIT BREAKER -- GEN. GEN. NO. 2 CIRCUIT BREAKER C - SHORE PWR CIRCUIT BREAKER POTENTIAL TRANSFORMER. .CT. - CURRENT TRANSFORMER
0 -NORMAL SUPPLY CIRCUIT BREAKER E -0ISTRIBUTION CIRCUIT BREAKER F -LOCAL 4 5 0 / 1 2 0 VOLT XFMR ClRCUlT BREAKER FU FUSE
A
-
PT-
~
Fig. 3 Typical main switchboard one-line diagram
r 450 VOLT, 3 PH, 60 CPS
GENERATOR BUS
I
1
SYN LTS
Q
Q
1
T
C
TO 120 VOLT BUS
SHUNT TRIP
'
------LEGEND VR -VOLTAGE REGULATOR FM -FREQUENCY METER AM -AMMETER SYN SYNCHROSCOPE PT -POTENTIAL TRANSFORMER WM - WATTMETER TM - TEMPERATURE METER CT -CURRENT TRANSFORMER VM -VOLTMETER GOV -GOVERNOR B GENERATOR CKT BKR A -GENERATORCKTBKR FU -FUSE . Fig. 4 Typical main switchboard generotor control panel --line diagram
-
-.
MARINE ENGINEERING 450 VOLT, 3 PH, 6 0 CPS EMERGENCY BUS
619
ELECTRIC PLANTS
I20 VOLT, 3 P H 6 0 CPS. BUS
24 VOLT DC BUS
I
?
?
-
4 5 0 VOLT, 3PH, 6OCPS / F!NAL EMER BUS .-
T
I T
?A
?
TWO BREAKER TYPE ABT INTLK
I
4 5 0 1 120 VOLT TRANSFORMER
t
4 5 0 VOLT POWER
120 VOLT, 3 WlRE FINAL EMER BUS ? ? T
AC/DC MG OR CONVERTER CONTINUOUSLY OPERtTlNG
'
I
450/120 VOLT TRANSFORMER
FINAL EMER LOADS
120 VOLT, 3 WlRE TEMP EMER BUS
1
I
b t AUX
TO
GEN HTR
I
EM GEN CKT BKR AUX CONTACTS
I I
9-' I
EMER LTG
?VITAL AC LOADS
I I
I
I
LEGEND A
- QEN CIRCUI,T BREAKER
B -NORMAL SUPPLY CIRCUIT BKR C T -CURRENT TRANSFORMER WM- WATTMETER U V UNDERVOLTAGE TRIP VR + VOLTAGE REGULATOR
-
-LTG XFMR CIRCUIT BKR C D - 4 5 0 VOLT DISTR CKT BKR E -120 VOLT DlSTR CKT BKR F U -FUSE FM -FREQUENCY METER HTR HEATER
-
-
F BATT CHARGER CKT BKR G 2 4 VOLT DC DISTR CKT BKR PT -POTENTIAL TRANSFORMER AM- AMMETER VM- VOLTMETER SW- SWITCH
CONTACTOR TYPE ABT
LEGEND AUX SW8 ON MN GEN BKRS
AM -AMMETER VM -VOLTMETER F M -FREQUENCY METER WM -WATTMETER MG -MOTOR GENERATOR P T -POTENTIAL TRANSFORMER C T -CURRENT TRANSFORMER ABT- AUTOMATIC BUS TRANSFER
/
i NORMAL SUPPLY Rg. 5
Typical cargo ship or tanker emergmcy switchboard me-line diagram
SKETCH "A"
-
FU FUSE A GENERATOR CKT BKR B NORMAL SUPPLY 450/120 VOLT XFMR CKT BKR C D -DISTRIBUTION CKT BKRS (AS REQ'DI E -MOTOR GENERATOR CKT BKR , F BATTERY CHARGER CKT BKR
-
-
Rg. 6
- AUTOMATICALLY TRANSFERS EMERGENCY LIGHTING TO BATTERY UPON LOSS OF NORMAL SUPPLY @ - AUTOLATICALLY TRANSFERS VITAL AC 6 0 CPS LOADS
ABT @
ABT TO MG UPON LOSS OF NORMAL SUPPLY ABT @ -AUTOMATICALLY TRANSFERS VITAL DC LOADS TO BATTERY UPON LOSS OF NORMAL SUPPLY (MG SET)
Typical passenger vessel emergency switchboard me-line diagram
4
both of which are electrically interlocked so that only one breaker can be in the closed position at a time. The bus transfer device is arranged to initiate an automatic start of the emergency generator set upon failure of the normal power supply. When the emergency generator ig up to rated voltage, the bus transfer operates so as to transfer the emergency bus from the normal supply to the emergency generator. Upon restoration of the normal power supply, the emergency loads may be manually transferred to the normal supply and the emergency generator manually stopped. Retransfer to the normal supply, and particularly shutdown of the emergency generator, should be done only after the continued operation of the normal supply is reasonably assured. See Figs. 5 and 6 for typical emergency switchboard one-line diagrams. On some small-capacity installations, the bus transfer device is a contactor type in lieu of a circuit-breaker type principally because of cost considerations. Functionally the operation of a contactor type is the same as the conventional two-breaker type transfer; however, the arrangement is different as noted in Fig. 5.
When the normal supply is not available, it is often desirable to feed emergency power back into the ship service system. To satisfy this condition when using a circuit-breaker type transfer, provision is made to defeat the electrical interlocking feature, thus allowing both circuit breakers to be in the closed position a t the same time. When using contactor-type transfers, a manually operated feedback circuit breaker is provided to complete r the normal a circuit between the emergency g e n e ~ t oand supply bus tie. When so arranged, electrical interlocking should be provided to prevent inadvertent paralleling of the emergency and ship service supplies. For a passenger vessel the emergency switchboard has two sets of buses, one designated 'final emergency" and one designated "temporary emergency. " The "final emergency" buses (450 and 120 volt) are normally energized through a bus tie from the main switchboard as described previously. The "temporary emergency" buses (120-volt Bwire, 120-volt 1-phase, and 120-volt d-c) are normally supplied from the "final emergency" buses through associated transfer switches and are automatically transferred to a storage battery supply
upon failure of the normal power. This provides instantaneous restoration of the emergency power supply, either directly from the battery or by means of d-c/&c conversion equipment supplied from the battery, to those loads requiring an uninterrupted power source. See Fig. 6 for a typical emergency switchboard one-line diagram for passenger vessels. c. Load-Center Switchboards. Load-center switchboards are essentially remotely loc&ted sections of the main switchboard distribution section. They are supplied from the main switchboard via a bus feeder and in turn supply power to local lighting and power loads. Load centers are centrally located regarding the loads supplied for reasons of convenience and economy and are normally installed on only large passenger vessels having considerable power requirements located throughout the vesael. (Figure 18 is a one-line diagram of a typical load-center application.) 3.8 Selectivity. Selectivity provides for maximum continuity of service under fault conditions through the selective operation of various protective devices, that is, the isolation of a fault with the least interruption of vital senrices.
Selectivity is obtained through the coordination of trip devices of the various breakers in the system. The time bands, both long and short time, of the open-frame type of breakers require special attention for coordin& tion, especially when the spread between breaker trip ratings is very narrow. When fused breakers are employed, very careful attention must be given to the coordination of fuses with the breakers, and with other breakers in, the system. Complete information with regard to generator maximum and sustained fault current and system impedance must be available in order to develop a properly coordinated system. For a more detailed discussion of selective systems see Sub,%section 6.7. 3.9 Circuit Protective Devices. Protective -devices such as oircuit breakers, fuses, reverse-power relays, and current-sensitive relays are installed on switchboards to provide protection against faults in the electrical distribution system. These devices are applied so as to isolate any fault with the least possible portion of the system being interrupted; the arrangement should be such that the generator circuit breaker is the last to open under fault conditions. Each protective device (circuit
620
MARINE ENGINEERING
breaker) must have an interrupting rating adequate to safely interrupt the maximum fault current obtainable at the point of application. Generator and bus tie breakers and breakers feeding combined loads in a selective system are of the openframe type and have long-time and short-time trips. Instantaneous trips may be used in those cases where the circuit breakers have a short-time rating below the interrupting rating, provided they do not defeat selectivity. Breakers feeding individual loads in the system should be of the molded-case or open-frame type with the only requirement being that the continuous and interrupting ratings are adequate for the application. When three or more generators are to operate in parallel, the generator circuit breakers should have instantaneous trips which are set a t a .value in excess of the maximum fault current obtainable from an individual generator. Fuse selection is based on system characteristics (voltage and current) and speed response (standard or time-delay) required a t the point of application. All fuses should be of the nonrenewable cartridge type and capable of interrupting the available fault current. Reversepower relays are provided to prevent a generator from operating as a motor when paralleled with another generator. This relay trips its associated generator breaker when power flows from the line to the generator in lieu of from the generator to the line. Usually the relay is set t o initiate generator breaker tripping within 10 sec when reverse power is approximately 5 percent of the generator rating. I t is sometimes necessary to employ current-sensitive relays for tripping circuit breakers ,at a predetermined current. This becomes of special importance when arranging for selectivity between the emergen:y generator gnd main switchboard bus tie breaker during feedback operation. Under normal conditions (with the emergency switchboard energized from the main switchboard), the bus tie breaker is coordinated with the main generator breaker, and has a relatively high trip value. During feedback operation it is necessary to trip the bus tie breaker should the combined load (both main and emergency) on the emergency generator approach a value likely to exceed the emergency generator rating. This is required so as to provide for continuity of emergency supply to emergency loads. The tripping scheme would normally consist of current transformers and current-sensitive relays properly coordinated and arranged to monitor the emergency generator total current and act on a shunt trip of the main-emergency bus tie breaker. The trip circuit would be electrically interlocked so as not to be effective a t any time other than when operating under a feedback condition. To avoid interrupting vital circuits as a result of overload tripping the generator circuit breakers, nonvital loads may be arranged for automatic tripping when the total load on any generator or bus tie circuit exceeds a predetermined value. Usually, the loads to be tripped
are connected to a common bus or buses; tripping the breaker feeding the bus thus disconnects all loads commonly connected. In lieu of the foregoing, nonvital loads may be sequentially tripped via a multiple-contact timing relay. The contacts are arranged to close serially a t predetermined intervals, and the closing of each contact trips one or more circuit breakers serving nonvital loads until the overload is reduced to an acceptable value. With the overload cleared, the relay contacts open and the timing relay resets. The circuit breakers to be tripped must have either a shunt or undervoltage trip device, undervoltage being preferred because of its fail-safe characteristic. Current transformers of the proper rating are required for each generator or bus tie circuit to be monitored, and a current-sensitive relay is required for each breaker or group of breakers to be tripped. The relay may be instantaneous or timedelayed so as to not initiate tripping on momentary overloads. 3.10 Types of Panels. The panels most commonly installed aboard ship for the many specialized service requirements are of the following types: a. Distribution Panels. Lighting and power system distribution panels have the same'function as do loadcenter switchboards, i.e., supply power to local lighting and power loads respectively. They are supplied from either the main switchboard, emergency switchboard, or a load-center switchboard. Distribution panels are normally of a dripproof construction and located in dry areas central to the loads they supply. Distribution panels should not be accessible to unauthorized persons. They are surface mounted except in passenger, crew, and public areas in which case they are flush-mounted in way of joiner work. Panels are fitted with multipole switches or circuit breakers having a pole and an overcurrent protective device for each associated circuit conductor. In general, 440-volt a-c panels are restricted to a maximum of 12 three-phase circuits, and 115-volt lighting panels are restricted to a maximum of 14 threephase circuits (42 overcurrent devices). Lighting papels are normally arranged for a threephase supply and single-phase distribution; power panels are normglly arranged for a three-phase supply and three-phase distfibution. b. Alarm Panels. Alarm panels are commonly provided for monitoring various criticaT checkpoints associated with the ship service genepators, emergency generator, generator prime movers, and vital propulsion plant auxiliaries and systems. Checkpoints may be monitored a t one common panel or separate panelstmay be provided for specific systems. Alarm panels should incorporate audible and visible indication of a system failure or "off-normal7' condition. Provisions should be made for silencing the audible alarm with the visual indicator remaining "on" to indicate a standing fault. Common practice is to have a "flashing" light indication for the initial fault and a "steady" light
ELECTRIC PLANTS
I
I
I
.
I I
for indication of a standing fault. Suitable means should be provided for indicating an "out of service" condition and for simulation of a fault for testing each system. Alarm circuits should be designed for operation from a 24-volt or a 120-volt supply unless the circuits are simple, in which case a 6- or 12-volt supply should be satisfactory. Each alarm panel should have a control power transformer, power supply fuses, power-available light, and power supply control switch. In some instances remote power supplies are provided for specific systems, and, when so arranged, power-available indicating lights should be provided for each such system. c. Test Panels. A panel arranged to provide test voltages at all values utilized on the vessel should be installed in a convenient work area for the purpose of testing electrical appliances and components. Normally, the *C power supplies (440 volt and 115 volt) to the test panel are from the ship service systems and the d-c supply is from a rectifier of the proper voltage and rating. Complete instructions for operating all test devices should be mounted either on or adjacent to the test panel. d. Special Consoles. A console arrangement provides for the grouping of selected navigation, communication, and/or propulsion plant indicators and controls on
or within a common enclosure usually having a sloping desklike top. Typical examples are wheelhouse command consoles and engineers' operating station consoles provided on vessels having automated propulsion plants. *Typicalexamples of these types of consoles are given in Chapter 21. On vessels having centralized control for propulsion plants, it is desirable that' the h i p service generator control be adjacent to or in the vicinity of the propulsion plant controls. Unle~sthe eritire ship service switchboard can be conve+6ntly located, the generator controls are grouped on a control unit apart from the switchboard so as to permit a choice of location. The control unit may be a vertical type of the same design as if it were part of the switchboard, or it may be of the console type. A generator console type of control unit permits grouping the generator circuit breaker control switches and instrument control switches on the console sloping top. Instruments, indicator lights, and less frequently used control switches are mounted on the vertical area of the console. Rheostats and voltage regulator equipment are usually located in the bottom section of the console. A console type of design is preferred when several generators are to be controlled from one location or when a large number of controls are required for remote devices.
Section 4 Power Eq~~ipment 4.1 Genered Requirements for Motors and Controls. types of enclosures and methods of ventilating motors General recommendations for the construction and most commonly used are as follows: application of motors and control apparatus for marine 1 Open, self-ventilated-used only where an adequate service are contained in the IEEE Standard No. 45. enclosure is provided by the housing of the driven Certain specific requirements are also contained in the machine. This type of enclosure should have ventilaregulations of the classification societies and in the tion openings which permit passage of external cooling United States Coast Guard Code of Federal Regulations air over and around the motor windings. -Title 4Mhipping, Parts 1 to 149. 2 Dripproof protected, self-ventilated-used for most Motor and control equipment for marine service may applications in dry, sheltered locations. This enclosure be classified into several standard types relative to is so constructed that drops of liquid or solid particles mechanical and electrical characteristics. The proper falling on the motor at any angle not greater than 15 application of motors and controls for shipboard deg from the vertical cannot enter the motor. The auxiliaries involves, therefore, the selection of those ventilating opefLings are riormally protected with wire standard types having mechanical and electrical charac- screen, expanded metal, or perforated covers to prevent teristics that are most suitable for the location and na- personnel from contacting electrical parts. These ture of the driven auxiliary. covers also keep out rats and mice that might use the The ratings and characteristics of electrical motors and equipment as nesting places. controls fop a typical singlbscrew cargo liner are listed 3 Totally enclosed, fawcool&-generaly used for in Table 2. motors in spaces where lubricating or fuel oils are present; they are also used for applications subject to splashing, 4.2 Mechanical Characteristics of Motors spraying, or hosedown. This type of enclosure prevents a. Enclosure and Method of Ventilation. A wide the free exchange of air between the inside and outside variety of enclosures and methods of ventilation is' of the housing but is not sufficiently enclosed to be available for motors in marine service. The specific termed airtight. types selected depend on the particular environmental 4 Waterproof, nonventilated-used for practically all condition to which the motor will be subjected. The motors mounted on weather decks, or where heavy
I
I
M A R Y E ENGINEERING
622
ELECTRIC PLANTS
1
L+s
asga 3
L
5
.e
4
a 0
g
0)
L
"
I
r B $
, "
2
FX'B $3 'E: 13 e'&;'z 3% 2e 1 0 0 0 0 0
aaa aa .pba .+a
a a .a
aaaaaa .4a.4a.4a
-48
mmaaa a aa 4aaa4a
YYY ?y 228 st
Y =??V) 1"
a
ge
rn-m-ga m
U
a
3.p Bm
U
B a
U
8 &
UUUU
Ha Ba Ha Ha
a .a .a .a .a.+a.a
UUCU
UU
U
U
U
H l H B 18 B B H aaaa aa a a a
aaaa aa
a
a
a
B
a
I -5
a aaaz 2aaa a,: a a a a
-
?* qr w55 =?r 2 w Rw-*
o a ,?S urw> a-
d
e4
\
,A
j l
? ? w
d
i3g
U
U
B B a a
5
4
623
9599 5 .9 9 5% 5% 59
-.. . " f 9
-$.$
1 xa 1a a' ~a :a1a1 12a a' Ea: a~ ~a . aB ' E1a : 2a a2x1 a
Pi Pin PI
g:ga3
1 2 2 '81 a a
a aa a .a.a
a ~ 28~2 3e
21
,*
8 .~.g:
12 ~5
:. :.
.g:
m e 9 9 9 9d 9d9ddd 99 d dd d dd
a
a
aa
a .a .a-
.a .a a a a a a a a a a a a aaaaaa aasa3
? Y 2 * Yw
Y
E
.2.2.2-2 +; ;. 0 0 0 0 AAAA + A
.p 5.; E
+E
.2.2 .$k 3.8 Jg!
-2 ;. ~
+
A
A
.g
a.; A
a g . 0
.z
+;
AA A A ~A A ~P
j j j $3
MARINE ENGINEERING
ELECTRIC PLANTS
$f$
444
33s
d m -
0 0 0
aaa
A48
.333 zzz
e
J 3 b,
a
3
4 a p
UUU p"
PPP
a aaaa a a aa a az,o,o _o _o 23 3
23.
8~
ccc
-g
2s2
8z
e
Y Q
aaa 332
8
g E CI
%
PlPlPlPlPlPl 333333 ddddl4l4
g h a m 3333 d dl4l4l4
fY%flf
I lllf "0 %%%% ea ea ea ea
CCCCCC
0 0 0 0 0 0
ea*aOa eaeaeae
a"k g . . . .
aaaaaa aaaaaa 'd '6'd 'd 'd'd
-[
UUUUUU
da ssssss 0
N
8 C
30g
0
0
0
0
0
$C$C
Ndndnd
'C 'd '6.6 'C 'C 222232 N N N N N N
u du uu 0
0
0
0
0
%
ss ss
a
2s
.z .;.a
-0
2-
3.2
2 22 22
u uuuu 0
In
2
0
0
0
0
Lciuamo
t."4N
.z .g.g.g.g N N N N
a aaza
u u uuu u
U U . U U U U uuu uuu u
s s
zs
0
q d
0
0
0
0
$$@
0
$i
0
0
0
0
0
0
%Zg$
0
0
0
ZZs 9% 5
0
$ ?'?$ $ d ddn 4
'? t.8
1'?q
-odd
n d m
e$$ A d d
00
.a
.>2
e.z
2 2 2g.8
% c
.a
. -m nn
.a.z.a.a .a.a.; %.az g
2 $ 2222 2s.2 $ 2 $
a
washdown or possible transient submergence may be encountered. 5 Explosion-proof, fan-cooled-used in an atmosphere containing an explosive mixture. This type of enclosure is required to be capable of withstanding an explosion of a specified gas or vapor which may occur within it while preventing the ignition of the gas or vapor surrounding the enclosure. Explosion-proof equipment is not watertight and must be enclosed in a watertight housing if it is subject to weather conditions. The cooling fan should be constructed of nonsparking material and be protected by a guard. 6 Submersible, self-ventilated-used for those limited and special applications that may require normal operation in air and emergency operation when submerged. A positive means of providing the required capability is the use of a bell type of enclosure. The motor and pump combination is vertically mounted and covered with a close-fitting bell that is open a t the bottom. The bell must be of sufficient depth so that the required submergence will not force water onto the motor windings. Under special and restricted conditions established by the U.S. Coast Guard, submersible motor-driven pump assemblies may be used for pumping out cargo tanks of liquefied methane, propane, ammonia, and other unusual cargos, the primary restriction being that air must be excluded from the cargo tanks at all times so as to prevent the possibility of an explosive mixture existing within the tanks. 7 Dripproof protected encapsulated, self-ventilatedused for those applications that may be subject to temporary submergence, splashing, spraying, or hosedown. Witb encapsulated motors the windings around the end coils and in the slots are completely encased in a protective insulating coating that permits exposure to specified liquids, and the bearing cavities are made watertight. This type of enclosure is suitable for many applications that heretofore have required total enclosure. b. Terminal Boxes. All motors are normally furnished with terminal boxes having threaded pipe taps for ship's cable entrance terminal tubes. Motor terminal leads and ship's cable are mated by means of cable connectors within the terminal box. The degree of enclosure required for terminal boxes is usually the same as that provided for the motor. The desired location of terminal boxes and the number and size of tapped holes for cable entrance are normally specified by the shipbuilder. c. Insulation. Insulating materials for use in motors are divided into categories according to their ability to withstand high temperatures for long periods of time. These categories are Class A, B, F, or H. Class A insulation has the capability of operating a t a maximum temperature of 105 C; Class B is rated at 130 C; Class F a t 155 C; and Class H a t 180 C. These temperatures in each case represent the insulation syqtem material capability and are the summation of the ambient temperature, motor-winding temperature rise above
625
ambient, plus an estimated temperature gradient referred to as the hot-spot allowance. The regulatory bodies have specific requirements regarding temperature rise limits for the various classes of insulation for different applications. Bearings and bearing lubricants should be selected based on operating temperatures encountered with each class of insulation. d. Special Treatment. All motors for use in marine applications should be given a special impregnation to make the winqiigs resistant to salt water, salt air, oil fumes, and fungus. Metal parts are made of corrosionresisting materials or are treated to render them corrosionresistant. For example, small hardware is usually zinc-plated, and the shaft inside the frame, exposed laminations, and brackets are usually treated to prevent corrosion. Small hardware for motors exposed to the weather is normally stainless steel. e. Ambient Temperature. Motors for machinery spaces are designed for an ambient temperature of 50 C, normally using Class B insulation. Motors for use in areas other than machinery spaces are designed for an ambient temperature of 40 C, normally using Class A or B insulation. Exceptions involve unusually hot areas, such as a t the tops of boilers or adjacent to smoke uptakes, and certain axial-flow exhaust fan applications for which the motors are rated for 65 C using Class F or H insulation. The trend is to use Class F insulation for both 40 C and 50 C spaces, to reduce the overall size of the motors. f. Space Heaters. Motors.subject to wide variations of temperature or excessive moisture conditions are often provided with space heaters to prevent moisture condensation in the motor when idle. The heaters may be resistance units bolted to the inside of the lower frame or a phase winding energized through a low-voltage transformer. I n either case, the heating circuit is electrically interlocked so as to remove heater power whenever the motor is energized. g. Shafts. For flexible and rigidly coupled drives, National Electric Manufacturers Association (NEMA) standard short-shaft extensions are used. However, in some instances shafts are provided with the end tapered, threaded, and equipped with a nut and washer for ease in disassembly and reassembly. Brake mqtors or motors using shoe brakes are provided with front-end shaft extensions as required by the brake application. NEMA long-shaft extensions are provided for pulley-driven auxiliaries. Special long shafts are provided for impellers of close-coupled pumps and axialflow fans. Carbon steel sha@ are normally provided for coupled drives and for freshwater closecoupled pumps. Shafts made of corrosion-resistant materials (e.g., stainless steel or monel) and fitted with sle$*s are generally required for pumps handling corrosive liquids. h. Bearings. With few exceptions a-c and d-c motors are equipped with greasable ball bearings; axialflow ventilation fan motors are usually equipped with prelubricated sealed ball bearings, since they are located in ductwork and not easily accessible. Sleeve bearings,
i II I
I
!
I
1
1
626
-
TORQUE (PER CENT OF FULL L O A ~ T O R Q U E ~ -
TORQUE (PER CENT OF F U L L - L O A ~TORQUE)
I
I
Fig.
ELECTRIC PLANTS
MARINE ENGINEERING
7 Torquespeed curves for induction motors rated 30 to 50 hp
designed for flood lubrication or forced lubrication, are used only in special applications. i. Speed Reduction Drives. Separately mounted coupled reduction gears or integrally mounted reduction gears are used to decrease or increase the speed obtained from the drive motor. Motors and gears must be carefully selected to suit the speed, torque, and duty of the driven auxiliary. A V-belt drive is often used for air and refrigerating compressors and slow-speed fans. The use of a V-belt drive requires the motor to be mounted on a base that affords adjustment of the belt tension. j. Mounting. Motors are designed for mounting in any required position, i.e., horizontal, inverted horizontal, vertical with the drive shaft up or down, and in some instances inclined. Most shipboard auxiliaries are driven by horieontal motors mounted on a common bedplate with the driven machine. However, the use of vertically mounted motor-driven centrifugal pumps provides a saving in deck area and a preferred piping arrangement. In addition to general-purpose mountings, NEMA has standardized two end mounts, types C and D, and two flange mounts, types P and PH. Types C and D are used either horizontally or vertically, with the relative location of the face and .feet fixed by the standards. Types P and PH flanges are used for vertical pump applications. Each type of mounting should be coordinated with the driven auxiliary to insure a satisfactory fit and performance. For heavy assemblies (such as an overhung motor with a disk brake attached by a flange), it is advisable to provide a foot-mounted motor to afford rigidity. 4.3 iElechical Characteristics of Alternating-Current Motors. The speed, horsepower, and duty rating of a
motor is fixed by the required input to, and the operating cycle of, the driven machine. Duty ratings. (operating cycle) for shipboard applications are classified as continuous duty or intermittent duty. Continuous duty is a requirement of service that demands operation a t a
Fig. 6
---
Design D Induction motor torque-speed curve for various slip values
substantially constant load for an indefinitely long time. Intermittent duty is a requirement of service that d e mands operation for alternate periods of (a) load and no load; (b) load and rest; or (c) load, no load, and rest, with such alternate intervals being definitely specified. The applicable IEEE Standard No. 45 contains a description of duty ratings for specific applications. The majority of the motors used aboard ship are squirrel cage induction motors. The design designations, Design A, B, C, and Dl for three-phase squirrel cage induction motors, are based on torque, current, and speed requirements. Each design offersdifferent torque, speed, and current characteristics to meet various operating requirements as may be seen from Figs. 7 and 8. Design A motors have a normal starting torque, high starting current, and low slip. This motor is not used for the usual shipboard applications because of its high starting current characteristic. Design B motors have a normal starting torque, low starting current, and low slip;. This is the motor most commonly used on shipboard; it is generally used for centrifugal pumps, fans, blowers, motor-generator sets, and compressors that are not loaded when started. Design C motors have a high starting torque, low starting current, and low slip. This motor is normally used for applications such as steering gear, anchor windlass, plunger-type pumps, and compressors that are not unloaded when started. Design D motors have a high starting torque, moderate starting current, and high slip; this motor is normally used for capstans, winches, valve operators, conveyors, elevators, and hoists. The squirrel cage motor may be designed for one, two, three, or four speeds. In addition, this motor is suitable for adjustablespeed operation when used with a combined frequency and voltage control; adjustable speed by voltage control only requires a specially designed highslip motor. Wound-rotor motors may be used for adjustablespeed
or constant-speed applications. For adjustablespeed applications, up to 50-percent speed reduction can be obtained by inserting different values of resistance in the rotor circuit by means of multistep controllers. Each set of resistance values inserted in the rotor circuit results in different torque and speed characteristics as shown in Fig. 9. The use of wound-rotor motors for constant-speed applications is limited on shipboard; however, its use should be considered when it is necessary to start a large motor from a relatively low-capacity generator. The advantage in this case would be the low starting current and normal starting torque that are available with proper selection of secondary resistances in the rotor circuit. Typical applications requiring adjustable-speed selection on shipboard are forced-draft blowers, fuel-oil service pumps, and main circulating pumps. Although wound-rotor motors may be used, multispeed squirrel cage motors are more commonly used for these applications. , Synchronous motors are used in shore practice for improving the system power factor and are usually applied to drive large continuous loads such as motorgenerator sets, compressors, pumps, and fans. They have practically no application on shipboard but may be used for such applications as large-capacity circulating pumps and motor-generator sets to improve the system power factor.
2EROEXTERNAL RESISTANCE 2-4 CENT SLIP / PER AT FULL LOAD
Fig. 9
Typical torque-speed curves for a wound-rotor motor with various external resistances
loads with lower armature current and better commutation than with a shunt winding only. The compound winding is desirable for loads of high inertia such as certain direct-connected centrifugal fans, as well as for such shipboard applications as the propeller shaft 4.4 Electrical Characteristics of Direct-Current Motors. turning gear, valve operators, compressors without The major distinctive characteristic of d-c motors is unloaders, and positive-displacement pumps. The crane or winch type of winding is a compound their type of field windings. Direct-current motors may winding consisting of a light shunt field and a heavy have a shunt winding, stabilized-shunt winding, series series field. These motors are deigned for a specific winding, cornPofind winding, or a crane or winch type of load and afford many desirable characteristics such EM winding. Shunt or stabilized-shunt windings are used for high torque for heavy loads and high speed for light applications requiring constant speed rkgardless of load loads with the light shunt field providing constant-speed variation. Typical shipboard applications are fans, characteristics. Typical shipboard applications of this type of winding are cargo winches and anchor windlasses. blowers, centrifugal pumps, and elevators. Motors with stabilized-shunt windings are provided 4.5 Mechanical Characteristics of Motor Control with a light series field in addition to the shunt field to Equipment prevent a rise in speed as the load increases; hence the a. Starter and Controller Panels. A grouping of term "stabilized". This type of winding is desirable for loads of high inertia such as direct-connected centrifugal several motor starters housed in a free-standing deckfans or pumps, since acceleration is accomplished with a mounted structure is known as a group control or motor lower armature current and better commutation than control center. Each motor starter within a group control is energized from a common power supply feeder with a common shunt winding. Series-winding motors are used for loads requiring a through individual circuif, breakers mounted in each very high starting torque. They are also used in starter. Group controls &-metal-enclosed unjts having applications requiring operation over a wide speed range metal barriers between starters. In general, individual starter enclosures are either such that the motor develops a high-speed operation at light loads, and at low speeds a comparatively light dripproof, watertight, submersible, or explosion-proof as current- and a high torque. Series motors are partic- required by their location. Starters mounted in group ularly suitable for the operation of warping capstans control switchboards are generally of the "open" type which require a high torque a t low speeds for handling mounted in a dripproof enclosure. Cabinet enclosures are usually designed for bulkhead loads and a high speed at light load for retrieving lines. The friction loss of the machinery is generally sufficient mounting in smaller sizes and for deck mounting in weights over 150 lb. Consideration should be givenib , to limit the no-load speed to a safe value. Motors with compound windings develop a high the need for rear access on large starter panels for starting torque and have the ability to handle peak connections and inspection.
1 628
MARINE ENGINEERING
ELECTRIC PLANTS
i
Suitable provisions for cable entrance are made in all cabinets, usually a t the top for the feeder and at the bottom for the motor and control leads. Watertight terminal tubes or other adequate seals to exclude moisture should be provided where cables enter at the top of dripproof cabinets. Cable clamps or other adequate means for holding cables in place to prevent chaffing from vibration should be provided where cables enter the bottom of dripproof cabinets. All cable entrances in watertight starters are made through terminal tubes. In general, practically all motor starter enclosures, excepting explosion-proof starters, are provided with removable drilling plates so as to avoid the danger of metal chips dropping in the working parts as could occur if drilled after delivery. Explosion-proof enclosures are often purchased with the number and size of tapped holes for cable entrance terminal tubes specified. Separately mounted resistors, if required, should be installed in well-ventilated spaces and provided with protection from dripping liquids. The ambient temperature and class of insulation for motor control equipment are determined in the same manner as previously described for motors. b. Master Switches. The broad category of master switches includes pushbutton, drum switches, selector switches, pressure switches, temperature switches, interloclr sw~itches,float switches, and any other type of remote pilot device required for proper operation of the controlled motor through starter equipment. Local switches and indicating lights which form a part of controllers should have the same degree of enclosure as the controller. For remote locations, the enclosure of switches should be dripproof, watertight, submersible, or explosion-proof to suit the requirements of the location. In general, for weather or corrosive areas, watertight enclosures of cast bronze or brass for small switches and bronze, nodular iron, or stainless steel for winch control switches are provided. The cable entrance into master switches should be through a bushing in the bottom of dripproof switches and through watertight terminal tubes in all others. The regulatory body rules generally require that controllers be mounted adjacent to the driven auxiliary. With group control and in those situations where the controller cannot be mounted in sight (as for certain fans and forced-draft blowers), remote pushbuttons, designed so that the stop button $an be maintained open, are located a t the driven auxiliary. c. Speed-Regulating Rheostats. For shipboard use, speed-regulating rheostats for shunt field control of direct-current motors are provided with dripproof covers and enclosed terminals and are arranged for bulkhead mounting. This construction permits a convenient location of the rheostats with a minimum probability of damage to the rheostat or injuries to personnel. 4.6 Electrical Characteristics of Alternating-Current Motor Controllers., Motor controllers are designed to
perform definite electrical functions regarding the control and protection of motors. The characteristics and
applications of the various types of controllers that provide these functions are as follows: a. Controller Operation. Controllers are designated as being either manual or magnetic in operation: Manual controllers are normally used for applications of less than 2 hp that require only "on-off" operation. Magnetic controllers are used for all other applications and may be classified as automatic or nonautomatic. Automatic controllers start and stop the motor in response to some controlled factor with no attention from an operator. Nonautomatic controllers require manual operation of a pushbutton or switch to initiate a start or stop. After the initial manual operation, the controller completes the starting or stopping of the motor. Nonautomatic controllers are used for practically all applications other than those auxiliaries that require automatic cycling such as air compressors and refrigeration compressors. Typical control-circuits for a-c magnetic controllers are shown in Figs. 10 and 11. b. Types of Controllers. Controllers of the acrossline type are used for practically all shipboard auxiliaries since the ship's generating plants are usually of adequate capacity to handle the starting currents of all motors installed. However, controllers of the autotransformer type are used when it is necessary to limit the starting current of a motor so as to avoid imposing an excessive load on the generating plant. This type of controller should be designed for closed circuit transition so as to avoid high transition currents. Standard starting transformers for motors above 50 hp have taps of 50, 65, and 80 percent of full voltage; only 65- and 80-percent taps are provided in sizes below 50 hp. The starting current drawn from the line is proportional to the square of the percent voltage tap; i.e., 80 percent tap equals 64 percent of the across-line starting current. A typical application is for fire pumps that must be started from relatively small-capacity emergency generators. A primary resistor type of controller could be used to limit the starting currents of large motors and also for speed control of small motors; however, their application is very limited because the motor starting current is not substantially reduced and therefore they provide little or no advantage for shipboard use. Controllers of the secondary resistor type are used to limit the starting currents and provide speed control for wound-rotor induction motors; typical applications are forced-draft blowers and main circulating pumps. A wye-delta type of controller could be used to limit the starting currents of large motors. With this arrangement the motor is started in the wye connection and then reconnected, with closed transition, to the delta running connection. This arrangement requires a six-lead motor and the starting current would be approximately 33 percent of the across-line starting current. A typical application is a bow thruster. See Fig. 12 for typical methods of starting a-c induction motors. c. Controller Protective Features. Low Voltage
4
I
4
1 t
f I
1
-
,I
Protection (LVP) is the feature that is provided to cause the controller to disconnect the motor from the power supply upon reduction or loss of voltage; the motor remains disconnected until the voltage is restored and the motor is restarted by manual operation of its starting pushbutton, as shown in Fig. 13. This feature is provided as a means of preventing the simultaneous restarting (after an interruption of the supply voltage) of a large number of motors such that their large starting currents are additive. Low' Voltage Release (LVR) is the feature that is provided to cause the controller to disconnect the motor from the power supply upon a reduction or loss of voltage; the motor remains disconnected until the voltage returns, .and then automatically reconnects the motor to the power supply to restart it, as shown in Fig. 14. This feature is usually applied to only those vital auxiliaries that must be automatically restarted immediately upon restoration of power. Typical examples are lube-oil service pumps, main and auxiliary condensate pumps, main circulating pumps, and control air compressors. It is usually desirable to use timedelay relays with LVR controllers to obtain staggered starting and prevent simultaneous restarting of all LVR auxiliaries. Motors that are automatically controlled by pressure-switches, and similar devices, have an inherent LVR feature unless a low-voltage relay, which opens upon failure of line voltage, is provided in the control circuit. All manual-type controllers provide LVR characteristics. Overload protection is the feature that results in the controller operating to disconnect the motor from the power source when excessive currents (not short circuits) occur that could cause overheating of the motor. This feature is provided by overload relays; separate relays are required for each winding of multispeed motors. Overload relays may be either thermal or magnetic. Thermal overload relays generally consist of a heatsensitive element and a heat-generating 'element. The heat-generating element may be a heater or coil in series with the motor load circuit. An excessive motor current passing through the heat-generating element causes the heat-sensitive element to react to open the overload relay contacts, thus breaking the circuit to the operating coil of the main line contactors; this in turn causes the contactor to open the motor circuit. Since the tripping characteristics of the thermal overload relay depend on both the length of time of application and the amount of overload current, the relay can be, and normally is, designed to follow approximately the timecurrent heating curve of the motor. This curve represents the values of the current that a motor can carry for different lengths of time without damaging the motor insulation. Thermal overload relays should be compensated against possible ambient temperature changes. Thermal overload relays are generally of the solder-pot, bimetal, single metal, or induction type. Practically all applications of overlaad relays on shipboard are of the thermal type.
Magnetic overload relays generally consist of a coil in series with the motor load circuit and a tripping armature or plunger. When the amount of overload current for which the relay is set passes through the series coil, the tripping armature is actuated to open the overload relay contacts, thus breaking the circuit to the operating coil of the main line contactors; Bhis in turn causes the contactor to open the motor circuit. Magnetic overload relays are not affected by variations in the ambient temperature and &quire no temperature compensation. Magnetic overlo*ad relays are of the instantaneous or time-delay type and have limited application on shipboard because they do not use heat in their operation and consequently do not follow the heating curve of motors. Overload relays are provided with a means of resetting so that the motor controlled can be restarted with overload protection. Tripped thermal overload relays must be allowed to cool before the tripping mechanism can be reset. Magnetic overload relays can be reset immediately after tripping. The three forms of overload relay resets are manual (hand), automatic, and electric. The manual form is the most common for shipboard use and consists of a rod or lever which, when operated, causes the tripping mechanism to be returned to its original position. The automatic reset has no practical application on ships. The electrical reset is limited to those applications where it is desirable to reset a relay from a remote operating position. Emergency run features should be provided on controllers for certain auxiliaries, such as elevators, in which case stopping in the middle of an operating cycle could be highly undesirable. This feature, which is initiated by operating a pushbutton or lever, renders the overload relay tripping mechanism inoperative so that the auxiliary can be operated with the motor running in an overload condition until the operating cycle is completed. In general, each control wire that leaves a controller should be provided with short-circuit protection. Such protection may be provided by a fuse, located in the controller, if the lead is not already protected by a current-limiting device (coil or resistor) located i11 the enclosure. When a secondary source of power greater than 24 volts is brought into a motor controller for alarm, indicating light, or other circuits, a suitable interlock is usually provided to disconnect the secondary source upon opening of the controller door. In lieu of disconnecting the secondary source as noted in the foregoing, an independent disconnect deyice may be used for this purpose. This independent CP-iconnectihould be-located adjacent to the motor and controller disconnect, and a sign should be provided on the main disconnect to warn that both devices should be operated to disconnect completely the motor and controller. 4.7 Electrical Characteristics of Direct-Current Motor StaHers. Direct-current motor controllers may be of
either the across-line or resistor type. Full-voltage across-line controllers are usually manually operated and
MARINE ENGINEERING
630 LINES L2
LI
ELECTRIC PLANTS LlNES L t L21 L31 TDS TDS TDS
L3 &
1
& A - - 4--ADS
I
H, :I: a
ELEMENTARY DIAGRAM II M OL II
, , OL N.
TI
N
T2
MOTOR
I
H2
CR 26
L(LA 80
*
PS M
(
UNLOADER SOLENOID
(L) LOCAL (R) REMOTE
DESCRIPTION OF OPERATION
ELECTOR IN "HA ": TO START. PRESS A START BUTTON ENERGIZING SELF MAINTAINING RELAY "cR': THE "CR" E o N T A c T s CLosEYNEasIzING MAIN CONTACTOR "M? THE * ~ ~ ~ c o r y CONNECT A c n THE MOTOR ACROSS THE LINE AND ENERGIZE.7HE UNLOADER SOLENOID. TO STOP PRESS A STOP BUTTON REMOVING OCR" "M"AND THE MOTOR FROM THE LINE. SELECTOR AUTO^: PLACING THE SELECTOR s w ~ i c n IN AUTO^ INSERTS A N.B. PRESSURE SWITCH, *PS*: IN
~k
Fig. 11 Typical elementary and wiring d l a g r m for a potable water pump application
...- --.. ...----,..
THC r n N T D n 1 I CD
TO START PRESS A START BUTTON ENERGIZING RELAY "CR". WHEN THE "PS~CONTACTSCLOSE, MAIN CONTACTOR "M" IS ENERGIZED CONNECTING THE MOTOR ACROSS THE LINE. WHEN THE n ~ "Ma'IS DROPPED REMOVING THE MOTOR FROM TUC I INC TO STOP. PRESS A STOP BUTTON R E I I N AVOLTAGE FAILURE CONDITION, A L L vv~la PRESS A START BUTTON AFTER RESTORATwn vr r v ~ ~ r r utc~. v n IN AN OVERLOAD CONDITION, THE 0.L.CONTACTS OPEN REMOVING THE RESTART, PRESS A RESET AND THEN A START BUTTON. (OVERLOAD PROTECTIOIY
~
w
~
~
~
Ra. 10 Tvdcal e l e m ~ t aand ~
I.
~
~
~
~
~
~
~
~
TO START THE MOTOR MANUALLY, TURN THE SELECTOR SWITCH TO THE MAN POSITION. THIS ESTABLISHES A CIRCUIT TO THE MAIN CONTACTOR M. THE MOTOR WlLL START AND CONTINUE TO RUN AS LONG AS THE SELECTOR SWITCH IS IN THE MAN POSITION. TOOPERATE THE MOTOR AUTOMATICALLY UNDER THE CONTROL OF THE PRESSURE SWITCH PS. TURN E SWITCH W ~ L LKENE ESTABLISH A-~IREUIT THE SELECTOR-SWI~CHTO THE AUTO POSITYON.Y ~ PR&RE TO THE MAIN CONTACTOR M. THE MOTOR WlLL START AND CONTINUE TO RUN UNTIL THE PRESSURE SWITCH CONTACTS OPEN. THE MOTOR WlLL THEN CYCLE OFF AND ON AS THE PRESSURE SWITCH CONTACTS OPEN AND CLOSE. AVOLTAGE FAILURE WlLL CAUSE THE MAIN CONTACTOR M TO OPEN AND DISCONNECT THE MOTOR FROM THE LINE. WHEN VOLTAGE IS RESTORED, M WILL RECLOSE AND START THE MOTOR (LOW VOLTAGE RELEASE). AN OVERLOAD WILL CAUSE THE OL CONTACTS TO OPEN, DEENERGIZING M AND STOPPING THE MOTOR. TO RESTART. PRESS THE RESET BUTTON. THE MOTOR MAY B E STOPPED AT ANY TIME BY TU8NING THE SELECTOR SWITCH TO THE STOP POSITION. ~
LINES
,
REAR VIEW OF DOOR
FRONT VIEW OF PANEL
ITO
LINES REAR VlEW OF DOOR
, 1-
FRONT VlEW OF PANEL
limited to those applications involving a maximum of 2 hp subject to the driven machinery withstanding the resulting high starting torque and the motor handling the high inrush current. Good practice restricts this type of starting to small galley appliances, small pumps, and ventilation fans. Controllers of the resistor type are used for most d-c motor control apylicatibns. They provide mean8 of
reducing the motor starting current to prevent motor damage and prevent overloading the generating plant. The resistor type of controllers may be either manual or magnetic. Most controllers of the manual d-c resistor type are similar to acros9-line controllers except for the addition of a starting resistor and an accelerating contactor. The control circuitry is designed such that as the motor comes up to speed the accelerating con-
tactor is actuated to by-pass the starting resistance and connect the motor to full voltage. Manually operated "dial" or "face plate" starting rheostats may be used for starting and speed regulation of nonreversing motors. They are used for light-starting auxiliaries such as ventilation fans but should be limited to ratings below 2 hp. Practically all d-c controllers are of the magnetic, reflistor type. Magnetic controllers may be of either the nonautomatic or automatic type and consist of a main contactor, control relays as required, one or more accelerating contactors, an overload relay, emergency run feature, and starting resistance. The main contactar is controlled by a manually operated master
switch (nonautomatic control) or by an automatically operated device such as a pressure-regulating switch (automatic control). When the main contactor closes, the motor is connected to the power supply through the starting resistances. The p n t r o l ciquitry contains current-sensitive series relays and accelerating contactors so that as the motor comes up to speed, individual steps of the starting resistance are by-passed until all steps are out of the motor circuit and the motor is connected across the line. Another method of control starting uses definite time accelerating contactors to close automatically and shunt out steps of the starting resistance. Speed adjustment for d-c motors may be obtained by
MARINE ENGINEERING
ELECTRIC PLANTS
REDUCED-VOLTAGE AUTO TRANSFORMER
LINES
I
633
ELEMENTARY DIAGRAM
START
DESCRIPTION OF OPERATION
Fig. 13 Typical motm canhol circuit with low-voltage prohctim feature
(01 ACROSS LINES
TO S V R T THE MOTOR PRESS TYE START BUTTON. THIS ENERGIZES CONTACTOR M. CONNECTING THE MOTOR ACROSS THE LINE. M MAINTAINS ITSELF THRU ITS OWN AUXILIARY CONTACTS. TO STOP THE MOTOR, PRESS THE STOP BUTTON. A VOLTAGE FAILURE WlLL CAUSE M TO OPEN STOPPING THE MOTOR. TO RESTAkT, WHEN VOLTAGE IS RESTORED, PRESS THE START BUTTON. (LOW VOLTAGE PROTECTION). AN OVERLOAD WlLL CAUSE THE OL CONTACTS TO OPEN CAUSING M TO OPEN STOPPING THEMOTOR. TORESTART.PRESSTHERESETBUTTONANDTHENTHESTARTBUTT0N.
(b) CLOSED TRANSITION LINES
SERIES
CLOSED- CIRCUIT TRANSITION
Fig. 12 Typical mehds of starting a-c induction moton
IM WOUND-ROTOR INDUCTION MOTOR
IA a Z A ARE DELAYED IN CLOSING BY A PRESET TIME
R I CONTACTOR SEO.
(C) REDUCED-VOLTAGE
PRIMARY RESISTOR
MOTOR
(dl FULL VOLTAGE SECONDARY RESISTOR
using a rheostat in the motor shunt field circuit or by varying the amount of resistance of continuous-duty resistors in the motor armature circuit. 4.8 Brakes. Electric brakes for marine service are of either the disk or shoe type, each type being spring set and magnetically (solenoid) released. Solenoids or magnets are usually energized (brakes released) through contactors whenever the associated motor is in opera-
tion. Deenergizing the motor also deenergizes and engages the brake. Disk brakes may be either a-c or d-c operated; shoe brakes are usually d-c operated. For motors that are 50 hp and larger, d-c operated brakes are usually provided. I n general, brakes for suspended loads are rated a t 200 percent of the motor torque. For other types of loads,
the brakes are rated to stop the load under any operating condition involved; in these cases, the brakes are usually rated a t 100 percent of the motor torque. Brakes located in weather locations should be of a watertight construction and have electric heaters to prevent the accumulation of condensation during nonenergized periods. Brakes should be provided with a means of being mechanically released for emergency operation in the event of a power failure. Disk brakes are attached directly to the motor fronb end bracket and require no special foundation; from a space point of view, disk brakes lend themselves to an
economical installation. Most shoe brakes are foob mounted and require a special foundation for proper alignment with their motor. 4.9 Transformers. Transformers _ are utilied to supply alternating-currefif* loads which cannot be operated on the ship's primary voltage. Transfortners are normally of a single-phase, 60-cycle, air-cooled, dry type that is designed for continuous duty with dripproof enclosures and suitable for connecting in a threephase bank. Each transformer in a bank should have an identical rating with the primaries connected in delta and the secondaries connected in delta or wye as re. quired for the intended service.
634
MARINE ENGINEERING LINES LI
L2
635
ELECTRIC PLANTS
ELEMENTARY DIAGRAM
lead-acid or alkaline type. It is advisable that all storage batteries provided on a given vessel be of the same type, to prevent the possibility of contaminating any battery electrolyte through the inadvertent use of a common hydrometer. Principal applications for storage batteries are: (a) emergency power for radios; (b) no-break power supply, (c) diesel-generator cranking; (d) lifeboat engine starting;
L3
IDS IDS IDS MOTOR OL
(e) emergency lighting when an emergency diesel generator is not installed; (f) fire-screen doors and watertight doors; (g) forklift trucks; (h) general alarm system; and (i) certain interior communication loads. The regulatory bodies h a w published specific detailed requirements regarding the construction, rating (capacity), ventilation, installation, and arrangement of i storage batteries.
OFF
-RUN
I
Section 5 Lighting Fixtuns and Equipment
2
DESCRIPTION OF OPERATION TO START THE MOTOR POSITION THE SELECTOR SWITCH &T"RUN". THIS ENERGIZES CONTACTOR M CONNECTING THE MOTOR ACROSS THE LINE. TO STOP THE MOTOR POSITION THE SELECTOR SWITCH AT"OFF~. A VOLTAGE FAILURE WILLFAUSE M TO OPEN STOPPING THE MOTOR. WHEN VOLTAGE IS RESTORED, THE MOTOR WILL IMMEDIATELY RESTART (LOW VOLTAGE RELEASE). AN OVERLOAD WILL CAUSE THE OL CbNTACTS TO OPEN CAUSING M TO OPEN STOPPING THE MOTOR. TO RESTART, PRESS THE RESET BUTTON.
LINES REAR VIEW OF DOOR
Li L21 T" T T
VIEW OF PAWL
L3f
I
.
,
Fig. 14 Typical motor m t r d cirwl with law-valtage please feature
4
I t
I I
RESET BUTTON PUSH TO RESET
u
MOTOR
06
'
The kva ratings of transformers should be based the connected load plus a reasonable allowance for installed spares and future development. Every effort should be made to balance the load between phases of transformer banks. 4.10 Motor Generators. Motor generators are provided to supply power to loads requiring special voltages and/or frequencies that differ from the ship's primary power. The principal applications for motor-generator sets are: (a) Providing d-c power for cargo handling equipment; (b) Providing a-c power for the "temporary emergency" loads on passenger vessels; in this applic* tion, upon loss of normal power, the motor-generator set
is supplied with power from an emergency battery and in turn provides a-c power to selected vital emergency loads until the emergency diesel-generator can start and assume all emergency loads; and (c) Providing power for automatic elevators on passenger vessels. 4.1 1 Reetiflers. Rectifiers are provided to supply loads requiring d-c power that is not available from the ship's primary power. Principal applications for rectifiers are: (a) Electronic equipment requiring d-c power; (b) Battery charging; and (c) Fire-screen doors. Specific details regarding requirements for rectifiers are contained in IEEE Standard No. 45. 4.12 Batteries. Storage batteries are usually of a
i
5.1 General. Marine fixtures and appliances must be of a special design and construction to suit the various requirements incidental to shipboard installations. They must be rugged to withstand normal shipboard vibration. Furthermore, enclosures must be compatible with the particular environment in which they are located; also, they must be of a corrosionresisting material or have an effective corrosion-resisting finish. To prevent rapid deterioration of the finish, decorative fixtures with polished or plated surfaces must be protected from the salt atmosphere by some form of protective coating, such as lacquer. The possibility of radio interference from fluorescent or high-intensity electric-discharge lamps also requires consideration in the selection of light sources. All lighting fixtures, with few exceptions, are required to meet the standards of the Underwriters Laboratories, Inc., publication UL595, Marine Type Electric Lighting Fixtures. Fixtures not covered by this standard must have U.S. Coast Guard approval for each specific application and vessel. In developing the design of the lighting system, the "Recommended Practice for Marine Lighting, " published by the Illuminating Engineering Society, should be used for determining the minimum footcandle requirements. 5.2 Types of Fixtures. Fixtures may be classified by the types of lamps used. Fluorescent, incandescent, and high-intensity electric-discharge lamps are the main types of lamps used. Of the three, the fluorescent offers the most advantages and is replacing the incandescent for practically all general illumination requirements. The color rendition of fluorescent lamps has been steadily improved in recent years and is now almost the equal of the incandescent lamp in this respect. Their high luminous &cacy makes them much better suited for use in air-conditioned spaces due to lower heat generation; and their longer life considerably reduces maintenance. The incandescent lamp provides a compact, highbrightness source of light which can easily be directed by a small, simple luminaire. This feature makes the incandeacent lamp superior for spot and detail illurnin* tion. It operates readily on either &C or d-c supplies. It is less sensitive to ambient temperatures than other light sources and, consequently, is preferred where
illumination is required in severe hot or cold temperatures, i.e., refrigerated spaces, weather locations, detail lighting for boiler gages, etc. Incandescent lamps are the most practical for minimal illumination requirements, such as small locker rooms and indicator lights. The high-intensity electric-discharge lamps, in general, are compact, with high brightness sources and have an even higher luminous efficacy and a longer l i e than the fluorescent. %lost of these lamps have an acceptable color of light-----.__ W w o r k areas, such-aScZF6 KoIds,"aeEk -' fEGIh$tTng, and engine~ooni~;but-arenot consiaered-~ ~ e ~ T o ~ 5 E 5 a T l J m m m - ~ ~ M color -rendition. .Their opertttion 'and maintenance ecGmYS-eGen more pronounced than for the fluore* . cent. Their long lamp -life, ranmgto 24;MlU-%r and 45i%t,er1 makes them an ideal lamp for fixtures mounted high on masts and kingposts and other locations not easily accessible. The problem of relamping is reduced such that the maintenance for this type of fixture is practically negligible. Of the high-intensity electricdischarge lamps, the mercury-vapor is most acceptable for shipboard use. For certain applications, supplementary instantaneous lighting may be necessary since, in the event of power failure, instantaneous relighting is not accomplished by electric-discharge lamps. The use of some of these lamps is limited due to certain hazardous characteristics (lamps containing sodium may cause a fire when broken on a wet surface). Serious corrosion condition8 may also result if the mercury from a ruptured mercury-vapor lamp comes into contact with aluminum. Even with these limitations, however, the mercury-vapor lamp is finding its rightful place on shipboard. Lighting fixtures may glso be identified according to their application; i.e., cedi% lights (located overhead in ceiled spaces), deck fixtures (located overhead in unceiled spaces), bulkhead Mures, detail lights (desk, berth, mirror, etc.), floodlights, navigational lights, and I miscellaneous lighting fixture4 --'%Sa*-faclr baiic types of fixture enclosures for shipboard use: watertight, dripproof, explosion-proof and nonwatertighi--Waterti@itTiiiiipment is installd b r d - ~ p t t . ~ e ~ Wwould i ~ i tbe exposed to seas, splashing, or severe moisture conditions, and for all installations in the weather. Dripproof equipment is installed in the % _
MARINE ENGINEERING
overhead in other wet or damp locations. Explosionproof equipment is installed in all areas that are subject to volatile gas accumulations; this equipment is so constructed that an explosion of a specified vapor within the equipment will not cause the ignition of the volatile atmosphere surrounding it. Nonwatertight equipment is used where moisture and volatile gases are not problems. 5.3 Passenger Stateroom Lighting. General illumination is provided for passenger staterooms by direct or indirect lighting supplemented by b m l i g h t s , mirror lights, table lamps, desk lights, bracket lamps, etc., to s G t 3 h e =gement %fTGmiture and decorative scheme. The trend is to the use of fluorescent coves, which may be combined with curtain valance lighting and incandescent down lights. A berth light, either separate or incorporated in a table lamp, is provided a t the head of each berth or bed. It should provide adequate light for reading but avoid annoyance to occupants of other berths in the room. Illumination for grooming at the dresser may be provided by bracket lamps on each side of the mirror, table lamps on each end of dressers, or by a mirror light recessed in the ceiling. These lights should be located or arranged for best illumination of the person seated at the dresser. I n bathrooms, general illumination is provided by one or more ceiling fixtures, with one or more mirror lights provided for each mirror. In small baths, where the mirror light provides the required level of general illumination, the ceiling lights are omitted. A receptacle is provided at each dressing table and a t each lavatory for the convenience of the passengers. 5.4 Officers' and Crew Quarters. Ceiling fixtures in the messrooms, lounges, recreation rooms, offices, and staterooms of the officersand crew usually are fluorescent with diffusing lenses. Berth, lavatory, table lamps, and desk lights are provided to suit furniture arrangements. Receptacles are installed at the lavatories, desks, and also for bracket fans when required. The toilet and shower spaces are illuminated by a ceiling light and a mirror light at the toilet case. I n small spaces where the mirror lights will provide the required illumination, the ceiling lights are omitted. 5.5 Passage Lighting. I n ceiled passageways, fixtures are usually fluorescent of the cornice type with concealed wiring. In unceiled passages, a similar type of fixture located on the bulkhead below interferences and angled to illuminate the center of the passage should be used. Passage lights are spaced from 7 to 9 f t apart and a light iii-IoCded'at the-interie6tlrdn+ofathwar2ship and fore-and-aft passages. Receptacles spaced about 40 f t are provided in passages for the attachment of vacuum cleaners and floor polishers. 5.6 Public Space Lighting. Architects and interior decorators generally design the passenger spaces, including the selection of furniture and fittings. They also determine the general character of the lighting installation and the type of fixtures to be used. The electrical
designer is responsible for the detail $esign of fixtures, as desired by the decorator, in accordance with sound engineering principles and for their suitability for marine installation with respect to rigidity, accessibility for servicing, ample ventilation, and adequate intensity of illumination. For illumination of the lounges, smoking room, dining saloon, and similar spaces, including adjacent foyers and passages, established practice favors simplicity with dependence on murals for decoration and with totally indirect trough or cove lighting arranged to provide adequate general illumination and to accentuate the decorative features. If direct illumination is provided, the detail design is determined by the fixture manufacturer's specialists in interior decoration, but all fixtures must be checked carefully by the electrical designer for compliance with marine practice. Special fixtures& public spaces include desk lights, table lamps, and floor lamps which are usually provided with built-in switches. Illuminated signs are provided on each deck a t passenger stairways and elevators for identification of public spaces, and for guidance of passengers to the debarkation decks in event of an emergency. Illuminated signs also are provided for all public toilets. When a toilet entrance is located in an athwartship passage, an additional sign is installed at each end in the main fore and-aft passages. Receptacles are provided, when desired, in floors or baseboards for lamps and vacuum cleaners. Depending on conditions, floor receptacles may be of the shallow type with the box for the fitting and conduit for wiring imbedded in the plastic floor covering, or they may be of the deep type with the box extending through the deck for wiring below. 5.7 Commissary Space Lighting. Careful consideration should be given to the illumination required for the preparation and inspection of food. General illumination for galleys, pantries, and service areas is provided by overhead flush-mounted dripproof fluorescent fixtures. Incandescent lights are installed under the exhaust hoods for the illumination of the ranges, griddles, etc. These fixtures must be constructed to withstand the intense heat from the range tops. Refrigerated spaces are lighted by guarded watertight deck fixtures controlled by a switch located within the space near the door. A fixture with a red globe, wired in parallel with the refrigerated space lights, is installed outside the door to show when the lights inside the compartment are "on". 5.8 Hospital Lighting. General and detail illumination for the hospital space is installed similar to a crew stateroom. For vessels having an operating room, a special operating light (or lights) is provided over the operating table; and all electrical equipment installed less than five feet above the deck must be explosion-proof. The Coast Guard rules should be reviewed to determine additional requirements as may be applicable to operating room installatibns. 5.9 Workshop Lighting. Illumination should be
637
ELECTRIC PLANTS
concentrated over the workbenches, machines or other areas where close visual tasks are performed. The lighting fixtures should be so located that the work area is not shadowed by the workman. Incandescent machine tool lights are also provided on lathes, drills, etc. Where fluorescent lighting is used in workshops having rotating machinery, altkrnate fixtures should be connected to different phases to prevent a stroboscopic effect. 5.10 MachinerySpace Lighting. The general illumination in machinery spaces is usually supplied by multilamp fluorescent deck fixtures utilizing 20- or 40-watt lamps controlled at the distribution panel. The location and spacing is generally dictated by interferences with pipes, vents, and other obstructions. Fixtures should be mounted as high as practicable to give good coverage, but they must be accessible for cleaning and replacement of lamps. Where necessary to avoid interference or shadows, the fixtures are stooled down from the deck. The fixture supports must be designed to avoid vibration and must be rigidly braced. The general illumination is supplemented by&--)-fi$t guarded watertight fixtures located as required by the arrangement of the space. This supplementary illumination is particularly necessary in the bilges and for gages, oil sight-flow glasses, and similar fittings. Switchboard illumination must be designed carefully to provide adequate light for the reading of instruments without reflection or glare. A continuous inverted a trough with a line or lines of fluorescent lamps frosted glass cover provides ideal illumination. Auxiliary machinery spaces, such as compartments for the steering gear Or generally are lighted guarded waterti%ht incandescent fixtures' fluorescent fixtures are used if the spaces are heated. Watertight receptacles, or combination switches and receptacles, are provided in main and a~xiliah'mach3inery spaces for the attachment of portable lights and tools. I n the efigine room, receptacles are spaced about 40 to 50 f t apart, and in the smaller spaces at least one double receptacle is installed. 5.1 1 Cargo Hold Lighting. There are two concepts for cargo hold lighting. I n one the illumination is provided by portable "cargo clusters' only; each cluster consists of a fitting accommodating from one to eight incandescent lamps, installed in a plastic or steel bowl reflector, and protected by a wiremesh guard. Receptacles for the attachment of these portable fixtures are installed a t each cargo hatch. The second, and more prevalent, concept is to have permanently installed fixtures for general illumination supplemented by cargo clusters as discussed in the first concept. Where lighting fixtures are installed on the deck over for general illumination, they should be installed between the deck beams or adjacent to the deep beam surrounding the hatches, where they are more readily protected from damage. By providing the fixtureg with resilient mounts, they are further protected from possible damage due to shock from heavy cargo dropped on the
deck above, from shifting cargo caused by heavy seas, and from rough cargo handling. To reduce theft losses, the fixtures in cargo holds and other spaces accessible to casual laborers should be protected by guards requiring special keys or wrenches for removal, and receptacfe caps should be of the hinged type. Watertight q u b l e receptacles, usually two at each hatch, are provided for the connection of portable lights for use in cargo holds and on the deck in the vicinity of cargo hatches.
5.1 2 Weather Deck and Caw0 Handling Lighting*
Fixtures for open decks are required to be of watertight c o n s t ~ ~ t i o nOn . Passenger promenade decks, they are U S U ~ ~ deck ~ Y fixtures ~ n s i s t i n g of a screwed type holophane or a frosted globe and a cast bronze body without a guard- Else\vhere, they may be standard guarded, watertight deck or b u l k h e a d - t ~ ~fixtures. e Exterior lights visible from ahead should be shielded. Lights on the navigating bridge deck and in other areas which cause interference to navigation should be controlled by switches located in the wheelhouse. Portable floodlights, with local control, are provided PO& and starboard to illuminate the gan&cwaYs. They should be mounted on brackets that are so constructed that they may be s w n g outboard and locked in place for overside lighting. -lighting of the oa'g0,-:handling gear and the sumunding deck, watertight floodlights, usu"~ of 309- to 500-watt rating, are permanently mounted high o ~ m ~ O r - E n g p o s t sThe . lights and particularly the mounting supports must be of rugged design to suit the exposed location. The lights are ordinarily not be set and secured in the position adjustable and which directs the beam to best advantage. 0, boat decks, lights are provided to facilitate the loading and lowering of lifeboats. These should be wide-beam 500-watt incandescent watertight floodlights located on the rail or on a vertical pipe stanchion between each pair of adjacent boats. ~h~ mounting arrangement should permit turning the floodlight inboard for deck illumination or outboard and vertically downward to light the water alongside. A quick-release clamp is pmvide&to secure the light in any position. For identification purposes, the owners usually require floodlights for illumination of the smokestack insignia on each side. Vessels on regular passenger runs may have the name boards lighted for identification and for the . convenience of the public&The trend is to the use of electric-discharge mercuryvapor floodlights for all open-deck applications except for lifeboat handling. 5.13 Navigational Space Lighting. When the ship is underway at night, all lights provided for general illumination in the wheelhouse and all exposed lights in the forward part of the vessel are extinguished for better vision of the navigating officers. Any illumination that would be a detriment to navigation must be properly controlled. J
MARINE ENGINEERING
ELECTRIC PLANTS
In the wheelhouse, only lights built into the binnacle, telegraphs, and other necessary instruments remain lighted. The character and intensity of these lights should be such as to permit reading the instrument without interference with outside vision. A shaded light controlled by a switch is provided over the chronometer box t o illuminate the faces of the chronometers when reading is necessary. I n the chart room there should be provided over the chart table one or two bulkhead-mounted adjustable arm type fluorescent desk lights fitted with a switch and a red filter. The arrangement should be such that the lights can be moved in a horizontal plane over the entire table for intense lighting of any portion. The log desk in the wheelhouse should have a similar light. 5.14 Navigation and Signaling Lights. All vessels must be equipped with running and signal lights in compliance with applicable International and Inland Rules of the Road as enacted into law by the United States Congress. The United States Coast Guard publication "CG-169 Rules of the Road, InternationalInland" provides detailed requirements regarding these lights. All fixtures are watertight, substantially constructed of corrosion-resisting material, and are fitted with Fresnel lenses. Navigation and signal lights fitted on ocean-going vessels are listed in Table 3. A typical arrangement of navigation and signal lights for a cargo vessel is shown in Fig. 15. The side lights are located port and starboard on the navigating bridge level so that each is visible at a distance of at least 2 miles on its respective side from
directly ahead to 22.5 deg abaft the beam. These lights are fitted with inboard screens projecting at least 3 f t forward from the lights so as to prevent these lights from being seen across the bow. The masthead and range lights are identical and screened so as to show forward through an arc of 225 deg; that is, to 22.5 deg abaft the beam on either side. These lights are located usually one on the forward mast and one on the after mast, in a line with and over the keel so that the forward light (masthead) is a t least 15 f t lower than the after light (range) and visible at a distance of at least 2 miles. The horizontal distance between the two lights must be at least three times the vertical distance between the lights. For an on-coming vessel the alignment of masthead arid range lights shows the course, and the color of the visible side light indicates whether itwill pass to starboard or port. If both side lights are visible, the vessel is coming head-on. The stern light is located on the centerline at the ship's stern and is screened so as to show aft through an arc of 135 deg (67.5 deg to port and starboard of the centerline); it must be visible from a distance of at least 2 miles. The forward anchor light is located at the bow of the vessel and at least 20 ft above the hull; the after anchor light is located near the stern at a height not l e ~ sthan 15 f t lower than the forward anchor light. Both lights are required to be visible all around the horizon at a distance of at least 3 miles. The forward light is usually permanently mounted on top of the jack staff and the after light on top of the ensign staff. The not-under-command light installation consists of two lights located in a vertical line, one over the other, not less than 6 f t apart and visible all around the horizon at a distance of at least 2 miles. These lights Table 3 Navigation and Signal Lights on Oceangoing Ships are normally portable but may be permanently installed. The towing light installation in accordance with International Rules consists of two lights located in a vertical line with the masthead light, one over the other, not less than 6 f t apart and screened to show forward green 112.5 Starboard side Port aide 112.5 red through an arc of 225 deg (22.5 deg abaft the beam on 225 clear JMasthead either side). On tugboats and other vessels expected to 225 clear Range operate frequently with a tow, thege lights are installed 135 clear ustern 360 clear Forward anchor permanently; otherwise they are portable. After anchor 360 clear The running lights (masthead, range, stern, and side) Wot under command 360 red must be constructed so as to have a backup light in event 225 Towing dear
NAVIGATION SEARCHLIGHT -STERN
LT.
1=.---RANGE
LT. NOT-UNDER-COMMAND
----S I G N A L SEARCHLIGHTS (BRIDGE WINGS)-'
L
~
MASTHEAD LT.
LTS.
LTS.l PORTa ~ ~ STBD.
Fig. 15 Typical arrangement of navigation and dgnal llghh f a a cargo vessel
I
of failure of the primary light. One of the two methods used for complying with this requirement is the use of a bi-filament lamp in a single-compartment fixture. The other method is the use of a two-compartment fixture with a single-filament lamp in each compartment. The latter method is considered the most dependable and is the most commonly used. The signal lights (anchor, not-under-command, and towing) are singlecomparb ment with a single-filament lamp. Each compartment of single and two-compartment fixtures is fitted with a No. 14 AWG &conductor flexible cable having a 3point plug for attachment to its respective lighting supply receptacle. These plugs and receptacles should be of a unique design to preclude their use for other purposes. Vessels equipped with a steam whistle are usually provided with a whistle light to illuminate the steam released when operated. In addition to the foregoing, tank vessels are required to display a red warning light during transfer of bulk cargo; this light should be located above the wheelhouse so as to be visible on all sides. The exact location of running and signal lights and the provision of mounting arrangements, screens, and rigging are generally the responsibility of the hull designer. 5.15 Navigating and Signal Light Controls. A navigating and signal light panel is installed in the wheelhouse for control of all running and signal lights. This panel combines an automatic or semiautomatic telltale navigating light section for audible and visual alarm and' control of the masthead, range, side, and stern lights and a signal light section for control of the anchor, not-under-command, and towing lights. The navigating light section is arranged to indicate failure of each primary lamp or filament and is provided with a switch, either automatic or manual, for transfer to the secondary lamp or filament. The signal light section is arranged only for "on-off" operation of the individual light supply circuit. 5.16 Signaling Lights and Searchlights. A daylight signaling light is required on all ocean and coastwise self-propelled vessels over 150 gross tons and on tankers over 150 gross tons that are engaged on international voyages. It may be either a portable hand-held type, a permanently fixed and wired lZinch unit mounted on top of the wheelhouse, or a semi-fixed 12-inch unit with arrangements for quick mounting and electrical connections at either wing of the navigating bridge. The portable unit may be complete with a self-contained battery or energized from a special low-voltage receptacle.
Searchlights, as such, are not required by the regulatory rules and are installed only when requested by the owners. When requested, u+ly an l&in. incandescent searchlight is installed on the top of the wheelhouse with manual control of train and elevation from within the wheelhouse. ~p"on-offn switch would be located near the operating point. Ships traversing the Suez Canal are required by the Suez Canal Authority to have mounted over the bow a searchlight with a special diffused beam to illuminate the banks of the Canal. Normally, this light is rented from the Canal Authority and is energised from a special receptacle located at the bow. 5.17 Bmcket Fans. Bracket fans have become obsolescent in living areas due tq the extensive use of air conditioning; however, they are used in spaces not air conditioned mch as galleys, pantries, laundries, workshops, and the wheelhouse. Bracket fans are either 12 in. or 16 in. and of the oscillating, 3-speed marine type. Mounting on thin partition bulkheads, where vibration may result in objectionable noise, should be avoided. Receptacles are provided for each fan and are located so that the portable cable length is minimal. 5.1 8 Wiring Appliances. The bodies of watertight fixtures and the special mounting boxes of other fixtures are used for wiring connection boxes as far as circumstances permit. Additional branch boxes are provided where necessary. For exposed wiring, they should be of watertight or nonwatertight type as required, and surface mounted. I n ceiled spaces, they may be flush mounted with sheet metal covers if the decorative character of the space permits; otherwise, they should be installed behind removable panels for accessibility. Switches and receptacles in public spaces, and living areas in general, should be of standard commercial type in the smallest enclosure obtainable; the depth, in particular, is limited. Switch and receptacle plates are usually of a standard sise but may be a special narrow type to fit on joint strips and door frames. They may be of brass or aluminum suitably finished, or plastic material. The mounting boxes for switches should be of adequate size for wiring and are provided with arrangements for clamping the entering cables. Specially designed boxes to suit conditions may be required. In spaces requiring watertight installation, the appliances should be-of the standard marine h t e a i h t tGe. Where aluminum boxes are used, particularly in the weather, care should be taken in their mounting to avoid the possibility of deterioration through electrolysis.
640
ELECrRlC PLANTS
MARINE ENGINEERING
I
Section 6 Lighting and Power Distribution 6.1 General. Energy for lighting and power loads is supplied from the ship service generators (through their associated switchboard and via the ship service distribution system) or from the emergency generator or battery (through the emergency switchboard and via the emergency distribution system). Normally, the emergency switchboard and the emergency distribution system are energized through a bus tie from the ship service switchboard. If the ship service power fails, the emergency distribution system is automatically transferred-fromits normal source to the emergency generator. Subsection 3.7 contains details regarding this automatic transfer. There are many different arrangements for distributing power to the various types of electrical loads installed on shipboard. Figures 16, 17, and 18 are typical generator and bus tie diagrams for a tanker, a dry cargo or containership, and a large passenger vessel, respectively. It should be noted that these diagrams are typical in layout and should not be considered as being restrictive or fixed for any type of vessel.
On large passenger vessels two or three sub-distribution, or "load-center" switchboards are provided for lighting and power distribution. Generally, one will be located in the forward part of the vessel, one aft and, if the size of the vessel warrants, a third will be amidship. Each should be centrally located regarding the loads supplied. Each of the load-center switchboards is supplied from the ship service switchboard by a bus feeder, as shown by Fig. 18. This arrangement is much more economical than providing numerous long feeders from the ship service switchboard to all parts of the vessel. Each load-center switchboard should be installed in a suitable compartment. These compartments usually serve also as centers of electrical serviceand maintenance, and each may be provided with a workbench and with bins 6nd a locker for spare lamps, fuses, and other electrical supplies. 6.2 Lighting Distribution. The lighting bus of each distribution switchboard is supplied by a threephase transformer bank; each bank consists of three 450/12U-
volt singlephase transformers connected delta-delta. In some installations that use rapid-start (without starters) fluorescent lamps, the secondary of the transformer bank is wyeconnected, in lieu of delta-connected, re to with the neutral-_grou ensure reliab6tarting o All lighting distribution panels are supplied by threephase feeders from the lighting bus of the applicable distribution switchboard. These panels are arranged for a three-phase supply and singlephase distribution. The single-phase loads are connected to the threephase aupply bus so as to ensure approximately balanced loading per phase. a. Lighting Feeders. All ship service lighting requirements are supplied by feeders from ship service distribution switchboards through lighting distribution panels. In general, it is economically good practice to limit the load supplied by each lighting feeder to less than 100 amps, so that the feeder may be supplied from its bus through a 100-amp circuit breaker. At least two feeders are provided to senre the lighting requirements of each machinery space.
Separate feders are provided for lighting in cargo spaces. One feeder is usually provided for each cargo hold so they can be disconnected at the switchboard when the vessel is a t sea, thus eliminating the possibility of electrical fire hazards in thpse unmanned spaces. Separate feeders are provided as necessary to supply all lighting requi7ements in working and living areas not covered in the foregoing. For passenger vessels that are subdivided into zones by fire-screen bulkheads, separate feeders are provided for each zone as necessary to supply the lighting requirements between adjacent fire-screen bulkheads. Ship service feeder and emergency feeders that supply the same or adjacent areas sxould be routed so as to be separated as widely as practical to minimize the possibility of damage to both feeders from the same casualty. q e based on 100 percent of Lighting. feeder c-a,b&-..ggs the total connected load plus t6e +errqge-iF,tive circuit load for each spare switch or circuit breaker on the panel being supplied. b. Location of Lighting Panels. For machinery spaces, the ship service lighting panels are usually /
CIRCUITS TO INDIVIDUAL LIGHTING LOAD CENTER FWD CARGO HOLDS
1
3 0 . 4 5 0 / 1 2 0 VOLT TRANSFORMER LOADS AS REQ'D
I
t
1 I
I
I
I
I
I
I
I
I
I
1
I
I
I
4
I
I
--
i
LOADS AS REQ'D
INTERLOCK
TO CARGO H O ~ D LTG PANELS AS REQ'D
I
-
LIGHTING AFT CARGO LOADHOLDS CENTER
-- -
---
j
I I 2 0 VL !T
EMERGENCY BUS
4 5 0 VOLT EMERGENCY BUS
--
--
I
LOADS AS REQ'D
I
+
I
I
L REQID o A m
t
4
L,
icLonos As R E Q Q ~
EMERGENCY SWITCHBOARD
2s
I
! I
!
I I II
90 I
!
_1
I
SHORE POWER TERMINALS
1
!
I
EMERGENCY SWITCHBOARD
3 0 , 4 5 0 1 1 2 0 VOLT TRANSFORMER
II
120 VOLT BUS
/ - r e
I
I
-+-+TO CARGO HOLD LTG PANELS AS REQ'D
I
r
CIRCUITS TO INDIVIDUAL LOADS
I
AS REQ D AND DI$TRIBuTIoN PANELS
I
- - --- I
,---A
-AS REQ'D
I
120 VOLT BUS 120 VOLT MAIN BUS
--
4 5 0 VOLT MAIN BUS
--
4 5 0 VOLT BUS
- ---
. ----
SHIP SERVICE SWITCHBOARD
SHIP SERVICE SWITCHBOARD
Rg. 16 Typical generalor and bur tie diagram f a tankers
230 VOLT BUS'
I
Fig. 17 Typical generator and bus tie diagran f a a dry cargo a containership
J
I
642
I
MARINE ENGINEERING
located on the main operating level, and where readily accessible as typically shown in Fig. 3 of Chapter 18. On some vessels these panels are located at the main entrance to the space. Panels for cargo lighting are usually located in cargo handling machinery deckhouses so as to be accessible, and so that the +igh+ing in each cargo hold may be deenergieed when loading is complete. These panels are not permitted to be located in the cargo holds. The number of panels for cargo lighting depends on the size and arrangement of the vessel; generally, one panel is provided for each cargo hold. The location of ship service lighting panels in paBsenger and crew spaces is determined, to a degree, by the structural and fire zone subdivision of the vessel. Generally, there will be one or more panels on each deck in each subdivision or fire zone; however, two or more decks may be supplied by a single panel, if the arrangement permits. Each panel should be located as near as practicable to the center of the areas served to limit the voltage drop in the branch circuitsj panels are usually installed on passageway bulkheads. In way of joiner work they should be of the flush type. I n public spaces the panels are located near the en-
ELECTRIC PLANTS
trance doors and where the operator can see the lights controlled. c. Lighting Branch Circuits. Lighting branch circuits may be 15-amp, 20-amp, or 30-amp capacities depending on the specific application. Fifteen-amp branch circuits are used for general lighting applications, and each circuit is limited to a maximum connected load of 12 amps (1380 watts) when wired with a No. 12 AWG conductor. When wired w i t h u E G coii&ctors, the maximum connected load is 880 wet&: - --Twenty-amp branch circuits are normally used to supply only fixed nonswitched lighting fixtures for cargo hold or deck lighting. Each circuit should be limited to a maximum connected load of 16 amps and be wired with not less than No. 12 AWG conductors. Thirty-amp branch circuits are normally used to supply only fixed nonswitched lighting fixtures having lamp holders of the mogul (oversize) type or lamps exceeding 300 watts. Each circuit should be limited to a maximum connected load of 24 amps and be wired with not less than No. 10 AWG conductors. Appliance loads, heater loads, and miscellaneous small motors utilizing lighting system voltage may be supplied from lighting distribution panels. Each branch circuit
should be limited to a maximum connected load of 30 amps. The connected load on a general lighting branch circuit is based on the actual lamp (incandescent) sizw installed, but not lei% than 50 watts per lamp unless the fixture design does not permit the use of lamps having a higher wattage than originally installed. The connected load for circuits supplying electric-discharge type lamps (fluorescent and mercury vapor) is based on the ballast i n ~ u tcurrent for each fixture. Receptacle outlets in'stalled for the convenience of the passengers or crew are not included as a connected load. Special lighting fixtureshaving a large number of lowwattage lamps are supplied by a three-phase circuit when the total load of the fixture exceeds 12 amps. The supply circuit is controlled only from the distribution panel, and the current in any conductor is limited to 12 amps. Overcurrent protection for lighting branch circuits is limited to 10-amp fuses or Isamp circuit breakers for the 880-watt circuits, I s a m p fuses or circuit breakers for the 1380-watt circuits, and 20-amp and 30-amp fuses or circuit breakers for the 20-amp and 30-amp circuits respectively. In general, lighting branch circuits in machinery spaces are arranged with alternate groups of lights on different branch circuits so that large areas wiU not be B'A 1- EY ' /3%450/120 VOLT r-7~ CHARGER 1 i put in darkness by failure of a single branch circuit. In I r TRANSFORMER these spaces no individual switches are provided as the SWITCH~OARO . FWD SHIP SERVICE SWITCHBOARD lights are controlled only by the panel switches. 4 51 0-0" (20;;BUS 4 5 0 VOLT BUS " 7 Each passenger stateroom and public space should be supplied by at least two ship service lighting branch circuits, so arranged that in event of failure of one branch there will be adequate light to permit use of the space. The ceiling fixtures in each stateroom, berthing space, storeroom, and similar small compartments are controlled by a switch at the room entrance, and located about 4 ft I .I . L I I i ~ i ~ i above the deck. Staterooms having more than one I I -I I entrance door should have a switch at each door for t t LOADS AS control of the ceiling fixtures. Berth, desk, lavatory, IC LOADS LOADS AS LOADS AS 1 i?) I REO'D AS REO'D' REO'D REO'D and dreeaing table -lights are usually controlled by I I individual switches mounted on the fixtures or located near the fixture. r--/30.450/120 VOLT, XFMR, ! I Separate branch circuits are provided exclusively for passageway lighting. Also, the lights in each passage should be divided between ship service branch circuits and emergency branch circuits, such that normal and emergency illumination requirements are satisfied. Separate branch circuits are generally provided for bracket fans; however, where relatively few widely distributed fans are installed, they are supplied from local lighting branch circuits. Branch circuits should not be routed through firescreen or watertight bulkheads. -. Typical arrangements of lighting branch circuits for L. . . . I I . . .. L.I . .1 I . ..1 . AFT LQhD CENTER SWBD a crew stateroom and a passenger stateroom are shown AFT SHIP SERVICE SWITCHBOARD FWD LOAD CENTER SWBD MIDSHIP LOAD CENTER SWBO in Figs. 19 and 20 respectively. Fig. 18 Typical wmratw and bur tie diagram for Q large panengar v-l 6.3 Power Dirhibvtion. Ship servicQpower system
-
1 '
loads consist principally of motor-driven auxiliaries and heating equipment, and are supplied either individually or in groups by feeders from a ship service distribution switchboard. Individud feeders are normally used to energize large propulsion pbnt auxiliaries that are located in the same space as the distribution switchboard, but may ,be used for large motors anywhere in the vessel. Grouped loads are supplied by feeders through distribution panels, these panels being centrally located to the loads supplied. a. Power Feeders. Separate feeders should be provided to panels and group control boards serving machinery space auxiliaries and refrigeration equipment that are not supplied individually. Machinery space ventilation fans, living and working space ventilation fans, and cargo space ventilation fans should be supplied by separate feeders. Each ventilation feeder circuit breaker should be provided with a means of remote control for deenergizing its feeder in case of fire; the remote control device provided for deenergizing msr chinery space ventilation feeders should be located in the passageway leading to, but outside of, the machinery space. For all other ventilation feeders, the remote control devices are usually located in the wheelhouse, but may be located in a space near the wheelhouse as permitted by the regulatory rules. The means of remote control for ventilation feeders,$onsists of a normally closed switch which, when 'operated to the "stop" position, deenergbes an undervoltage trip device on the circuit breaker, thus tripping the circuit breaker. This arrangement may be such as to trip several breakers (feeding various ventilation feeders) at one common trip switch. These switches are mounted in special locked enclosures having glass fronts which must be broken to operate the switch. Separate feeders should be provided for galley appliances, air heaters other than isolated units, and each group of cargo handling equipment. Equipment required to be operated when underway should not be supplied from the cargo handling equipment feeders, since these feeders are usually disconnected from the Wribution switchboard when at sea. Windlass and capstan motors may be supplied from these feeders if convenient. The steering gear should be supplied by two independent feeders, separated as widely as practical to reduce the probability of 10- of power from a single casualty. Both feeders are pormally applied from the ship service distribution switchboard; however, if the capacity of the emergency power source is adequate, one feeder may be supplied from the emergency switchboard. The steering gear distribution panel is arranged, through interlocked circuit breakers, to connect each motor or appliance to either feeder. A separate feeder should be provided from the ship service distribution switchboard to a shore connection box. The shore connection box should be suitably located for supplying power from shore to ship when tied up at shore facilities.
'
644
ELECTRIC PLANTS
MARINE ENGINEERING
BERTH LTS. 30 W EACH FAN 75W
11
LEGEND 0 RECEPTABLE X SWITCH OQ DOUBLE RECEPTABLE K SWITCH, 3 WAY Fig. 19 Typical lighting branch circuit for a crew stateroom
O< -SWITCH, 3 WAY @ -RECEPTACLE
8 -CONNECTION BOX
Fig. 20 Typical lighting branch circuit for a pauemger stateroom
The minimum current-cawing capacity of power system feeders to individual motors should be based on a t least 125 percent of the motor full-load current. In general, for feeders supplying a group of motors, the feeder current capacity should be based on 125 percent of the largest motor rating, plus the sum of the ratings of all other motors, plus 50 percent of the rating of the spare switches on the panel supplied. For feeders supplying cargo handling equipment, workshop tools,
windlass, and capstans, refer to regulatory rules for specific requirements in determining cable currentcarrying capacities. The current capacity of galley panel feeders should be based on 100 percent of the first 50 kw of load or one half of the total connected load (whichever is larger), plus 65 percent of the remaining connected load, plus 50 percent of the ratings of the spare switches. The current capacity of transformer feeders should be based on
100 percent of the rated primary and secondary currents. Steering gear feeder current capacities should be based on the total ratings of all equipment normally connected to it that operate simultaneously. b. Location of Power Panels. For mchiinery spaces, the ship service power panels, including those for ventilation fans, are usually located on the main operating level, so as to be readily accessible and central to the auxiliaries supplied. Panels supplying cargo handling equipment and cargo hold ventilation panels are usually located in cargo handling machinery deckhouses. The steering gear transfer panel is located in the steering gear - room. Ventilation panels serving living and working spaces, and general service panels are located as near as practical to the loads supplied. Where the loads served are located in a single compartment, the panel should be in the same space. Galley panels are located within the galley space or in a passageway and adjacent to the galley entrance. c. Power Branch Circuits. Separate branch circuits should be provided for each motor having a full-load current of 6 amps or more, and for each air heater regardless of rating. With these exceptions, several small loads, having a total rating not exceeding 7.5 amps, may be grouped and supplied by a branch circuit through a subdistribution panel; the circuit conductors should not be smaller than No. 14 AWG and should be protected by a circuit breaker of not more than 15-amp rating or fuses or not more than 10-amp rating. Also, groups of loads not exceeding 15 and 20 amps respectively may be supplied by branch circuits as noted in the foregoing, provided the circuit conductors are not smaller than No. 12 and No. 10 AWG, and the overcurrent protection is rated not over 15 and 20 amps respectively. Receptacle outlets should not be supplied from these branch circuits. The current-carrying capacities of motor .branch circuits should be based on at least 125 percent of the motor full load; the carrying capacities of all other branch circuits should b e based on a t least 100 percent of the connected full load. All branch circuits should be wired with not less than No. 14 AWG conductors. Each branch circuit should be protected by a circuit breaker, with thermal or magnetic trips, or by a fuse. Since there are many different requirements applicable to overcurrent protection for branch circuits, the regulatory rules should be reviewed for each specific application., The magnetic instantaneous setting for motor branch circuits .should be a t a higher value than the starting current of the motor; for branch circuits supplying loads through transformers, this setting should be higher than the inrush current of the transformer. Power branch circuit loads that require other than normal ship service voltages are supplied either by individual transformers for each load, or by a bank of transformers in the feeder circuit just ahead of the distribution panel. Typical loads of this type include
645
laundry equipment, galley equipment, and special refrigerated containers. . d. Special Features. Vent fans serving spaces that are subject to carbon dioxide flooding should be arranged to be deenergized when carbon dioxide is released into the space as the carbon dioxide will otherwise be spread to other regions of the ship. The means for deenergizing the fan motors is a 'presstireoperated switch in the carbon dioxide piping system. When the switch is actuated, by relgase of the carbon dioxide, it may deenergize an uidervoltage trip device on the circuit breakers supplying the fans involved, thus tripping the circuit breaker and stopping the fan motors, or actuation of the switch mav cause the fan motor control circuit to be deenergized, thus stopping the fan motor. Also, a similar emergency means of stopping fuel-oil service pumps, fuel-oil transfer pumps, and forced-draft blowers should be provided if the machinery space in which they are located is subject to carbon dioxide flooding. Stopping these motor-driven auxiliaries would avoid the possibility of fuel oil feeding the fire from a ruptured oil line, and reduce the probability of the forced-draft blower discharging the carbon dioxide from the space. Circuit breakers supplying the fuel-oil service pump, fuel-oil transfer pump, and forced-draft blower motors should be provided with a means of remote control for deenergising each motor in case of a fire in the machinery space. This means of remote coiitrol may consist of a normally closed switch which, when operated to the "stopJ' position, deenergizes an undervoltage trip device on the circuit breaker, thus tri$ping the circuit breaker. This arrangement may be such as to trip each breaker ~ These switches are mounted at one common t r i switch. in special locked enclosures having glass fronts which must be broken to operate them. The switches are required to be located outside the machinery space and may be located adjacent to the emergency stop switches provided for machinery space ventilation fans. The Coast Guard rules should be reviewed to determine the specific requirements, as may be applicable to each vessel, regarding the following: watertight door system, fire-screen door system, lifeboat winches, steering gear, air heaters, motion picture projection rooms, hospital operating room, and locations where gasoline or motor fuel is carried in vehicles. These requirements involve: types and location of equipments, special codes that must be complied with, cable size and overcurrent protection restrictions, and special features for control circuits. For a discussion regarding the automation of the machinery space, includins bridge control of the propulsion plant, see Chapter 51. On many vessels, an impressed current cathodic protection system is provided to prevent corrosion of the rudder, propeller, and the submerged hull. The system supplies a low positive voltage to submerged anodes from which current flows through the seawater to the hull, the purpose being to suppress the normal flow of small currents between areas of a submerged
646
ELECTRIC PLANTS
MARINE ENGINEERING
hull so as to prevent corrosion. System components usually include : 1 Hull-mounted reference electrodes, in quantity as required, to provide the degree of protection desired. 2 Hull-mounted anodes, in quantity and location as required, to provide the desired protection. 3 Controller, usually located on the operating level in the main machinery space, to select, monitor, and provide the proper signals to the power supply units for the desired protection. 4 Power supply units, in quantity and location as necessary, to provide the necessary regulated d-c power to the anodes. 5 Propeller shaft grounding assembly to provide a path of low electrical resistance from the propeller to the hull. 6.4 Emergency Lighting and Power. Some form of emergency lighting ordinarily is provided for every vessel equipped with an electric lighting system. Exceptions include, (a) small passenger vessels that operate only between sunrise and sunset, and (b) small passenger vessels that operate not more than ore of having their source of general lighti the propulsion plant and located above the bulkhead deck. The temporary source of emergency power required on large passenger vessels is of limited capacity and is rated to supply only selected emergency loads, for a short time, while a large-capacity final emergency source is being started. J X e following loads are reqhired to be energized from the tgmporary source when available: 1 Navigation lights. 2 Adequate number of lights in the machinery space to permit the performance of essential operations and observations under emergency conditions and to facilitate the restoration of service. 3 Lighting for passageways, stairways, escape trunks, passenger quarters, crew quarters, public spaces, machinery spaces, and work spaces adequate to permit the passengers and crew readily to find their way to open decks and to lifeboat embarkation and assembly points with all watertight doors and fire-screen doors closed. 4 Illuminated signs bearing the word "exit" in red letters installed as required so that the direction of escape to the open deck is apparent (passenger vessels only). 5 General illumination for safe operation of poweroperated watertight doors. 6 One or more lights in galleys, pantries, radio room, steering gear room, emergency power room, chart room, wheelhouse, crew's mess, and recreation rooms. 7 Lighting for boat and embarkation decks and passenger assembly points for safe embarkation into lifeboats. 8 Electric communication systems essential under temporary emergency conditions and which do not have an independent storage battery source of power. 9 Power-operated watertight door system. 10 Emergency loudspeaker system.
11 Fire-screen door holding and release system. 12 Supply to motor generator or other conversion equipment when a temporary source of a-c power is necessary for essential communication systems, emergency, or safety requirements. The entire emergency load, consisting of the following, is required to be supplied from the final source of emergency power : 1 All loads energized from the temporary source. 2 Illumination for the safe operation of the lifeboat and life raft launching gear and the lifeboats and life rafts in process of and immediately after launching. 3 Charging panel of temporary emergency battery and of starting battery for the diesel engine driving the emergency generator. 4 One bilge pump, one fire pump, and one sprinkler pump, if required to be supplied from the emergency source. 5 Daylight signaling lights. 6 Smoke detector system. I n addition to the aforementioned requirements, regarding the loads supplied from the final source of emergency power, the United States Coast Guard rules recommend that the following loads, if installed, be supplied from the normal emergency source when the capacity and character of the emergency source permits:
Radio installation Radio direction finder Loran ,+A Radar plan position indicator \g Gyrocompass Depth sounder 9 Electric whistle and siren control The emergency source of power is required to have sufficient capacity to supply only those loads that are required to have an emergency supply; however, nonemergency loads may be supplied from the emergency source provided there is adequate capacity to supply all of the loads that may be connected simultaneously. The location of the emergency generating set and its characteristics are described in Subsections 2.4 and 2.5, respectively. When the emergency power source is a battery, its location requirements are the same as noted for the emergency generator. Batteries used for the emergency power source should have characteristics as specified in the rules of the regulatory bodies. The arrangement of the emergency switchboard, including automatic transfer equipment, is discussed in Section 3.7b. The requirements for emergency supply for various types of vessels are included in the United States Coast Guard rules. The Coast Guard rules are revised periodically, but are typically as follows:
a. Passenger Vessels Over 65 ft in Length 1 I n ocean and coastwise service, the emergency source should be a storage battery of sufficient capacity to supply continuously the full emergency load for 36 hr, or
a diesel or gas turbine driven emergency generator with capacity to supply continuously the full emergency load for 36 hr and a temporary source of emergency power consisting of a storage battery of sufficient capacity to supply the "tem~orary"emergency load for not less than hr. Arrangements must be provided for an automatic transferof the full emergency load to a battery supply upon loss of the normal supply, or an automatic transfer of the U t e m ~ o r a load ~'l the battery s uand ~ ~ automatic starting of the emergency generator, transfer of the full emergency load to the emergency generator upon loss of the normal supply, as applicable. 2 In than Ocean and coastwise service, for vessels of 100 gross tons and Over, the source should be a diesel or a gas turbine driven emergency generator or a storage battery of sufficient capacity supply ~ontinuousl~ the full emergency load for 8 hr, or twice the time of run, whichever is the smaller. Arrangements must be provided automatic transfer of the load to battery upon loss of supply, or automatic shrting ofthe emergency generator and the transfer of the emergency load to the emergency generator upon loss of normal supply, as applicable. 3 In other than ocean and coastwise service, for vessels over 15 gross tons but less than 100 gross tons, the emergency source should be a diesel or a gas turbine driven emergency generator of a storage battery of sufficient capacity to supply continuously the full emergency load for 8 hr or twice the time of run, whichever is smaller. Arrangements must be provided for either automatic (as just noted in a.2) or manual transfer of the emergency load to the emergency source. Manual transfer requires that only a single operation of an "onoff" switch will cause the emergency system to supply its connected load. This "on-off" switch is located in the wheelhouse or as necessary to be under the control of the chief engineer.
b. and and Tank Ships
Se&?opeued
1 In all-waters service, for vessels of 1600 gross tons and over the emergency source is the same as just given in a.3 except that the emergency load is required to be camed for 12 hr. 2 I n all-waters service, for vessels of 300 gross tons and over, but less than 1600 gross tons, the emergency source may be the same type as noted in a.3 or may be relay-controlled battery-operated lanterns. The emergency source selected must have sufficient capacity to supply the full emergency load for 12 hr or twice the time of run, whichever is smaller. When batteryoperated lanterns are used for the aforementioned emergency service, they are required to have rechargeable batteries, incorporate an automitic battery charger that w i l l maintain the battery in a fully charged condition, and not be readily portable. Also, the minimum period of operation of these lanterns may be less than 12 hr but not less than 6 hr.
6.5
647 Emergency Distribution System
a. Lighting. In general, the emergency lighting system forms a part of the ship servicelighting system and is energized at all times when the passengers or crew are aboard. Separate emergency lighting feeders are provided for machinery spaces, crew and passenger areas, and lights controlled from the wheelhouse. The machinery space feeder supplies only those emerlgency ~ , lights loc,ted in the machinery space; this rupply is through a didribution panel usually located above the operating l e e 1 and near the main entrance to the space. Usually the crew and passenger area feeders supply distribution panels suitably located for the control and distribution of the emergency fights throughout these areas. For passenger vessels subdivided into mnes by firescreen bulkheads, a t least one emergency feeder is provided to supply the lighting in each zone between adjacent fire-screen bulkheads. A separate feeder, supplying a panel in the wheelhouse, is provided for emergency lights located in or controlled from the Loads supplied from this pane] include the navigation light panel, signaling lights, emergency lights on open decks, lifeboat lights, and chart room lights. The navigation light panel is generally supplied by a through feed, without protection, from the wheelhouse emergency panel; but, as an alternative, the navigation light panel, signaling lights, and lifeboat lights may be supplied by separate feeders from the emergency switchboard. I n addition, on vessels that have two "islands" (e.g., those with midship and aft houses), lighting for lifeboat-associated areas located remotely from the wheelhouse "island" may be controlled from a central location within the remote "island " in lieu of from the wheelhouse. Floodlights for adjacent lifeboats should be supplied by different branch circuits. Emergency lighting fixtures usually are of the same type as the ship service lighting fixtures installed in the same spaces. I n public spaces where fixtures may contain more than one lamp, only the necessary number of lamps to give the required illumination are connected to the emergency circuit. For ready identification, all lighting fixtures on the emergency system are identified by a metal tag stamped with the letter E (in red) a t least % in. in height. This tag is secured to the deck or paneling immediately adjacent to the fixture and, for fixtures having lamps on both the ship service and emergency circuits, an additional tag is secured to the emergency supply cable within the fixture mounting box. b. Power. Separate &ergency power feeders are routed from the emergency switchboard to each power system load individually. Only a limited number of power system loads are required to have an emergency supply and they are not centrally located with respect to each other. c. Communication and Alarm. It is general practice to supply vital machinery space indicating, monitoring,
648
Table 4
Correction Factor for Cable Calculations
CABLE AWG
POWER FACTOR OF LOAD
fl
No.
1.00
0.95
0.90
0.85
9.80
0.75
0.70
0.65
2 1 n
1.00 1 .OO
1.01 1.03 1 0.5
0.99 1.01 1.04
0.96 0.98 1.02
0.92 0.95 0 .gg
0.84 0.88 0.93
0.76 0.80 03 5
0.68 0.71 0.77
r nn
ELECTRIC PLANTS
MARINE ENGINEERING
,
The percent voltage drop in 3-conductor, 3-phase circuits 52,600 CM and smaller is 173 I R L Voltage drop = (2) CMV
92.2' TO CALCULATED -----LOAD CENTER . 8 9 0 % DROP--
n
---
For 3-conduchr, %phase circuits 661400 CM and larger the percent voltage drop is 173 I R L CF Voltage drop = (3) CM V
L and alarm loads from an emergency panel that is conveniently located on the machinery space operating level. This panel is supplied directly by a feeder from the emergency switchboard. Other communication and alarm loads are supplied individually from the emergency switchboard. d. Electronics. Electronic loads are usually energized from an emergency panel, centrally located to these loads, that is supplied by a feeder from the emergency switchboard. Electronic loads include radio, radio telephone, radio direction finder, loran, and depthsounding equipment. Radar equipment is usually supplied directly from the emergency switchboard, since this equipment often requires a power supply voltage different from the remaining electronic loads. e. General. The distribution panels used for emergency system distribution are of the same type as those for the ship service distribution system. Emergency system cables that are not required to terminate at equipment located within machinery spaces, uptakes, or casings should be routed to avoid penetrating the boundaries of these spaces, and should be kept clear of the decks and bulkheads forming these boundaries. Voltage drops in emergency circuits are calculated in the same manner as for ship service circuits. For lighting circuits the voltage drop from the emergency switchboard lighting bus to the most remote fixture should not exceed 3 percent. For all other circuits the total voltage drop should not exceed 5 percent when supplied from the main bus of the ship service switchboard through the bus tie to the emergency switchboard.
CM V CF
m
--
2.17AMPS .#v
-"
.483%DROP
1-
m
-".178%
.--
-
".104%
-1-
-- -
.077%
'
.G8%
1
k---FIVE
5 0 WATT 115 VOLT LAMPS-CALCULATED VOLTAGE DROP . 8 9 0 % ' F 0 ~ LONGEST BRANCH CKTNOTE: 4 4 9 7 CM CABLE USED IN<,LIEU OF 4 1 10 CM (NO.14 AWG) SWITCHBOARD 120 VOLT SECTION
= current, amps = resistance of conductor, ohms/CM-ft
(for copper, R is nominally 12) = single length of circuit, f t = Area in circular mils of each conductor or total area of conductors in parallel = Circuit voltage = Correction factor determined by system power factor and cable constants (see Table 4)
a. Lighting Circuits. Lighting circuit wire sizes are usually fixed by voltage drop; however, wire sizes of short runs of feedersand mains, in combination, for such spaces as engine and boiler rooms, may be determined by currentcarrying capacities. The IEEE rules establish the maximum values of current that may be carried by each size and type of cable. In general, the voltage drop in any lighting circuit should not exceed 3 percent from the distribution sw,itchboard lighting bus to the most remote fixture. In calculating the voltage drop for large lighting cables, the correctionfactor, CF, in equation (3) may be considyed as 1.00, since lighting loads consist generally of incandescent lamps or high power-factor fluorescent lamps. However, if a significant portion of the load is low power-factor fluorescenh-hmps, a correction factor should be used as noted for power system circuits. Figure 21 shows a typical lighting branch circuit for a passageway or cargo hold, with no subbranches and with all outlets uniformly rated. It is assumed that this branch has the greatest drop of any supplied by the feeder under consideration. The percent drop indicated for each section was calculated using equation (I), and the total percent drop for the entire branch was determined to be 0.890 percent. To determine the minimum s2e of the feeder conductors, the branch circuit percent drop is subtracted 6.6 Determination of Wire Size for Lighting and from 3.00, and the remainder is substituted in equation Power Circuits. Classification society rules establish the (2). For the feeder shown in Fig. 21 the minimum CM maximum voltage drop limits for various electrical is determined to be circuits. .In order to determine that these limits are met, cable voltage drop calculations are made for each circuit sup~lying:power to electrical loads. The voltage drop, inpe;ce&,-for a 2-conductor circuit (d-c or single-phase The next larger size is No. 2 AWG, with a sectional area a-c) can be computed as of 66,400 circular mils, and would be the proper selection. 200 I R L From wire tables it may be noted that No. 8 AWG with Voltage drop = a sectional area of 16,500 circular mils, for rubber or CM V
t -
-
---- +
2.1 I % ALLOWABLE DROP SIZE N0.2-AWG ( 6 6 , 4 0 0 CM)
In the foregoing equations:
I R
120 VOLT. 3 p H FEEDER ~~AMPERES2 14'-0"-
2.17 AMPS 91'-0"TO LOAD CTRSELECTED BY INSPECTION .878Yo DROP
Fig. 21
OTHER BRANCH CIRCUITS-I0 ACTIVE AND 2 SPARE WITH TOTAL LOAD OF 61.83 AMPS AT 115 VOLTS INCLUDING THE SPARE CIRCUITS (SPARE CKTS ARE AN AVERAGE OF ACTIVE LOADS)
Typical lighting branch circuit for passageways and cargo holds
thermoplastic insulated cable, would be adequate if current-carrying capacity instead of voltage drop determined the cable size. A shorter method of calculating the voltage drop in a branch circuit is based on the assumption that the entire circuit load is concentrated at the "load center" and the fact that the approximate location of this point usually can be determined by inspection of the circuit as drawn on the deck wiring plan. In fixing the exact load center, the load on each outlet as well as its location must be considered; in other words, the load center is analogous to the center of gravity of a series of weights of total value equal to the load on the several outlets. From an inspection of Fig. 21 it is evident that the load center is near the third outlet, 91 f t from the panel. Substitution of this number for L and the total circuit current (2.17 amps) for I in equation (1) shows a drop of 0.878 percent, which is within 0.012 percent of the calculated value. Detailed calculations prove that the load center is actually 92.2 ft from the panel and the use of this value in equation (1) gives the precise drop. Figure 19 shows a typical crew stateroom branch circuit. The total branch circuit load is 465 watts; convenience receptacles are not included in the total load. All conductors are 4497 CM (Navy type cable). In calculating voltage drop, the ceiling fixture loads should be oonsidered as being at the respective switches. The exact method of calculation shows the maximum drop to be 1.65 percent at 91.8 f t from the distribution panel: Assuming the load center to be at the second ceiling fixture in the stateroom, the short method gives a drop of 1.53 percent. The method of determining the size of the feeder and the total drop to the switchboard would be as described for Fig. 21. 4 Table 5 i$ a typical tabulation listing a lighting feeder and the branches it supplies; it Also notes all circuit characteristics including calculated percent voltage drops and selected cable sizes. This type of tabulation is prepared for the entire lighting system, and includes every lighting feeder and branch circuit on the ship; the
I
I
complete tabulation is usually identified as the "list of lighting feeders and mains. " b. Power Circuits. In contrast with lighting conductors which are usually fixed by voltage drop, the size of power conductors is almost invariably determined by carrying capacity. As an example, assume that a 240volt d-c, ten-circuit power panel supplies a group of seven motors, one of which is rated 50 hp, three at 25 hp, two at 10 hp, and one at 5 hp. The three spare switches are rated a t 100 amp, 70 amp, and 50 amp respectively. The feeder length from the switchboard to the panel is 400 f t and the longest branch is 150 f t and supplies a 25-hp motor. The total motor current, including 125 percentvofthe largest motor full-load current, is approximately 596 amps and the allowance for the spare switches is 110 amps; therefore the feeder rating should be 706 amps. This is greater than the capacity of the largest conductor usually installed; therefore each leg of the feeder could consist of two 400,000 circular-mil cables in parallel. The rated current of the branch serving the 25-hp motor is approximately 115 amps (125 percent of the motor rating) and the branch circuit could be No. 00 AWG (133,000 circular mils) twin-conductor cable. From equation (1) the voltage drop in the feeder is Voltage drop
=
(200) (706)(12)(400) (soo,ooo)(240)
=
3.53 percent
and the voltage drop in the branch circuit is Voltage drop
=
(200)(115) (12)(150) (133,000)(240)
=
1.29 percent
therefore the total drop ffum the switchboard is 4.82 percent. Lengths as great as those assumed in the foregoing are seldom, if ever, encountered on an ordinary ship; hence it is obvious that, as previously stated, conductor sizes are usually determined by carrying capacity rather than by voltage drop. Nearly all merchant ships with an %C electric plant utilize 450 volts as the generated voltage. The probability of conductor sizes being dictated by voltage
Table 6
L-Cene-tor No. l-nhip's service
Yemore conneotionbox -us tie to emergency d b . Main wnd. oirc. pump no. 1 Main wnd. cim. pump no. 2 Main cond. cnda. pump no. 1 Main wnd. cuds. pump no. 2 Forced-draft blower no. 1 Forceddraft blower no. 2 Lube-oileervicepump no. 1 Fuel-oil service pump no. 1 Fuel-oil service pump no. 2 Aux. cond. circ. pump Aux. cond. cnde. pump Fire, bilge & olean bdllrat pump Panel P14 mech. aux. 1P14 Saltwater semce pump no. 1 &PI4 Bilge & oily ballaat pump , no. 1 3P14 Engine room bilge - Dump . no. 1 4P14 Bilge & ballaat priming pump no. 1 5P14 Control air compressor 6P14 .LPsteam pen. feed pump no. 1 7P14 36" main cond. ovbd. valve 8P14 Atmw. cleen draii tank pump no. 1 9P14 28" main oond. high suotion valve 10P14 Stsck vibrator (4) -
1P15 2P15 8P15 4P15 5P15 6P15 7P15 8P15 ,9P15 10P15 llP15 i2pi6 13P15 14P15
-
10" aux. SWsystem valve Spare Spare Spare Penel PI5 mmh. aux. Saltwater service pump no. 2 Bilga & oily ballaat pump no. 2 .,' ~ - e robm bilge pump no. 2 Bilge LbdlaBt priming pump no. 2 Fuel-oil traafer DumD Boiler cold s t a r t ~ u m i Lube-oil pwiEer 28" mein dond. low suction wlve Freshwater tnrnefer pump & priming Pump 6"6re. bilge &clean ballast dichg. valve F r d l w a k r DUmD no. 1 b h ~ a t ~priminp pump Spare 8-
PUIUP
-
DC
Hp
PF
Load
Workii Load
T
Y
CIB. BKB.
21' 6" h t fr. 137 8 ,440-34
BBMOd
llP14 42P14 -PI4 14P14
AC
-AM-EBFull
Smam Cm. BKB. LEAQTE, E ~ M E A T ft
~
- --Cm. MILE
AND
,-mA
21'6'' h t fr. 147 8 bridge dk. fr. 163 8 .
2000A lw0A 250A 175A 175A BOA BOA 225A 225A 70A 40A 40A 30A 20A
8' 0" level fr. 134 8 8' 0" level fr. 138 8 8'0" level fr. 132 P 8' 0" level fr. 132 P upper dk. fr. 148 8 upper dk. fr. 148 P 8' 0" level fr. 148 8 8' 0" level fr. 148 8 8' 0" level fr. 150 8 8' 0" level fr. 145 8 8' 0" level fr. 142 S 8' 0" level fr. 150 P 440-34 8' 0" level fr. '131 8 440-34
125A 250A
1,48)350.000M-T-350 -(8)350.000 ~WT-350 (3)400,000 3-T-400 350,000 TSMI 168,000 T-168 168,000 T-168 41,700 T-41 41,700 T-41 300,000 T-300 300.000 T-300 33,100 T-33 9,016 T-9 8,016 T-9 9,016 T-9 8,016 T-Q 83,700 250,000
T-83 T-250
8' 0" level fr. 143 P
440-39
40A
9,016
8' 0" level fr. 148 P
440-34
MIA
16,500
T-16
8' 0" level fr. 152 8
440-34
15A
4,487
T-4
8' 0" level fr. 153 P 440-34 8' 0" level fr. 143 8 440-34
15A 40A
4,497 16,500
T-4 T-16
8' 0" level fr. 151 P 8' 0" level fr. 136 P
440-34 440-34
30A 15A
8.016 4,497
T-9 T4
8' 0'' level fr. 132 S
44-34
20A
8,016
T-9
15A 15A
4,497 4,497
T-4 D-4
15A 15A 50A 15A 350A
4,497
T-4
400,000
T-400
40A
9.016
T-Q -
50A
16,500
T-16
8' 0" level fr. 133 8 440-34 21' 6" flat fr. 147 8 440-14 8' 0" level fr. 142 S
440-34 440-34 440-34 440-34 8' 0" level fr. 143 P 440-39 8' 0" level fr. 144 P
440-34
8' 0" level fr. 148 P
440-34
8' 0" level fr. 154 P
440-36
lWA
15A
8'0" level fr. 153 P 8' 0" level fr. 150 P 21' 6" level fr. 147 P 8'0" levelfr. 148 S
440-34 440-34 440-39 440-34
l00A lOOA l00A l00A
15A BOA 15A 15A
8' 0" level fr. 135 9
440-34,
lOOA
15A
8' 0" level fr. 152 P
440-34
lOOA
70A
8' 0" level fr. 151 P 8' 0" level fr. 149 P
440-34 440-34
15A 30A
8'0" level fr. 154 P
440-34 440-34 440-34
15A 90A 15A
-
L C A B7 ~ % Vom ?BOP SIQ Feeder M ~ U L Totd
i
A n e r s t o r No. 2-ahip's
-
-
VOLTS
7
LOCATIOH
Typical List of Power System Feeders and Mains
T-9
ELECTRIC PLANTS
MARINE ENGINEERING
circuit protective devices, a fault-current analysis of the entire electrical generating and distribution system should be made. I n calculating the total magnitude of fault currents, it is necessary to determine not only the contribution of short-circuit current from the generators, but also the contribution from motors connected to the system. The contribution from induction motors decays very rapidly; however, the time of decay usually spans the time range of circuit breaker operation and should be considered. The fault-current analysis should be based on the total number of generators, including spare units, that may be operated in parallel, the number of motors expected to be operating, and the reactance and resistance of cables and transformers in the circuit in question. The currents obtained are an average of the maximum asymmetrical rms currents of the three phases at % cycle after the inception of the fault; these currents determine the interrupting rating required for the protective devices. The value of fault current may be calculated from the ex~ression Short-circuit current = EnK (4) 2, where Fig. 22
Ratio of average rms current in three phasos at one-half cycle to rms value of symmetrical current
drop with this voltage is much less than that with the previously discussed 240-volt d-c system; however, in order $0 prove compliance with classification society rules, all distribution circuits should be checked for voltage drop. Table 6 is a typical tabulation listing of a power feeder and the mains it supplies. Table 6 also no$es circuit characteristics including calculated percent voltage drops, selected cable sizes, and the size and type of circuit breaker and element for each main and the feeder. This type of tabulation is prepared for each power system circuit on the ship; the complete tabulation is usually identified as the "power system ligt of feeders and mains. " 6.7 Short-Circuit Analysis, and Selectivity and Coordination of Circuit Protective Devices. The steady
growth in the use of electric power on merchant vessels, with the resulting increase in the capacity of the generating plant, has resulted in a considerable increase in the mag'njtude of possible short-circuit currents throughout the electrical distribution system. To maintain continuity of electrical gervice, with the least possible interruption from fault currents, it is necessary to provide adequately rated circuit protective devicesproperly coordinated with each other-throughout the distribution system. These protective devices are usually circuit breakers; however, fuses may be used for many applications. To determine the proper selection and application of
En = rated line-to-neutral voltage of generator (260 volts for a 450-volt generator) K = Asymmetry factor dependent on ratio of total reactance to total resistance in cii-cuit (see Fig. 22) Zen = equivalent system impedance in ohms per phase between point of fault and source of current As an example, the following calculations are for a shipboard electric plant with two 1250-kw, 450-volt, %-phasegenerators operating in parallel, and an induction motor load of 1800 amps a t the time of a fault. A three-phase fault is assumed to occur on the generator ' switchboard main bus. System characteristics are: gene~atorsubtransient reactance = 0.0285 ohms = 0.0087 ohms generator armature resistance = 0.0002 ohms generator cable reactance = 0.0003 ohms generator cable resistance By combining the foregoing -values, the total resistance and total reactance of one generator and its cable to the ** switchboard main bus are: resistance = 0.0090 ohms/phase reactance = 0.0287 ohms/phase For two generators operating in parallel, and including generator cables, the impedance is
+
Zen = d(0.0045)~ (0.0143)2 = 0.015 ohms The ratio of reactance to resistance is 3.18, and K, in equation (4), as read from Fig. 22, is 1.06. Substituting the values for K and Z,, in equation (4), gives the total generator contribution to the fault:
Short-circuit current =
(260)(1.06) = 18,373 amps (0.015)
The average asymmetrical motor contribution to the fault is equal to approximately 3.5 times the motor current, which is (3.5) (1800) or 6300 amps. Adding the motor contribution to the generator contribution gives a total short-circuit current available a t the switchboard main bus of 24,673 amps. This value is the minimum interrupting rating for the circuit protective devices installed on the generator switchboard. Figure 23 shows maximum fault currents at different points of the distribution system used in the previous example. The fault-current analysis should be extended to include calculations of minimum fault currents for remote points of the system to determine that a sufficient current is availabwa mcure the proper tripping of each _..-_---device. Under this condition, the smallest protective generator which can be connected to the system, with negligible motor contribution, is used for the calculations. Using the fault-current analysis as a basis for the selection of protective devices, a sequence of circuitbreaker tripping can be determined that will isolate any fault in the distribution system with a minimum interruption of power to other services. In the event of a fault, the nearest protective device on the supply side of the fault.should open to isolate the faulted circuit; other protective devices on the supply side of the fault should remain closed. All circuit breakers and fuses in the distribution system should have an interrupting rating equal to or greater than the available shortcircuit current a t its paint of application. Breakers ..___ .--trip devices should have a3hortequipped with selective time fatlKggequal to or greater than the short-circuit current at 'iis"point of application. The basic selective overcurrent tripping characteristics used on marine circuit breakers are long delay, short delay, and instantaneous. The long-delay element provides overload protection and operates in seconds or minutes a t low multiples of its coil rating; a breaker having this type of overload protection should not be applied above its continuous current rating to avoid exceeding its thermal capacity. The short-delay element operates in a few cycles and provides protection against high values of fault curreht; this type of delay also allows the circuit breakers connected in series to operate. The instantaneous trip element introduces no intentional time delay and may be used to provide short-circuit protection to load circuits. n
' TWO 1250 KW, 4 5 0 V, 6 0 CPS GENERATORS
9 MAIN SWBD 24.700AMPS
SHORE SUPPLY
I)
A
Q
, '
I
I
ENGINE ROOM VITAL AUXlLlARlESl POWER PANEL\
1
4
16,000 r' AMPERES
I
I
I
I
UlAJ
Cr(n\
I
, -
REFRIGFRATION GROUP CONTROL CENTER
, AMPERES
LIGHTING BUS\
1 1,000 AMPERES
EMER SWBO 17,000 AMPERES
I, 30,450/120 V TRANSFORMER
\A
EMERGENCY LTG BUS 5 , 0 0 0 AMPERES
-
Fig. 23
1
1
6' I,
--
150KW 4 5 0 " ~OCP~,EMER OlESE DRIVEN GENERATOR
Fault currents at various points of o distrib~tionaystem
To properly determine circuit breaker selectivity, a set of time-current curves should be prepared. These curves should show the tripping characteristics for the various types of circuit breakers involved, and also the generator decrement curves. -Regulatory body rules require that the long-time setting of the generator circuit breaker should not exceed 115percent of the full-load current of the generator. The short-time setting of the generator breaker should be set to trip a t the lowest values of current and time which will coordinate with the trip settings of the distribution system feeder breakers. Where three or more generators are arranged for parallel operation, each generator circuit breaker should be provided with an instantaneous trip set a t a value in excess of the maximum asymmetrical short-circuit current available from its associated generator. Using the curve of the generator breaker as a base, curves of the other breakers are added to ensure that there is no overlapping of tripping characteristics, thus establishing proper selectivity between circuit breakers. Selective circuit-breaker operation should be provided between the generator, bus tie, bus feeder, and feeder protective devices. For vital services, selective tripping should be extended to include feeder and branch circuits.
654
MARINE ENGINEERING
Section 7 Interior Communications 7.1 General. Interior Communications are those means used for signalling or transferring information from the source to one or nlore points by indicating, recording, voice, or audible equipment. This includes alarm systems, telephone systems, navigational sensor systems, and elements of the ship's control systems. The legal requirements stipulated by the U.S. Coast Guard and the Federal Communications Commission must be observed in the design and installation of these systems. The U.S. Coast Guard requires that certain critical circuits be supplied with power from emergency or battery sources and recommends that others be connected to the emergency supply if sufficient capacity is available. These requirements lead to a variety of supply systems. Generally, the IC loads associated with the machinery space are supplied from a local IC power panel with an emergency power feeder. Those circuits associated with the wheelhouse are generally connected to the emergency supply via an IC and electronics panel located in, or in the vicinity of, the wheelhouse. Unique systems, such as the gyrocompass, may have a requirement for electrical power not available from the ship's supply or the switchboard. In this case, the equipment will include power supplies to convert the ship's power to the required characteristics. The interior communication circuits are normally identified by alphanumeric symbols similar to those adopted by the U.S. Navy and IEEE. Symbols identify the circuit primary function and simplify references on plans and correspondence. Interior Communication systems can be classified as indicating, alarm, communication, and control systems. Each of these classifications and the types of circuits which they include are discussed in the remainder of this section. 7.2 Indicating Systems. Indicating IC systems transmit status information to areas of control aboard the ship. The receiving device may be a dial, a light, a digital display, or an input to a unit of control equipment. Indicating systems are also considered to include the sensors, such as the underwater log or gyrocompass initiating the information. Typical indicating systems are as follows: a. Refrigeration Temperature Indicating System (Circuit RT). Vessels with refrigerated compartments are required to have a refrigeration temperature indicatr ing system which indicates the temperature maintained within the refrigerated spaces. Temperature sensors, strategically placed in the areas, are connected to indicators located on an indicator panel outside the area. Variations to the basic system consist of added alarm devices to alert the crew before damaging temperatures are reached. Recorders can also be provided to give a printed record of temperature readings versus time.
b. Shaft Revolution Indicator System (Citcuit K). The shaft revolution indicator system indicates the direction of rotation, speed in rpm, and the cumulative revolutions of the propeller shaft. A transmitter is coupled to each shaft. Its output is generally fed into a combination revolution counter and rpm and direction indicator located in the machinery space and into rpm and direction indicators located in the wheelhouse and other miscellaneous spaces as desired. c. Bearing Temperature Monitoring System (Circuit TM). The bearing temperature monitoring system is used to continuously monitor the temperature of selected bearings in the main propulsion plant. A sensor is installed for each bearing to be monitored, and is electrically connected to the monitoring alarm and indicating equipment incorporated in a status panel located in or near the ship's control panel. d. Rudder Angle Indicator System (Circuit N). The rudder angle indicator system provides a means of indicating, a t remote stations, the angular position of the rudder; the system normally consists of a waterproof enclosed transmitter, located in the steering gear room, and waterproof enclosed indicators, located in the wheelhouse and other selected stations. The transmitter is a synchro generator and should be connected to the rudder post through a mechanical linkage such that it will transmit the actual angular position of the rudder to each synchro indicator. The indicators generally consist of a dial containing a fixed amidship line and a moving pointer deflected left or right of the line to indicate left or right rudder angle. The system supply is usually 115 volts, 60 cps, single phase, AC, and is often supplied from the emergency switchboard. e. Pyrometer Indicator System (Circuit PB). Boiler flue temperatures are monitored by permanently installed thermocouples mounted in the boiler uptakes. Each thermocouple is contained in. a protective tubing, and by means of a rotary selector switch it may be connected to an indicator mounted locally and/or a t the engineering operating station. f. Salinity Indicator System (Circuit SB). The salinity indicating systems provide a means of measuring the degree of salt content of the w&er in potable water systems and in boiler feed and condensate systems. Separate systems are usually ~rovidedfor each application, one for each desalination plant and one for the boiler feed and condensate system. Each system consists of salinity cells with valve assemblies, in number as required, and an indicator panel. Salinity cells and valve assemblies are designed so that the cells may be removed and replaced without interruption to a continuously operating piping system. T O facilitate replacement and servicing, each cell is energized through a watertight plug and receptacle.
ELECTRIC PLANTS
Indicator panels are normally arranged with an individual control section for each associated cell, a common audible alarm, and a meter for reading the salinity content a t any 'cell by means of a cell selector switch. Individual control section circuitry is such that when the salinity content of the system, a t the point being monitored, reaches a predetermined value, the monitoring cell alarm light and the common audible alarm are automatically actuated. A silencing switch is provided in each individual cell section to deenergize the common audible alarm; operation of this switch should not affect the aural or visual signalling capability provided for other cells. Each cell alarm light is arranged to give indication of alarm until the salinity is reduced below the alarm setting. Salinity systems are normally set up to alarm when the salt cantent reaches 0.25 grains of sea salt per gallon. Salinity systems that monitor the desalination plants also incorporate a solenoid-operated dump valve that will automatically dump, to the bilge, or by-pass all processed water that exceeds the predetermined salt content; dumping or by-passing will also take place upon loss of power to the salinity indicator, the dump valve solenoid, or the desalination plant pumps. Dumping for any reason should actuate the common alarm on the indicator panel. The feed and condensate salinity system is usually identified as circuit 1SB and the desalination plant salinity system as circuit 2SB. g. Underwater Log System (Circuit Y). The underwater log system measures the ship's speed through water and the distance traveled. A rodmeter projects through a sea valve and extends approximately 30 in. below the shell of the ship. The rodmeter contains a button electrode on each side. As water flows past the electrodes, a voltage is generated proportional to the ship's speed. The signal generated is amplified and drives digital speed indicators a t various locations on the ship. The system is not normally provided on most merchant vessels due to its high installation and maintenance costs. h. Doppler Sonar Speed Log (Circuit Y). The doppler sonar s p d log is an electronic system for accurate measurement of the ship's foreand-aft velocity relative to the bottom a t depths to 500 f t and relative to the water mass a t greater depths. Operation is based on the principle that a signal transmitted from a moving object and reflected back from a stationary surface will indicate an apparent shift in frequency. This frequency shift is proportional to the velocity of the moving object in relation to the stationary surface. Its principal uses are for "speed trials" and for continuous monitoring of the ship's velocity in coastal waters and harbors. i. Gyro Compass System (Circuit LC). True north and, if desired, ship's roll and pitch information are supplied by the gyrocompass system. The compass h e a d i i information is furnished to steering and bearing repeater compasses located a t various navigational
655
stations and to other electronic navigating equipments, such as RDF, course recorder, and the radar displays requiring this information. Roll and pitch information can be supplied to stabilizing equipmenta if desired. 7.3 Alarm Systems. Certain alarm and warning systems are essential for the safety of the ship equipment and ship personnel; the deaign and function of such systems are closely controlled by the U.S. Coast Guard. The alarm indication may be visual or audible. When two or more ~udiblealarms are required in a single space, it isd~bmmonpractice to provide a common alarm panel which provides one common alarm with visual indications of the alarm source., A brief description of typical alarm circuits is as follows: a. Lubrication-Oil Low-Pressure Alarm System (Circuit EC). The lubricating-oil low-preasure alarm circuit provides a warning alarm when the pressure in the lubricating-oil lines to vital machinery becomes dangerously low. The alarm is actuated by pressureoperated contact switches installed in the lubricating-oil pipeline. The switch contacts are held open as long as a predetermined pressure in the line is maintained. When the pressure falls below the acceptable level, the switch contacts close, energizing the alarm signal in the machinery control spaces. b. Feedwater Low-Level Alarm System (Circuit FW). The feedwater low-level alarm system sounds an alarm a t the propulsion control station when the level of the water in the tank supplying the boiler feed pumps becomes low. The equipment consists of a floatactuated switch located a t $hidfeedwater heater or tank, alarm bells, indicating lights, and cutout switches. A low level of water causes the switch contacts to close, energizing the alarm indicators. An added feature to this circuit can provide for the cutoff of fuel oil when the feedwater reaches a dangerously low level. c. Fire Alarm System (Circuit F). The fire alarm system detects the presence of fire in the protected compartments or spaces and transmits an alarm signal to the wheelhouse or fire control station where audible and visual alarm indications are provided. The detecting system is divided into zones to restrict the area covered by any particular alarm signal. The system is designed so that the officer on watch can determine from the fire alarm switchboard the compartment or zone in which the fire exists. The dktecting system may be electric or pneumatic. The electric system uses mercury or bimetallic temper* ture detectors with an alarm switchboard. This system is generally supplemented by manually operated contact makers connected into the system to provide an alarm to the switchboard. The pneumatic system samples air drawn from protected areas to determine if smoke is in the space. Photocells and light beams connected into an alarm board provide an automatic alarm when smoke is detected in the sample being checked. Provisions are made for automatic sampling of the protected spaces by such means as relays and timing devices.
MARINE ENGINEERING
-
On passenger vessels a fire alarm system is required by the U.S. Coast Guard and the system design and components require Coast Guard approval. d. General Alarm System (Circuit G). A general alarm system, the purpose of which is to warn all persons on the ship of an emergency, is required on all manned vessels of over 100 gross tons. The system consists of electric vibrating alarm bells, manually operated contact makers and distribution panels. The alarm bells are energized by manually operating a normally open, spring-return-to-normal contact maker to the "alarm" position. The number of contact makers and their location should comply with the applicable U.S. Coast Guard requirements. To prevent unauthorized operation of the contact makers located outside the wheelhouse, they should be mounted within an enclosure having a breakable transparent window. The alarm bells should be at least eight inches in diameter, and produce a distinctive tone different from all other bells on the vessdl; these bells should be located throughout the vessel in passenger and crew living areas, work spaces, and machinery spaces in such numbers and locations as to assure that all persons can hear an alarm bell. In spaces where the ambient noise level is unusually high, the alarm bells within the spaces should be augmented by flashing red lights. All general alarm bells are supplied by feeders from a feeder distribution panel through branch circuit distribution panels. The feeder distribution panel is supplied from the system storage battery source. For vessels that are subdivided into zones by fire-screen bulkheads, at least one feeder should be provided for each zone, as necessary, to supply the bells between adjacent firescreen bulkheads. For vessels that are not divided into fire zones, the vessel should be divided into vertical zones, not exceeding 150 ft in length, and at least one feeder should be provided for each zone as necessary to supply the bells within the zone. Distribution feeder and branch circuit panels should be located above the bulkhead deck or above the freeboard deck, whichever is the higher, and outside the machinery casing. These panels should be provided with overcurrent protection for each feeder and branch circuit. Disconnect switches are not provided. Branch circuit distribution panels are provided as necessary for each zone. At least one fused branch circuit should be provided for each zone deck level. No more than five general alarm bells should be connected to one branch circuit, and a branch circuit should not supply alarm bells on more than one deck level. The power supply source for the general alarm system is either a single storage battery or duplicate storage batteries, located in a well-ventilated battery locker or battery room that is above the bulkhead deck or freeboard deck, whichever is higher, and outside the machinery space casing. On vessels having only one general alarm battery, the battery must be maintained in a fully charged condition a t all times by an automatic
charging panel. On vessels having duplicate storage batteries, the batteries should be connected, through a two-position transfer switch (no "off" position), so that one battery will be on charge while the other battery is set up to furnish power to the systems involved. For specific details regarding battery loads and capacities, refer to the U.S. Coast Guard lrequirements. For vessels having a temporary emergency bus on the emergency switchboard, the alarm system may be supplied from this bus in lieu of being supplied from separate general alarm batteries. The entire general alarm system including all component equipments are subject to compliance with U.S. Coast Guard requirements. e. Refrigerated Spaces Alarm System (Circuit RA). A refrigerated space alarm system is required in any refrigerated space, accessible to ships personnel during a voyage, that can be locked from the outside such that egress from the space is impossible. Each space is fitted with a waterproof pushbutton electrically connected to an audible alarm located outside the space unless a mechanical pull operating a jingle bell is provided. 7.4 Communication Systems. Communication systems fall into two general categories: those required for safety of the ship, and those installed as a matter of convenience. Communication systems aboard ship are typically as follows: a. General Announcing System (Circuit MC). All ocean and coastwise passenger vessels certified to carry 500 or more persons, including officers and crew, and all passenger vessels whose lifeboats are stowed more than 100 ft from the navigating bridge are required by the Coast Guard to have a general announcing system. This system enables an officer on the bridge to transmit orders or information to all lifeboat stations, all lifeboat embarkation stations, all public spaces used for passenger assembly stations, and throughout the passenger's and crew's quarters. The system is controlled from a single location on the bridge. As directed by the Coast Guard Commandant, certain loudspeakers are arranged for two-way conversation between the bridge and the lifeboat and embarkation stations. The distribution system, location of speakers, and the equipment used are subject to U.S. Coast Guard approval. b. Ship Service Telephone System (Circuit J). Ship service telephone systems %re used for general telephone communication, and are similar to shore installations. They may be either automatic dial or central switchboard control types. TrunMines are provided for connectipg to shore telephone systems when the vessel is in port. c. Sound-Powered Telephone System (Cicuit JV) and Voice Tube Systems. The Cosst Guard requirements specify that vessels shall be provided with an efficient means of communication between each of the following spaces: Wheelhouse
ELECTRIC PLANTS
Engine room Steering gear room After steering station In addition, the requirements specify that efficient communication be supplied between the following: Master gyrocompass and remote wheelhouse repeater compass Wheelhouse and remote radar plan position indicator Wheelhouse and remote emergency squad equipment storage spaces Wheelhouse and remote smoke detector cabinets Wheelhouse and bow or forward lookout station Wheelhouse and remote radio room and emergency radio room Wheelhouse and radio direction finding equipment
It should be noted that for this requirement, equipment located in a space which opens into the wheelhouse is not considered remote. These requirements are generally met with soundpowered telephone circuits 1JV and 2JV. Soundpowered telephones are self-contained in that they require no external power. This fact adds to their reliability. Most installations now employ the magneto ringing type and thus avoid the requirement for the external ringing circuit "E." I n noisy areas the magneto ringing circuit is used to operate a relay which controls an electric-powered Maxon. Installation costs have practically eliminated the use of voice tubes to fulfill these requirements. Their use is now restricted to short runs where the installation costs are competitive or the use and storage of sound-powered phones is impractical. The system design and the equipment used in this system are subject to Coast Guard approval. d. Sound-Powered Telephone Call Bell System (Circuit E). Sound-powered handset stations may be alerted by a call bell system. The calling station operates a switch which actuates an alarm at the station beine called. e. Call Bell System (Circuit A). Call bell systems are usually provided on all vessels carrying passengers; these systems provide a means for passengers to call for assistance, when needed, and normally consist of pushbuttons, buzzers, and annunciators. A pushbutton should be located in each passenger stateroom, convenient to the head of each berth; operation of the p~shbutton will cause a buzzer to sound, and an annunciator to indicate the stateroom initiating the call. Annunciators should be located where someone is always in attendance. The power supply should be 115 volts, 60 cps, single phase, a-c from the emergency lighting system. f. Ship's Entertainment System (Circuit SE). Circuit SE distributes speech, radio, and taped programs in the passenger and crew quarters where terminals exist. A centralized control station, with amplifiers and receivers, broadcasts the program collectively, or in
-
657
groups, over a series of reproducers. The system components vary to suit the type of vessel. g. Ship's Television System (Circuit TC). The ship's television system provides an outside antenna connected to outlets located in various spaces as desired. The system may also include an omnidirectional or rotating antenna as well ag a radio frequency signal booster. 7.5 Confrol Systems. Control systems are used for remote contro~.'of the ship's functions from the pilothouse or other designated control areas. They may transmit specific orders to be initiated manually, or they may power synchro or servo loops that actually initiate the necessary action. Typical control circuits are as follows: a. Engine Order Telegraph System (Circuit MB). The U.S. Coast Guard rules require that all self-propelled vessels have an efficient method of transmitting engine orders from each ship control station to each engine room along with a method of transmitting acknowledgment of engine orders from the engine room to each ship control station. This may be accomplished by mechanical, electrical, or other methods, each being subject to approval of the Coast Guard. However, for most vessels an electrical system is selected because of its reliability and overall economy in first cost and maintenance costs. The electrical system is designated as "MB," and is designed for operation on 115-volt, 60-cps, single-phase, a-c power; the source of supply is often from the emergency switchboard. An electrical engine order system could consist of (a) combined-type transmitter-indicator telegraph instruments located in the wheelhouse, engine room, and other ship control stations on the vessel; (b) a wrong-direction alarm unit located in the engine room; (c) an alarm bell located in the engine room; ( d ) A power-failure alarm panel located in the wheelhouse; and (e) a transfer relay panel located in the wheelhouse. Each telegraph instrument is fitted with a transmitting synchro and an indicating synchro. To initiate an engine order from a ship control station instrument, the transmitter handle (and synchro) is operated to the selected order position. This positions the transmitter synchro out of synchronism with the engine room indicator synchro and causes the engine room indicating synchro to rotate to the same relative position as the transmitter synchro, thus indicating the desired engine order. Acknowledgment of the order is accomplished by matching the engine room telegraph instrument transmitter pointer (an&- synchro) with t h e order pointer, which causes the ship control station indicator synchro (and pointer) to match its transmitter pointer nosition -4n audible alarm at each transmitter and indicator instrument rings when an order is transmitted and continues to ring until the order is properly acknowledged. The wrong-direction alarm unit consists of an audible alarm, a visual light, and a contact maker arranged to
-
658
MARINE ENGINEERING
provide alarms when the engine room throttle valve is operated in the wrong direction from the transmitted order. The power-failure alarm panel should be arranged to aut~matically sound an audible alarm and indicate visually when the ~ystempower source fails; also, a switch should be provided to silence the audible alarm. The transfer relay panel is arranged electrically, so that only one engine order transmitter instrument may be connected electrically to the engine room indicator instrument a t any one time. The reply pointer of all transmitter instruments should operate in synchronism a t all times. Passenger vessels of 20,000 gross tons or over that are equipped with an electric engine order telegraph system are required to have a standby engine order telegraph system, either electrical or mechanical. The entire engine order telegraph system including all component equipments is required to be watertight and is subject .to compliance with U.S. Coast Guard requirements. b. Whistle Operator System (Circuit W). The whistle operator system may be used in the "at will" or "automatic" mode of operation and from various conning stations. Moving a control switch to the "at will" position closes the circuit to a solenoid which opens the whistle valve or closes a contactor. Moving a switch oq the whistle timer control panel located in the pilothouse to the "automatic" position energizes a coding timer motor driving a mechanical cam that intermittently closes the whistle solenoid circuit. The coding timer can be set for 1, I%, or 2 min periods as required by international navigational rules. Mechanical whistle pulls are installed a t the navigating stations for emereencv - < use. c. Automated A number of IC circuits are included in the ship's automated control system; however, they will not be reviewed here as they are discussed in Chapter 21. d. Steering Control System (LP). The U.S. Coast Guard rules require that a means be provided for controlling the main steering gear from the pilothouse and from an alternative steering station; this means of control may be either mechanical, hydraulic, or electrical. The alternative steering station is required to be located on the after weather deck unless duplicate pilothouse controls are provided, in which case an alternative steering station on the after weather deck is not required, and the steering gear room may be considered as the alternative steering station.
-
-
ELECTRIC PLANTS
On most vessels, two independent electrical steering control systems are provided. These systems usually consist of a common (dual) steering stand, located in the pilothouse, and two control units, one for each main steering unit, located in the steering gear room. Each control system is arranged to transmit an electrical rudder order signal from the steering stand to its associated control unit. The control units, operated by hydraulic servo power, stroke their respective main hydraulic steering unit which moves the rudder to the ordered position. When the rudder reaches the ordered position the electric order signal is nullified, through potentiometers or synchros located in the steering stand and the control units, and the rudder is stopped at that position. Three modes of electric steering control operation are normally provided; automatic, hand-electric follow-up, and non follow-up. Selection of the steering control mode is made at the steering stand by means of a selector switch. In the automatic mode of operation, steering is controlled by the master compass, which senses any deviation from the ordered course and feeds a course error signal into the system to move the rudder and bring the ship back to the ordered course. The ordered course is set up by turning the steering wheel to the desired course position; no further adjustments are required until the ordered course is changed. In the hand-electric follow-up mode of operation, the steering wheel is turned manually to the desired rudder position. The rudder will stop a t that ordered position and remain so until the steering wheel is turned to another desired rudder position. In the non follow-up mode, the rudder is moved by port or starboard manual operation of a non follow-up controller lever. This system is arranged so that the rudder will stop moving when the lever ys released. The steering stand for dual control systems is provided with a selector s ~ t c hfor selecting either the port or starboard control system; this switch should also be arranged to start the main steering pump motor associated with the selected control svstem. Widelv se~arated control cable runs should be &stalled from ;he ;teering stand to the steering gear room; these port and starboard control cables are connected to their associated hvdraulic < power control unit located in the steering gear room. Each steering control system ira, supplied 115-volt, single-phase, a-c power from either its associated steering motor controller, or through a transformer fed from the steering gear power panel. Usually, when fed from the steering gear power panel, provisions are made for transfer to either steering power feeder. -
Section 8 Electronic Navigation and Radio Communication 8.1 General. Electronic navigation and radio communication systems are primarily used to ensure the safety of the ship and the ship's personnel. The exact suite of equipment installed is determined by legal requirements and the design function of the vessel. The physical placement of the units and the installation of the associated antennas are unique to each type of equipment, and must be designed to comply with the operational requirements of each device. The rules and regulations controlling the services, type of equipment, and installation requirements for the minimum acceptable suite of navigation and radio communications equipments are issued by the U.S. Coast Guard and the Federal Communications Commission and, in the case of vessels engaged in international voyages, they must conform to the International Safety of Life a t Sea (SOLAS) Convention. Installation layouts must be approved by the U.S. Coast Guard for navigational equipment. The Federal Communications Commission also requires the use of approved radio equipment and some navigational equipment. 8.2 Radio Communication. Radio communication installations consist of receivers, transmitters, and transceivers together with their associated terminals and antenna systems; these equipments are selected on the basis of the requirements of the individual ship and in accordance with the requirements established by the Federal Communication Commission and the U.S. Coast Guard as a minimum. SOLAS Safety Radiotelegraphy or Safety Radiotelephony certificates attesting to compliance with SOLAS requirements are required on vessels engaged in international voyages. Communications are normally maintained by voice or radiotelegraph; the more advanced communication systems may include radioteletype and digital data facilities. The radio installation for a typical vessel consists of the following: 1 A main transmitter, receiver, antenna, and power supply. 2 An emergency transmitter, receiver, antenna, and power supply, all of which must be separate from those required in item 1. 3 A radiotelephone transmitter, receiver with selective ringer, antenna, and power supply. 4 An automatic alarm receiver to monitor the 500kHz international radiotelegraph distress frequency. 5 An automatic alarm keying device to key the main and emergency transmitters with the international automatic alarm signal in case of an emergency. 6 A complete and portable transceiver for use in a lifeboat. On most vessels, the aforementioned radiotelegraph equipment, excluding the antennas and item 6, along
with the necessary components that are required for complete operation and supervision of the radio complex, may be combined into a communication console designed for yse by one operator. This console is located in the
660
s
ELECTRIC PLANTS
MARINE ENGINEERING
equipment provides- reliable radiotelephone communication over manv thousands of miles under normal atmospheric conditions. For reliable short-range communication up to 50 miles, VHF (very high frequency) transceivers are becoming increasingly popular. The units operate in the 156 MHz marine bands and employ the frequency modulation mode of transmission and reception. Selection ringing equipment is also available for this service. A bridge-to-bridge channel (channel 13-156.56 MHz) is incorporated in the VHF transceiver and is used for bridge-to-bridge communication between vessels for navigational purposes only. Channel 16 (156.8 MHz) is the allocated channel for distress, safety, and calling. Radio communication antennas may be of the verticalstanding whip type, or a wire type, depending on the system requirements and the shipboard configuration. Each antenna design and installation for compulsorily fitted equipment is subject to Federal Communication Commission and U.S. Coast Guard approval. A portable radio unit, complete 4 t h hand-cranked generator, is provided on each vessel for emergency use in lifeboats. The entire unit is required to be packaged so that it is buoyant and of a minimum weight; it is normallv stowed in the radio room in a s~ecialrack designei for easy and quick removal. 8.3 Navigation. The navigational systems installed on a ship are determined by the ship's mission. The capabilities of available systems range from position location, with an adcuracy of feet, to general position information, accurate to plus or minus 5 miles. Ranges of effectiveness vary from line-of-sight for certain station-keeping systems to a system that will allow a properly equipped vessel to determine its position with an accuracy of f mile anywhere on the surface of the earth. Ultraprecise and short-ranged systems are normally used by dredges, well-drilling ships, survey ships, and are temporarily installed on large ships during sea trials and performance tests. The systems may use the principle of determining the bearing angle of the navigating station to known points, determining the distance of the navigating station to known points, or a combination of both methods. Radar is also commonly used for navigating purposes when it is necessary to know the location of the ship in-relation to surrounding shorelines, buoys, and other ships. 8.4 Radio Direction Finding. A radio direction finding system aboard a ship provides a means of establishing the ship's position based on the location of two or more fixed transmitting stations. Radio direction finders are used for both navigation and for the safety of life a t sea involving search and rescue of ships or lifeboats in distress. A typical shipboard system could consist of a radio receiver-indicator located on the ship's navigating bridge or in the chart room, a fixed cross-loop antenna, and a sense antenna. The receiver-indicator is tuned to receive the specific frequency signal of the selected transmitting station. The relative intensity of this received signal is reduced by
rotating a movable coil, within the receiver, with respect to the fixed-loop antenna until a null position is realized in the signal received from the transmitting station. This null position indicates the bearing a t which the system is in electronic alignment with the transmitted wave front of the radio frequency transmission from the transmitting statioh. The indicated bearing from the ship to two or more fixed stations is noted and the reciprocal lines are plotted on an appropriate chart. The position of the vessel is indicated by the intersection of these lines. The sense antenna provides a means of preventing the 180-deg ambiguity that could result in bearing readout if the received signal is not properly oriented. The main radio transmitting antenna should be grounded (inoperative) when the radio direction finding system is in operation to prevent distorted signals from causing maloperation and to reduce error in the direction finding system; this grounding is usually accomplished at the main radio console through a switching unit that disables the main antenna and also provides power to the radio direction finding system. Also, special~onsiderations regarding guywires, stays, handrails, etc., must be taken into account and compensated for by proper design during installation of the radio direction finder system. The accuracy of radio direction finder observations is affected by fluctuations in the ionosphere and consequently vary in accordance with the time of day and time of year. Daily distortion of the sky wave component of the radio transmission known as "night effect" can cause considerable error in the plotted position. The shipboard power supply for the radio direction finder is usually 115 volts, single phase, 60 cps, a-c and is taken from the main radio console power supply through the switching unit noted in the foregoing. 8.5 Radar. Radar systems aboard a ship normally operate at wavelengths of 3.2 and 10 centimeters. They consist of a transmitter-modulator, receiver, a rotatable directional antenna (scanner), required power supplies, a master indicator, and remote indicators (when desired). The radar video presentation may take many forms. The most popular method is by the Plan Position Indicator (PPI) which displays a map presentation of an area 360 deg around t h e ~ h i p . The radius of this presentation can be varied by the operator from close ranges to the maximum capability of the radar. The maximum range capabilities of a radar system is a function of frequency and transmitter power, but is primarily controlled by the line-of-sight characteristics of microwave frequencies. Radar is quite important to safety at sea. The video presentation displays other ships, buoys, islands, and other navigational hazards and shows their relationship to the navigating station. Introducing compass information into the radar system display will permit the observed bearings to be measured as true north, magnetic- north, or relative
66 1
Fig. 24 Typical liner of position from two pairs of Loran stations
quantities. Circuitry can be provided to give a true motion presentation on the P P I tube; that is, echoes of stationary objects remain stationary and echoes of moving objects move. Normally, without this circuitry, all objects on the PPI tube appear to move because of the ship's motion. Many large ships use two radar systems, one operating at 3.2 centimeters and the other at 10 centimeters wavelength. '8.6 Hyperbolic Systems. A popular method' of navigational positioning measures the difference in distance between the navigating station and two or more known fixed stations. A navigating station positioned
between two stations a t known geographic locations can measure the difference in time of arrival of a radio frequency transmission from the two stations if their transmissions are synchronized. A line connecting all the positions between 'khe two transmitting' stations where the same time difference, or delay, would be measured will describe a hyperbola, having the transmitters as focal points. A similar measurement between one of the first two transmitting stations and a third station will describe a second hyperbolic line that will intersect the first hyperbolic line at the navigating station This method allows the position, as shown in 1'ig:24. installation of a network of stations and the preparation
662
MARINE ENGINEERING
663
ELECTRIC PLANTS
of navigational charts with the hyperbolic calculations identified. The navigating station takes two or more measurements and projects its position from the appropriate intersections on the chart. Loran A was one of the first operational hyperbolic systems and operates in the 1.5 to 2 MHz radio frequency spectrum. The transmitting stations are operated in pairs with one the master station and the other the slave station. The master statioii transmits a pulse 6f R1' energy; the slave station receives this signal and after a programmed time delay transmits a pulse of Rl? energy. The navigating station measures the lapse of time between receiving tlle master station signal and the slave station signal; the navigating station then measures the lapse of time bettveen receiving signals from a different pair (master and slave) of Loran stations. Each measured lapse of time is used to determi~iethe navigating station's proper line of position (with respect to each pair of transmitting stations) on a Loran (hyperbolic) chart. Where the two measured lines of position cross establishes the navigating station position; see position "V" on Fig. 24. Each of Loran stations is usually located several hundred miles apart. One station is often made common to two pairs. As noted in l'ig. 24, M is a common or double-pulsed station (master); S1 and 52 are singlepulsed stations (slaves). A double-pulsed station is co~lsideredto be two separate stations at the same location, since this type of station sends out two entirely distinct sets of pulses; each set of pulses is paired with the pulse from adjacent stations. A typical shipboard Loran A installation includes an antenna arid a receiver-indicator; the receiver-indicator should be installed in the chart room, above the chart table, and should provide direct in-line time-delay readings. Loran A transmissions are effective at ranges of about 700 nautical miles in the daytime and up to 1400 miles at night with accuracies between f1 mile and f,5 miles depending upon location of the navigating station in the hyperbolic grid, time of year, and time of day. Loran C uses the same principles as Loran A except for the frequency of transmission (100 kHz), which affords better and more stable area coverage. Each master station has at least two and sometimes three associated slave stations; using these multiple slave stations provides for a more accurate identification of lineofpbsition hyperbolic grid lines. The use of the lower frequency, as compared with Loran A, allows for greater range on the reliable ground wave and separation of the master and slave stations by as much as 800 nautical miles. A typical Loran C shipboard installation will vary according to the requirements of the particular vessel. It may range from a suite of equipment similar to Loran A to elaborate computer-controlled devices with automatic updating and readout capabilities. The trend on merchant ships is to provide a combination receiverindicator that is suitable for use with both Loran A and C.
The Omega navigational system is a long-range llyperbolic navigation system that uses synchronized signals from two or more radio transmitters that may be spaced several thousand miles apart. Omega stations transmit on low radio frequencies from 10 to 14 kHz; use of these low frequencies results in transmission of signals that are very stable, accurate, and predictable. Precalculated line-of-position (hyperbolic) charts are used to plot the navigating station position, based on measuring the particular signals received from pairs of stations, somewhat similar to the Loran system, except that all stations are synchronized from a common time source and share transmitting time with the other stations. Each station transmits for a specific duration of time (approximately one second) and all stations transmit successively and always in the same order. Ultimately the Omega navigating system will consist of only eight stations strategically located so as to provide worldwide coverage; it is expected that the Omega system will result in accuracies approaching nautical mile in daytime and one mile at night. The current installation consists of four stations, located in Korway, Trinidad, Hawaii, and New York. A shipboard Omega installation would consist of a comparatively short ship antenna and a receiver. The Decca navigation system is a highly stable continuous-wave radio position-fixing hyperbolic-type system operating in the 70-130 kHz band. Each system or "chain" comprises a central master station and three slave stations disposed in a three-pointed star configuration about the master station. Slave stations are 70-120 miles distant from the master station. There are upward of 35 such chains throughout the world, most of which are in the eastern hemisphere. The performance range is up to 250-300 miles by day and positions correct to within &50 yd can be obtained at a range of 50 miles. Unlike Loran and Omega, which are operated by the U.S. Government, Decca is a privately owned system. The Decca receiver and indicators (Decometers) are generally installed in the chart room. Navigation charts for areas served by the Decca chains are readily obtainable. In general, receivers for the foregoing hyperbolic navigation systems are located in the vessel's chart room. 8.7 Echo Depth Sounding Sonar. An echo depth sounding system provides a means of measuring the sea depth beneath the vessel by computing the time interval required for sound waves to trave1,"at a known velocity, from the ship's bottom to a reflecting surface and return. The measured depth may be visually indicated and permanently recorded. This system usually consists of a transducer, an electronic control unit, an indicator, and a recording unit. The transducer is permanently mounted in and flush ,kith the ship's hull a t approximately one fourth the distance from the bow to the stern, and a h e a r the ship's centerline as is practicable; this transducer converts electrical oscillations, from the control unit, into sound energy during the transmit cycle and converts the echo
sound energy into electrical oscillations during the receive cycle. The control unit contains the circuitry that produces, times, and controls the various signals necessary for system operation; this unit is usuaIly located in the chartroom. The indicator visually displays the depth of the water, and is usually located in the wheelhouse. The recorder produces a permanent record of the sea depth on a calibrated chart and should be located in the chart room above the chart table. Some vessels are equipped with an alarm device, located in the chart room, that sounds \\-hen a p r e determined shallo\v depth occurs. Echo sounding systems are also used for measurement of ground track by using t~vo,four, or more directional sonar beams transmitting impulses at an angle. The doppler effect on each beam is measured, indicating a rate of change of slant range in that particular direction. The ground track navigation systems are used by large vessels navigating in shallo~v coastal waters, survey vessels, and ships required to maintain station over a given point. Ground tracli information is also required by ships using a doppler shift method of satellite navigation, such as the Transit System. A typical ground tracli installation could consist of a receiver/driver, four transducers mounted in the ship's bottom (which may be combined in arrays), a dead-reclioning tracer, and a digital readout device displaying forward, aft, and athwartships velocities. 8.8 Special Systems. Extremely accurate position information can be computed using navigational satellites in polar orbit. The first operational development was the U.S. Navy Transit system; military require ments dictated that this be a passive system. Satellites in polar orbit are monitored by ground tracking stations. Some of the tracking stations have the capability of updating the information stored in the satellite-borne computers. Every tw-o minutes a satellite transmits a message containing the time of transmission and orbital track information. The transmission occurs on two frequencies, in the UHF spectrum, separated by approximately 300 MHz, to compensate for distortions in propagation encountered in ionospheric and atmospheric penetration. The navigating station receives the transmissions with a receiver/computer complex that measures the doppler effect on the transmissions caused by the satellites movement with respect to the navigating
station. The resulting quantity, which is the rate of change of slant range, is used to compute position. By taking several observations from successive satellites, accuracies to a few hundred yards are possible. A shipboard installation for the Transit system consists of an omnidirectional dual-frequency antenna, a receiver/computer console, and a precise frequency standard. Concepts for navigational satellites suitable for commercia.1,. use are being developed: these are directed toward reducing the complexity of the shipboard installation by removing the requirement for passive navigational stations. Using interrogating transmissions from the navigating station and a shore-based computing station, several methods are being considered to establish position information. 8.9 Portable Systems. Dredges, sea drilling ships, cable laying ships, and similar special-purpose vessels require medium to short-range navigational systems of high accuracy but do not require permanent shore stations for reference. Several systems are available using portable shipboard and shore station units. Raydist is typical of this category. Raydist DR-S (dual range, single sideband) systems consist of small portable shore stations with an antenna installation selected by range requirements, and the shipboard terminal consisting of a CW transmitter, a navigator, and a strip chart recorder. The shore stations can be placed as required by range coverage because no tra~lsmissiorls occur between the shore stations. Two shore statioris-are required. The system measures the distance from the navigating station to each of the shore stations by phase measurement. The use of single-sideband transmission techniques reduces the radio frequency spectrum occupancy required by the system. The system requires active participation by the navigating station in that CW transmissions to the shore stations occur. Ho~vever,in the case of multiship operations, three or more navigating stations can receive position information by proper range organization. For operating areas where the use of land-based reference stations is not feasible, battery-operated reference stations encased in buoys are used. The Raydist DR-S can furnish position information up to 250 miles and accuracies of from 1 to 3 yd. It is widely used to record the speed of ships during standardization and maneuvering tests.
Sectio~~ 9 Wiring Application and Methods 9.1 General. I n general, shipboard electric cables are specified to be constructed and tested in accordance with the specific requirements of IEEE Standard KO.45. However, as an exception to this requirement, most owners permit the substitution, for the equivalent IEEE cable, of cable constructed in accordance with military
-
specificationsMIL-C-915,MIL-C-2194, or MII~-C-23206. The use of cable in accordance with military specifications is allo~vedprincipally because of its reduced material cost and ready availability. The only limitation usually placed on the use of Navy cable is that the maximum current for any conductor should not exceed the current-
664
MARINE ENGINEERING
carrying capacities specified in the publication "Cablc Comparison Guide, " NAVSHIPS 250-660-23. The regulatory body rules contain specific requirements regarding cable application and installation. These rules also contain tables of allo~vablecurrentcarrying capacities, dimensions, weights, and physical and electrical properties of mercbax~t marine cables; NAVSHIPS 250-660-23 contains this same type of information for cables for naval ships. 9.2 Light and Power Wiring. Except for flexible service applications, all light and power cables should have a basket-weave armor of steel, aluminum, or bronze; bronze armor is normally used for those cables installed in the weather, and in damp areas. Except for lighting and power branch circuits installed in dry spaces, such as crew and passenger areas, all light and power cables should have a moisture-resisting jacket under the armor; this jacket may be lead, thermoplastic, or thermosetting compound. Because of the excessive material and installation costs involved, lead jackets are seldom used. Rubber- or thermoplastic-insulated cable may be used in spaces where the ambient temperature does not exceed 50 C. Varnished-cloth, asbestos-varnished-cloth, or silicone-insulated cables should be used where the ambient temperature exceeds 50 C; additionally, these types of insulated cable may also be used for other applications. Where varnished cloth insulation is used, the wire size should not be smaller than No. 12 AWG; this type of insulated tape on wire of smaller size will not withstand bending. Asbestos-varnished-cloth insulated, jacketed, and armored cable is generally used throughout merchant ships for lighting and power circuits. There are, however, the following exceptions to this generalization:
1
I
,
I
I
SJ, SJO, or SJ'l' should bc used for portablc lights, tools, and appliances. 9.3
Interior Communication and Electronic Wiring.
The IEEE rulcs list the various types of merchant marine electric cable designed for use in interior communicatio~lsystcms, telephone systems, and electronics systems. However, as noted previously, the trend is to the use of Navy-type cable, in lieu of the merchant type, for most shipboard applications. The general practice followed in the selection of IC and electronic wiring is as f 0110\\-$: a. IC Systems. Navy-type SGA cable is used in the 4000 and 9000 circular mil sizes; for larger sizes IEEE merchant marine types, as used for power and lighting systems, are usually employed. Navy type MSCA is used for all multicorlductor applications with the exception of tempqrature-indicating systems, in which case Navy types PBJX and PBTX are used. b. Telephone Systems. Navy cable types SGA and hlSCA are used the same as for I C systems; Kavy type TTHlCWA is used for all twisted-pair applications. c. Electronic Systems. Navy cable types SGA and MSCA are used the same as for IC systems; Kavy types TTRSA and RG coaxial are used for shielded applications. As recommended by the equipment manufacturer, special cable and conductor types are used for electronic circuits and antenna connections. Basket-weave armor and moisture-resisting jackets are provided on the foregoing cable types, the same as described for the light and power wiring in the previous section. 9.4 Cable Installation. All cables should be continuous between terminations and, insofar as practicable, be routed to avoid areas where excessive heat, moisture, or oil may be encountered. Cables should not be run For local lighting branch circuits that are installed through oil tanks or pump rooms unless they are enin spaces having ambient temperatures not exceeding closed within watertight trunks; also, cables should not 50 C, rubber-insulated and armored cable is,used in dry be run behind, or embedded in, heat insulation. Where spaces; in damp spaces, rubber-insulated, jacketed, and it is necessary to pass cable through insulation, the cable should be run in a continuous pipe. armored cable is used. Cables routed through cargo spaces should be proIn lieu of the IEEE type of cable, Navy cable (type tected against mechanical damage incidental to the SGA in sizes 4000 and 9000 circular mils and type handling of cargo; advantage should be taken of the MSCA) is used. protection afforded by beams, girders, and stiffeners. Three-conductor cable should be used for all a-c Routing of cables within a radius of 15 ft from any 3-phase circuits to neutralize the inductive effect that magnetic compass should be limitzd to the necessary may cause heating of adjacent equipment or structure. binnacle and local lighting circuits. The largest Bconductor cable normally installed is Cables should be installed in compliance with the 400,000 circular mils. For circuits requiring greater United States Public Health Rules so as to avoid than 400,000 circular mils, two or more 3-conductor harborage for rats. cables of the same size are connected in parallel; one Tanker vessel cable runs between "islands" and to the conductor for each phase of the circuit should be con- forepeak are routed either: (a) on the underside of the tained in each paralleled cable. fore-and-aft walkway, and protected by steel channels All portable cords should have a minimum size of No. or enclosed in cable trunks or pipes; (b) attached to a 16 AWG, and should be in accordance with the specific steel plate supported from the walkway handrail requirements of the Underwriters Laboratories Standard stanchions; or (c) in a wireway-type structure on the for Flexible Cords and Fixture Wire. Types S, SO, or weather deck, protected by steel channels and cover, and ST should be used for hard service applications; types located well inboard to afford protection from seas.
ELECTRIC PLANTS
Cables should not be bent to a radius of less than six diameters to. avoid damage to the insulation. Cables should be grouped and routed in main wireways as much as economical distribution permits and should be supported as shown in Fig. 25(a), (b), or (c). Small groups of cables should be supported as shown in Fig. 5 , ( ) (i), or (1. Figure 25(j) shows typical methods of supporting a maximum of four small cables located behind joiner bulkheads and ceilings. Where cables are run on thin nonwatertight bulkheads, they may be secured by through screws and nuts as shown in Fig. 25(i). Single cables should be supported as shown in Fig. 25(g) and (h). Cables may be secured in wireways by fitted straps, such as shown in Fig. 25(a), or by banding straps, as shown in Fig. 25(e); these straps are usually 5.8 to % in. wide. Cable supports are usually spaced not more than 18 in. apart on vertical runs and 14 in. apart on horizontal runs. Cable straps should secure the cable without damage to the armor or insulation. Cables should be strapped in position a t every hanger in vertical runs, and at not less than every fourth hanger on horizontal runs; at turns of horizontal runs the cables should be strapped a t every hanger. Cable hanger material, not exposed to the weather, should be steel; galvanized material is normally used in refrigerated spaces and galleys. Exterior cable hanger material should be corrosion-resisting material or steel with a corrosion-resisting coating with CRES, brass, or bronze nuts, bolts, and washers. A variety of the alternative methods of passing cables through decks and bulkheads is illustrated by Fig. 26. Cable penetrations of nontight declcs and platforms are usually through clear openings of adequate size for the cables involved. Steel collars, extending approximately 3 in. above the deck, should be provided all around the edge of the opening. Where single cables pass through nontight bulkheads or beams, and the bearing surface for the cable is less than % in., a bushing should be provided as shown in Fig. 26(c). Where the bearing surface is % in. or greater, a clearance hole with rounded edges is acceptable. Figure 26(h) shows a method of passing a group of cables through beams with insulation. Cable penetrations of watertight and airtight bulkheads are through stuffing tubes, as shown typically by Fig. 26. Cable penetrations of watertight decks are through kickpipes with stuffing tubes, except that watertight riser boxes with stuffing tubes are sometimes used for groups of cables not in the weather; these are also shown typically by Fig. 26. The following subsection contains a description of stuffing tubes and their application. Insofar as practicable, cable runs in finished crew and passenger areas should be concealed. When paneling is fitted overhead or on bulkheads, through wiring is run. behind the paneling and secured to the steel deck or bulkhead. Where heat insulation is applied to the deck or bulkhead, the wiring is run between the insulation and
665
paneling, and supportcd as shown in Fig. %(a) nrld (b). I n some instances, when vertical cable runs arc not near a stcel bulkhead, a ladder type of frame, from dcclc to deck, is provided to support thc cablc, and is corlccalcd by paneling. Switches, call buttons, and rcccptacles locatcd in paneled spaces are usually recessed into hollow H-section joints. All cables to these'outlets should be run overhead and then down in the H-section to their termination; each cable should be secured at the top of the H-section and a t the outjet. Wiring to light switches a t entrance doors is run in the hollow door frames on which the switches are mounted. Where concealment of cable runs in finished areas is not practicable, special troughs or channels are sometimes used to conceal individual runs of cables to surfacemounted outlets; otherwise, exposed wiring should be installed to be as inconspicuous as possible. 9.5 Stufflng Tubes. Stuffing tubes are either of the terminal type or bulkhead type as illustrated by Fig. 27. Terminal-type tubes are used to provide a watertight cable entrance into electrical equipment enclosures; they are made of either brass or nylon. I n general, terminal tubes consist of a body, gland rings, packing, and a gland nut or cap. A watertight penetration is obtained by tightening the gland nut or cap to squeeze the packing against the cable. For enclosures Y I G in. thick or thicker, tapped holes are provided to accept the tube male pipe threaded end. For enclosures less than a/,, in. thick, clearance holes*are provided; in this type of installation, brass tube' bodies are brazed to the enclosure, whereas nylon tube bodies are fastened and made tight by a locknut and O-ring assembly. Bulkhead type tubes are used to provide watertight and airtight cable penetrations through bulkheads and decks. Bulkhead tubes are usually steel, and are similar to brass terminal tubes, except that the tube end 'opposite the gland nut has standard IPS female threads to accept threaded pipes. For passing cables through bulkheads, a pipe nipple of adequate length is threaded into the tube body; the assembly is jnserted into a clearance hole in the bulkhead, with the tube body butting the bulkhead; the tube body is then welded to the bulkhea'd. For passing cables through decks, kickpipes are welded into the deck with a bulkhead-type tube threaded,on the upper end. The height of the kickpipe assembly should be a t least 9 in. to the top of the tube, with the pipe extended below the deck, or sheathing, as applicable, leaving approximately % in. of the pipe exposed. An additional method &-passing cables through tight bulkheads is by means of a "multicable transit," shown by Fig. 26(i). This method utilizes a special assembly, welded in a clearance hole in the bulkhead, through which several cables may be pulled. After the cables are pulled, they are positioned and maintained in place using a system of split insert blocks; the complete cable grouping is made tight by applying compression through special bolting and plates. Cable transits may be used
,
ELECTRIC PLANTS
MARINE ENGINEERING
* ELDED STUD
INSULATION
,-CONDUIT
COUPLING
PHENOLIC COUPLING WALL THICKNESS 3/8: OUTSIDE THREADED
STEEL CABLE BAND
( 0 ) SINGLE ROW OF CABLE SUPPORTED FROM INSULATED OR NON- INSULATED DECK OR BULKHEAD
(c)
WITHOUT INSULATION ON ONE SIDE
(a) FOR REFRIGERATED SPACE BULKHEAD
STEEL DK OR 81
FOR NON-TIGHT BHDS
K,,,,,
I
GAP FOR RAT HARBORAGE PROTECTION WHERE--J WIREWAY IS NOT BEHIND RAT PROOF SHEATHING IS PROVIDED BYSTEEL SADDLE BANDEDTOHANGER.
THREADEDWASHER
(el MULTIPLE ROWS USING BANDING METHOD
STEELHANGER
OR BHD E
METAL SHEATHING
STEEL STRAP AT EVERY 4 t h HANGER AND WHERE CABLE CHANGE DIREC
--
.I " .MAY ..-.. n l.A CABLE
kI-
SHEETMETAL
I-3"
L---KICKPIPE. I
( t ) SINGLE ROW ON INSULATION SHEATHING +--9"MAX.--+
CABLES MAY BE SUPPORTED ON EITHER TOP Ok BOTTOM OF HANGER BUTNOT BOTH
STEELDKORBHD
STEE CABLE Dl
STEEL CABLE STRAP ( g ) SINGLE CABLE
(b) SINGLE ROW OF CABLE SUPPORTED FROM INSULATED OR NON-INSULATED DECK OR BULKHEADS TO CLEAR OBSTRUCTIONS
LEETS- -
DK OR BHD WELDED STUD WlTH COLLAR
STEEL STRAP rSTRAP
I
7
MAX--+
EVERY 4 t h HANGER 8 WHERE WIREWAYS CHANGE DIRECTION DEC I(
(h) SINGLE CABLE WHERE EXCESSIVE MOISTURE IS PRESENT NON - WATER TIGHT STEEL BULKHEAD
STEEL FERRU
ONLY WHERE BEHIND RAT PROOF
i5" MAX. ALLOWABLESPAN FOR CABLE
ROUND INSIDE EDGES
-
SECTION A-A
( i ) SINGLE ROW MOUNTED ON BULKHEAD TOO THIN TO WELD.
SULATWN
WHERE CABLES ARE FEW
STEELSTRAP
-- -
ROUIYD EDGES
L~~~ PARTIAL LINER
A
J
(h) FOR NON-TIGHT BEAMS WITH INSULATION
\
(g) FOR BEAMS WlTH LIGHTENING HOLES
CHANNEL WELDED COLLAR THE SUPPORT OF MULTIPLE ROWS OF CABLE ON STEEL BULKHEADSISACCOMPLISHEDTHE SAME EXCEPTTHAT EACH TIERSHOULDBESTRAPPEDANDNOSPANOF CONTIGUOUS CABLES SHOULD EXCEED 9': WHERE SPANS UP TO 15" MAX. ARE REQUIRED, A HOLD-DOWN BOLT SHOULD SECURE THE RETAINING STRAP AT ITS CENTER. ASSEMBLIES SHOULD NOT BE SPACED MORE THAN 14" APART. CENTER TO CENTER.
RE FILLER BLOCKS EPROOF ELASTOME FOR FUTURE CABLE
( c ) MULTIPLE ROWS OF CABLE SUPPORTED FROM INSULATED OR NON- INSULATED DECK.
TRANSIT FRAME
-CABLE
JOINER BHD OR CEILING-
( d l TWO SMALL CABLES 3/4"DlA. MAX (LOCATED BEHIND JOINER BULKHEAD OR CEILING)
Fig. 25
STRAP-
CABLE STRAP JOINER BHD OR CEILING
7
(i) CABLE RUN OF FOUR SMALL CABLES ON STEEL DECK OR BULKHEAD WlTH AND WITHOUT INSULATION (BEHIND JOINER BULKHEAD OR CEILING)
Methods of supporting cable
INSERT BLOCK FIREPROOF ELASTOMER
THRU DECK APPLICATION
(i) MULTI-CABLE TRANSIT TYPE BULKHEAD PENETRATION ASSY (WATERTIGHT BHD) Fig. 26
Methods of pauing cables through deck and bulkheads
668
MARINE ENGINEERING
ELECTRIC PLANTS
EXPLOSION PROOF '
RUBBERSEAL (b) SECTION THRU MALE TERMINAL STUFFING TUBE SHEET META
(c) SECTION THRU BRAZED TERMINAL STUFFING TUBE UP TO 3/16
(a) SECTION THRU CABLE SEAL FITTING FOR EXPLOSION PROOF EQUIPMENT
BUSHING
WAS
ASHERS
WASHERS
NlPPL
-W
ENCLOSURE WALL
(dl
SECTION THRU NYLON TERMINAL STUFFINGTUBEFOR ENCLOSURES OVER 3/16
II (el
Fig. 27
SECTION THRU NYLON TERMINAL STUFFINGTUBE FORENCLOSURES UP TO 3/16
Cable rtufFlng rubes
for passing groups of cables through decks, provided a riser box is utilized; in this application, the cable transit would be installed in the top of the riser box, Fig. 26(j'). Cable entrance into explosion-proof equipment located in explosion-hazardous areas must be accomplished through an approved seal fitting, similar to that shown in Fig. 27(a). Proper installation would be as follows: nipple the seal fitting to the explosion-proof enclosure by a pipe not more than 18 in. long; remove all of the cable covering, except the individual conductor insula-
LBUSHING I L O C K NUT
-
tion, f ~ o mthe point of entering the seal fitting to its end; fill the seal fitting with a sealing compound through the filling plug hole. When the compound hardens and the filling plug is in place, the cable entrance is explosionproof. 9.6 Determination of Cable Lengths. The length of each run of cable may be determined from wiring deck plans, or isometrics, using a standard map measure. I n addition to the lengths measured from plans, the following allowances should be made:
10 ft for each vertical run between decks. 5 f t for each vertical run from overhead to bulkheadmounted equipment. 10 f t for each connection to a switchboard. I n addition to the foregoing allowances, 10 percent' should be added to feeders and branches to cover waste, losses incidental to making up cable ends, and for unforeseen bends. 9.7 Connections and Terminals. Connection of wires to terminals should ensure a good electrical contact without damaging the conductor. Since no splicing of wires is permissible, all terminations should be made by screw connections or approved cable connectors located within equipment enclosures or wiring appliances. I n general, all ship cable connections should be made by solderless lugs of an approved clamp or compression type; exceptions include twist-on pressuretype connectors with a securing device which are sometimes used for making connections in lighting outlets and connection boxes. Also, interior communication and electronic system wiring connectors may be of the soldered type or may have the wire end formed into an eye and be solder-dipped. Cable ends of vital circuits should be sealed against the admission of moisture. Usual practice is to select connectors of the proper size, but in special cases, where the wire size has been increased to reduce the voltage drop, strands may be removed to permit the use of smaller connectors. I n no case should the remaining cross-sectional area be less than that required to carry the maximum current.
9.8 Grounding. Except for final branches which may be grounded ut the supply end only, the metallic sheath on all armored cable should be groutded a t each cnd. Where armT%TZg;tile eiltGrs n box or wiring device, the armor should bc coritinucd into the box and secured by a clamp or corinector to assure good contact with the box. If this is impracticable, thc armor should b ~ ~ o u n d to ed the steel structure a t the ngrest cable strap. For each ground the armor and strap shouldJe we11 cleaged,at the point of contaot. Radio, radar, and miscellaneous communication equipmenta have various requirements regarding grounding which are specified by the equipment vendor. To ensure proper operation, the shipboard installation should comply - - with the vendor's recommendations. Stays, shrouds, and other standing rigging should also be permanently grounded to prevent the accumulation of static electricity. Rigging and railing in way of the radio direction finder loop and sense antenna need not be grounded but should have insulators inserted every few feet to prevent a shielding effect caused by nearby induced radio frequency currents. Enclosures and frames of electrical equipment generally have inherent grounding by support from the ship's structure. When they are not attached directly to steelwork, special grounding should be provided. All portable equipment should 'be grounded by a separate conductor in the flexible supply cable and a grounding device in the receptacle and plug.
*
PIPING SYSTEMS
e
CHAPTER XVlll
E. E. stephenson
I
ERmGE DECK
Piping Systems
Sectio~~ 1 Machinery Space Arrangement 1.1 Introduction. The development of a machinery space arrangement and the development of piping systems within a machinery space are closely related processes. I n fact, they are inseparable; the location of the majority of machinery components is predicated on the basis of piping system considerations. The considerations which must be entertained when developing a machinery space arrangement were outlined in Chapter 1. As is the case with the development of designs for many types of complex systems, the development of a machinery space arrangement is an iterative process. Such an iterative process is illustrated by the preliminary design spiral shown by Fig. 3 of Chapter 1. For most types of ships, such as passenger liners, oil tankers, and containerships, there is no uniquely optimum machinery space arrangement that is clearly discernible. In the usual case, a number of arrangements would be considered satisfactory; of course, some would be preferable to others. It is the responsibility of the marine engineer to assess the various alternative courses of action and select the one which affords the best compromises. In order to arrive a t a reasonably optimum machinery space arrangement, the marine engineer must possess a broad base of knowledge covering the operation and maintenance of propulsion plants, piping systems, ventilation systems, and electrical systems. Concurrent with the preparation of the specifications for a ship, sketches of the machinery space must be developed. One of the foremost considerations when developing a preliminary machinery space arrangement is to recognize that the arrangement is based on many components which are onlyftentative selections and that their dimensions are only approximate. Insofar as practicable, a reasonable allowance should be made for variations in dimensions and unanticipated developments. Since payload is not carried in the machinery space, the marine engineer is under constant pressure to decrease the size (particularly the length) of the machinery space. Sufficient space must, however, be provided for proper operation and maintenance of the machinery and supporting equipment. 1.2 Development of Machinery Space. One of the first decisions to be made regarding the design of a machinery space is its location. Although many varia-
67 1
tions are possible, particularly for ships with special applications, there are usually two choices with the more commonplace merchant ships; namely, the machinery may be located either in the aftermost region of the ship or in the region of the ship somewhat aft of amidship. The machinery spaces of oil tankers, as an example, are without exception confined to the stern of the ship; this arrangement generally simplifies the design of the ship, especially in that explosion hazards are minimized. Two factors permit such an arrangement: one is that the hull lines in way of tanker sterns are full so that there is sufficient hull width to accommodate the machinery. Another is that the ability of tankers to take on ballast enables the ship operator to adjust the operating drafts within rather broad limits; therefore, the heavy weight inherent with the main machinery presents no trim problems even when confined to the stern of the ship because seawater ballast can be taken on in the forward region of the ship so as to maintain a reasonably even draft. Many types of dry cargo ships, on the other hand, do not have the ability to arbitrarily adjust their operating drafts; consequently, the weight inherent with the main propulsion machinery would present a severe trim problem when the ship is lightly loaded. Furthermore, the hull lines aft on most dry cargo ships are so fine that the propulsion machinery will not fit within the confines of the hull in the aftermost region of the ship. For these reasons, the machinery space for fine dry cargo ships is located considerably further forward. When developing a machinery space arrangement, the designer must visualize the piping, valves, and other equipment associated with each component. Also, the functional requirements and interrelationships of the various components must be borne in mind both from a mechanical and the operator's standpoint. Maintenance i and overhaul play important roles in the arrangement of many components, particularly those containing tubes which occasionally require replacement. There is often a range of flexibility in the arrangement of a machinery space. Assuming for discussion purposes that overall considerations of the type outlined in Chapter 1 have led to the selection of a steam turbine type of propulsion plant, the boilers may be positioned forward of the main engines or they may be located aft and above the main engines as illustrated by Figs. 1-
I
MAIN D E C I
1
'
Fig. 1
Machiney arrangement-ale~ationat c looking to p a t
through 4 (which are a typical dry cargo ship arrangement). The latter arrangement permits a shorter machinery space, but the raised center of gravity of the relatively heavy boilers is disadvantageous as compared with the former arrangement, which permits the boilers to be located low in the ship. In selecting the boiler location, the space between, forward, and aft of the boiler must be studied from the viewpoint of operation as well as such maintenance
considerations as boiler tube renewal, soot blower tube removal, economizer tube removal, burner removal, and desuperheater removal (if appropriate to the boiler selected). The uptake location between the boiler and boiler hatch, as well as the forced-draft blower duct work, are investigated and normally shown on preliminary design sketches. Forced-draft blowers are usually located advantageously to draw the hottest air either from air casing surrounding the uptakes or from the top
PIPING SYSTEMS MARINE ENGINEERING
.
MARINE ENGINEERING
PIPING SYSTEMS
of the machinery space casing if uptake casings are not required. Machinery space casings are hecked to determine that their size is adequate to ccommodate the boiler uptalres and the ventilation air intakes with access around the uptalres. Trade-offs between the components selected and the space available are sometimes required as the equipment tentatively selected may not lend itself to an acceptable arrangement. For example, in the choice between horizontal and vertical pumps, horizontal pumps require more space but are easier to support Bnd are more readily overhauled. I n the case of heat-transfer equipment, the choice between vertical and horizontal types can also be made to favor the space available. Once the general location of the main engines has been tentatively established, the vertical position of the main engines is investigated. As discussed in u a p t e r 11, the location of the propeller is fixed by hydrodynamic considerations, but, by raking the shaft, a degree of freedom in locating the main engines is provided. As can be seen from Fig. 1, in the usual case there is scarcely adequate space beneath the main reduction gear for the lube oil sump, and the main condenser is so lo\\- that it is difficult to provide sufficient submergence for the main condensate pump without recessing the innerbottom. As a result, the main shafting is invariably raked upward going forward so as to alleviate these problem areas. Moderate amounts of shaft rake are not objectionable (see Chapter 11). If the main eneines are located in the aft end of the machinery sgace, a check must be made to ensure that there is adequate space around the reduction gears for passage and that sufficient access to the gear and pinion bearings is provided. Foundation girders for the reduction gears and the main thrust bearing are sketched in to confirm that sufficient rigidity can be provided. At this time the large machinery components should be studied to ascertain that they can be either unshipped through the hatch or removed through a hole cut in the hull. Frequently, unshipping considerations necessitate adjustments in the location of the machinery components. Several aspects of the main condensate system require special study during the early design stages. The main condensate pumps must be able to handle water under vacuum a t its saturation temperature. Flashing will occur a t the pump impeller entrance unless adequate head between the minimum operating water level of the condenser hot well and the pump suction is available. A short and direct lead slopingc-from the hot well to the pump is essential. Locating the pumps near the condenser fore-and-aft centerline will reduce erratic suction conditions in the event that the vessel rolls or assumes a ~ermanentlist. The main condenser seawater piping, due to its size, should be shown on the preliminary plans and analyzed with.the condenser overhaul space. The main circulating water pumps should be arranged to provide short sea suctions via high and low suction sea chests. The pump discharges to the main condenser and then overboard as
2
c2
675
shown bv kcin. 2. The seawater coolant for the main lubricatiig oz coolers is usually supplied by the main circulating pump. The lubricating oil pumps take suction from the oil sump under the main reduction gear and should be located to provide the shortest suction piping practical. Where a gravity system is employed, the lubricating oil pump discharges through a'strainer and cooler to a gravity tank that is located well up in the machinery space casing. The gravity tank must be of sufficient height to supply 'oil under the flow of gravity a t a pressure of about 10 psi a t the turbine and gear bearings. The oil completes the cycle by collecting in the oil sump beneath the slow-speed gear. Due to damage control considerations, naval ships are not provided with a gravity lubricating oil system; instead, the pressure in naval lubricating oil systems is maintained by pumps. Due to arrangement considerations, some merchant ships also employ a pumppressurized lubricating oil system. Other systems such as the auxiliary condensate system and feed system are treated similarly. As illustrated by Pig. 2, components such as fire pumps, bilge and ballast pumps, fuel oil transfer pumps, sanitary pumps, freshwater pumps, general service pumps, and air compressors, which are not directly related to the propulsion system, are studied and located to advantage. At this point, with a satisfactory arrangement tentatively established on the lower level, the main operating level can be established. When fixing the operating level height, due consideration-must be given to the space required beneath the level for piping, wireways, ventilation ducts, lighting, and the necessary headroom on the lower walking level. The operating level located outboard on each side of the main propulsion unit serves as a foundation for the turbogenerators, switchboards, distilling plants, contaminated evaporators, and other equipment as shown in Fig. 3. It is desirable from an operational standpoint for the operating level to be a t the same elevation a t all points; however, a lower level may be required between the turbines and over the main reduction gear for access to bearings and to permit observation of lubricating oil thermometers and sight flow indicators. The size and location of structure, including pillars, within the .machinery space may be determined to suit estimated component weights. Smaller equipment such as air ejectors, pump indicator panel, telephone booth, and log desk are located on the basis of o ~ t i m u mo~erationalconsiderations. The space provided for.#perating -areas and access should be confirmed as adequate. I n determining access and operating areas, the following should be considered minimum. The headroom in all working areas should be 6 ft-3 in. The clear operating space in front of the main throttle and control stations and the width of the boiler firing aisle should be 5 ft. The width of main access passages should be 24 in. ; however, the width of secondary or infrequently used passages q a y be 18 in.
676
MARINE ENGINEERING
The width of main access ladders should be 24 in. and the angle of sloping main access ladders should be 60 deg from the horizontal. The slope of infrequently used ladders may, however, be greater; the width of vertical ladders to infrequently used intermediate levels may be 18 in. with 12 in. as a minimum. At this point an overall evaluation must be made. I s there adequate space for access, operation, and maintenance? If not, there may be a necessity to increase the length of the machinery space. It is most unlikely to find that the space available is in excess of that required; but in the event of such fortunate circumstances, the excessive space should be assigned to the cargo.
PIPING SYSTEMS
1.3 Summary. Despite the amount of work that has gone into the design of the machinery space thus far, the arrangement developed is only tentative. Much of the arrangement is based on approximate data; furthermore, sketches of several alternate arrangements may have been developed for some areas. There now remains the task of evaluating the various alternatives and selecting those offering the best compromise from the standpoints of cost, performance, operation, accessibility and overhaul. If all of the foregoing is carefully executed, very little rearranging will be required during the development of the final working drdwings which are prepared for ship construction.
Section 2 P i p i ~ ~Desig~~ g Details 2.1 Piping Arrangements. Piping should be arranged in a neat, orderly manner and should not obstruct or interfere with operation of doors, hatches, or scuttles. Piping should permit free passage in walking areas and performance of work in designated working areas. The operation and control of machinery should not be impeded, and the interference of piping with the maintenance of equipment and ship structure should be minimized. Wherever practicable, piping should be kept clear of removable plates provided in the ship structure for shipping and unshipping machinery or equipment. The access to a compartment, boiler drums, or other components should not be limited by the piping. Piping should be portable in way of machinery and equipment that require dismantling for periodic overhaul, and wherever necessary for access to other piping systems or electrical systems. Cutout valves should be judiciohsly located to isolate sections of piping intended for removal during maintenance and overhaul of machinery and equipment, so that interference with continued operation of the remainder of the system is minimized. Piping should be located where it ~vouldnot normally be subjected to mechanical damage. When situations malie this impracticable, a means to protect the piping should be provided. Ample provisions should be made to accommodate the thermal expansion of pipes and movemerits due to hull deflections. This is often accomplished by means o f . exparisiori bends or offsetsin thc pipirig (see Scctiori 2.!5), where such can bc provided without causing uridcsirable poclccts; alternatively, in some cascs cxpansior~joints may bc used. l'ipirig arrangemerits which causc exccssivc turbulence det/rimental to tlic system or coritairi unlicccssary high or lo~vpoints should bc avoided. Wlicrc high and low points in piping or equipmerit are u1invoidable, vents, drains, or othcr mcans diould bc iristallcd to erisurc
proper operation of the system and its connected equipment. The amount of piping led through messing and livingt spaces should be minimized. Piping in such spaces should be symmetrically and neatly arranged and installed in the most inconspicuous location practicable, consistent with efficient operation and maintenance of the piping system. Except as necessary to serve the space, insofar as practicable, piping should not be run through medical and dental spaces, chain lockers, freshwater tanks, gasoline tanks and service areas and voids surrounding them, refrigerated spaces, and wiring trunks and enclosures. Steam and liquid piping should not be located where drips or sprays from leaks, condensation, or splashing from funnels could damage electric equipment. Where this is not practicable, adequate shielding of the equip ment should be provided. I n addition, flange or union joints should not be installed in the vicinity of such equipment. Where bilge or ballast piping is led through deep tanks, means should be provided to prevent the flooding of the compartmerits being served in the event of pipe leakage within the tanks. Such means may consist of an oiltight or watertight pipe tunnel, xJr the lines may be of schedule-SO pipe thickness, fitted with expansion bends, and all joints within the tanlcs welded. Where a pipe tunnel is installed, the watertight integrity of the bullheads should be maintained, and if the tunnel is riot of sufficient size to afford easy access, valves or fittings should riot be located within it. Bilge lines led through deep tanks without a pipe tunnel should be fitted with norireturn valves a t the bilge suctions. Where pipes are carried through watertight or oiltight bulklieads, declcs, or tank tops, the watertight iritcgrity of thc structure should be maintained.
I
1
I
i
Lead or other heat-sensitive materials should riot be used in piping systems which perletrate watertight subdivision bulkheads, where deterioration of such systems in the event of fire would impair the ~vatertiglit integrity of the bulkheads. Illsofar as practicable, piping with operating temperatures exceeding 125 1' arid vacuum piping should not be located in the bilge area. Pressure',gages, thermometers, level gages, and other instruments should be located so tliat they are visible from equipment operating areas. Every effort should be made to locate valve hand\\-heels such that they can be conveniently operated. Where this is impracticable, remote operatirig gear should be provided for convenient operation, or the valves should have attached gears or extension shafts for this purpose. Valves in horizontal piping below eye level should be arranged with their stems pointing above the horizontal wherever practicable. Globe and angle cutout valves may be arranged \\-it11 the pressure either above or belon- the disk, whichever is more advantageous for operation, protection, and repair of machinery and equipment served by the system. Examples of typical valves which should have the pressure under the disk are boiler stop valves, root valves, throttle valves, sea valves, and tank valves subject to tank pressure. Where a valve is subject to vacuum in the closed position, the vacuum should also be under the disk. Manifolds may be used wherever globe or angle valves are satisfactory, provided their use results in a simplified piping arrangement or affords an advantageous operational feature. Insofar as practicable, manifolds should be installed with their body just above the floor plates. Inlet piping to safety and relief valves should be short and direct. Where relief is to the atmosphere, open ends of discharge pip$g should be such as not to damage machinery or equipment, or endanger personnel. Relief of flammable liquids should be to a suitable lowerpressure part of the system. Relief of toxic or explosive gases from containers or systems shbuld be to the atmosphere above the weather deck. Air escapes from flammable fluid tanks and atmospheric reliefs from toxic and ,inert gas systems should not terminate where their discharge can be picked up by ventilation or forced-draft air intakes, or where their discharge would otherwise damage machinery or equipment or endanger personnel. Tailpipes should terminate a t the lowest point in tanks and should be installed in such a manner as to preclude their coming into hard contact with the bo4tom of the tank should the bottom of the tank deflect upward. The end of each suction tailpipe should be enlarged to provide an area not less than 1.5 times the inside area of the tailpipe. The height above the bottom should be one half the tailpipe diameter and the tailpipe location with respect to adjacent plating or other components should provide a free suction area around the open end periphery of not less than 1.5 times the inaide area of the tailpipe.
677
All of the foregoing requiremcrits arc desirable without exception; lio~vcvcr,it is riot uriusual that some must be forcgo~icand compromises must bc made a t the expelise of the less importarit arcas. Thc importaricc of a thorough working knowledge of the applicable rcgulatioris of tlic U.S. Coast Guard [I],' l'ublic Hcalth Scrvicc [2], and classificatiori societies (e.g., referericc [3]) cannot tfc ovcrempliasized. Definitive requiremcrits regarding marly aspects of pipirig systcms are cyitained in thesc regulations. It is important that these requirements be reflected in a piping system design during the prelimiliary stages as conformarice to the requirements a t a.later point call be costly and necessitate undesired compromises. 2.2 Materials. In order for a material to perform characteristics:in service, it must possess the follo\ving satisfactorily Suficielit strength or load-carrying ability a t the operating temperature to resist the imposed stresses. Retention of suitable ductility arid impact properties a t all operating temperatures. Satisfactorv corrosion and erosion resistance to the medium which it contacts. Resistance to galling with mating materials in moving contact. Many materials may be suitable for the same service. Factors which could influence the decision to select a particular material include the service, availability, cost, castability, forgeability, weldability, and lead time required in production. Metals cannot be classed simply as castable, forgeable, or weldable as such properties are a matter of degree. What might be considered as a satisfactory application for one category of service could be unsatisfactory in another application. Past successful practice plays an important role in the selection of materials. Pressure and tem~eraturelimits for the more common piping materials are giver1 in Table 1. Limits for recommended working stresses for piping materials a t various operating temperatures are given in reference [4]. A discussion of such metallurgical subjects as the affinity of materials to gall and galvanic corrosion of materials in seawater is covered in Chapter 22 and will not be repeated here. Regarding galvanic corrosion, however, the following factors may be considered when designing 9 system to minimize its effects. Reduce potential differences between metals by selecting materials close together in the galvanic series or metal combinations where one member polarizes easily. Avoid bimetallic couples where possible, by insulation or by proper choice of materials. Where bimetallic couples cannot be avoided, keep cathodic (noble) metal area small in relation to the anodic metal. Important members such as fasteners, -
Numbers in brackets designate References at end of chapter.
.
MARINE ENGINEERING Table 1
Material Pressure and Temperature Limits DESIGNLIMIT
MATERIAL Steel Pi e 8 e a m k carbon Seamless alloy
Electric-resistance welded Butt welded Wrought Iron Pipe Lap welded Butt welded Plates Carbon steel Iron Castings Malleable iron
ASTM Smc.
h y r e , Temp., deg F
A53-A, B ... A106-A, B, C ... A335-P1 ... A335P2 ... A335-Pll ... P12, P21, P22 A312TP321. ~ ~ i 4 7. .
.
875
1200
A53-A, B A135-A, B A63
roughness. Empirical values, such as those of Moody [5], are commonly used for the friction factor. Most practical piping systems contain valves and fittings whose resistance to flow is significantly greater than that corresponding to their actual length. Their resistance contribution is usually assessed by assigrliilg them a nondimensional equivalent length, this being the length of straight pipe (measured in diameters) of the same nominal size required for a n equal pressure drop. For piping runs containing n valves and fittings each having an equivalent length (L/d)i, the term L/d in equation (4) 11-ould be replaced by
where
A72 A72
p = fluid pressure, psf r = fluid density, lb/cu ft V = fluid velocity, fps g = acceleration of gravity, ft/sec2 z = elevation of fluid, ft T = fluid frictional shear stress, psf R = mean hydraulic radius of flow channel, ft L = pipe length, ft
A201-A, B
A47-32510, 35018 A197-(Cupola) Cast iron A126A, B, C A278 Copper and Copper Alloy Pi e and Tubing Eeamless copper B42 B75 -. BSS Copper alloy B43 (Red brass) B111 (Adm. Metal) B l l l (Al-brass) B l l l iAl-bronze) B l l l ( ~ e brass) d B l l l 70-30 Cu-Ni) . B l l l [8&20 Cu-Ni B l l l (90-10 Cu-Nil Brazed copper ... Co per Alloy Castings ironze B61 (Steam bronze) B62 (Ounce metal) B14.3 (Tin bronze) ~ 1 7 8 (PVC) 5 . Plastic Piping
difference between specified points along a pipe in terms of flow rate, pipe dimensions and roughness, and properties of the fluid. The tools of analysis are derived from the momentum relation (an expression of Newton's second law), continuity (an expression of conservation of mass), a friction factor, the general energy equation, and the equation of state. The f i s t three of these may be combined to form the following differential. equation which describes the condition of fluid in motion:
PIPING SYSTEMS
If the fluid flow is nearly horizontal or is a ga& the third term can be dropped. Additionally, if the fluid can be considered as incompressible and there are no significant changes in flow area, the second term can be dropped and the remaining terms can be integrated to obtain the following expression for the pressure loss over a length of pipe.
- ---
75
320
300 150 300 150
450 366 450 140
valve seats, and critical components should be the cathode in a bimetallic system. Paint or coat large cathodic areas. Remove mill scale from steel surfaces. When using stainless steel, avoid crevices and regions which will contain standing water. Do not rivet stainless steel or use lap joints for underwater service. Waterproof all faying surfaces involving stainless steel with other bimetallic connections. Structures under cathodic protection ahould be electrically bonded together to provide low-resistance connections. Provide proper drainage to prevent accumulation of standing water. Choose appropriate metals for specific environments to prevent stress-corrosion cracking. 2.3 Determination of Pipe Sizes. A fluid flowing through a pipe must be supplied with a pressure gradient to overcome the frictioilal resistance of the pipe walls. Tllc typical engineering problem is to find the pressure
The fluid frictional shear stress is defined as
where f is a dimensionless friction factor. R is defined as one fourth the pipe diameter, dl for a circular pipe; therefore, the head loss, hL, over a length of pipe car1 be determined from equation (2) as
The use of equation (4) is not confined to incompressible fluids (i.e., liquids) ; it can also be used in connection with compressible fluids (i.e., gases) when the pressure difference along the pipe is so small that the fluid density is nearly constant between the points of interest. I n the design of main steam piping, for example, equation (4) can bc used because there is a negligible change in fluid density between tlic supcrheatcr outlet arid the turbine throttle. On thc other hand, boiler escape piping offers cquation (4) is inacan example of a system for wl~iicl~ curate because of thc largc prcssurc drop, and hcncc steam dcrisity change, in thc piping. Thc friction factor is, in gc~lcral,a function of I
Values of equivalent lengths for common valves and fittings can be found in handbooks such as reference [6]. Piping is generally sized on the basis of pressure drop and flow requirements except where other criteria or operating conditions specified for a particular system impose additional considerations. where a main serves two or more units of machinery or equipment through parallel piping circuits, the design and arrangement of the piping should ensure adequate flow to each during all operating conditions. Where necessary and practicable, throttle valves, orifices, or similar devices may be used to regulate flo~v. However, since these components tend to create turbulent flow, they should be installed only where similar results cannot be obtained by a more optimum piping system design. Piping design and component selection should be such that it is not necessary to install restrictions in the flow to a major component to assure adequate flow to an auxiliary component. Table 2 contains a tabulation of suggested fluid velocities which may be used as a guide in selecting pipe sizes. The upper velocity limits listed f r ~Table 2 should not be exceeded even for the larger sizes because of erosion, water hammer, or the possibility of noise. The velocities suggested in Table 2 should be disregarded if the corresponding pipe sizes are so small as to incur head losses exceeding the pump suction or discharge head ratings. Particular care must be taken to analyze the h-- -~ n-. d loss on the suction side of pumps. The suction piping should be checked on the basis of design details for each specific case. For oil systems the suction losses due to high viscosity plus static lift can easily exceed the rated suction lift of the pump and should be analyzed i n detail. 2.4 Pipe Wall Thickness. Definitive requirements which control the minimum permissible thickness of pipe walls are promulgated by the U.S. Coast Guard [I] and the classification societies (e.g., reference [3]). Pipe wall thickness requirements are continually updated and revised; therefore, details of the current requirements are not repeated here. 2.5 Flexibility. Flexibility is a major consideration in the design of high-temperature piping systems. Hightemperature piping systems must be designed to have
679 Table 2
Fluid Design Velocities
SEUVICE Condensate primp suction Condensate pump discharge Condensate drains Hobwater suctions Hobwater discharge Feedwater suction Feedwater discharge Cold freshwater suction Cold freshwater diskharge Lube-oil service $ump suction Lnbe-oil discharge Fnel-oil service suction Fuel-oil service discharge Fuel-oil transfer suction Fuel-oil transfer discharge Diesel-oil suction Diesel-oil discharge Hydraulic-oil suction Hydraulic-oil discharge Seawater suctions Seawater discharge Steam, high pressure Steam exhaust, 215 psig Steam exhaust, high vacuum a
b
/
FLUIDVBT~OCITY, fps Nominal. Limit dd 3 3& 8 0.3Cd 1 & 3 8 3& 1.3& 4 4dd 10 3 4 15 5dd 20 dd 4 2& 6
&
1.5&
da
2& 2dd 5dd 1.5& 8& 343 5dd 50& 75& 75&
4 6 6 15 7 12 8 20 156 15b 200 250 330
d is the pipe internal diaketer in inches. 9 fps for galvanized steel pipe.
sufficient flexibility to prevent thermal expansion or contraction from causing excessive stresses in the piping material, excessive bending moments a t the joints, or excessive forces or moments a t points of connection to equipment and a t anchors and guide points. Piping flexibility calculations should be made when the system design temperature exceeds SO0 F and reasonable doubt exists as to adequate flexibility of a system. Reasonable doubt of flexibility for a two-anchor system of ferrous piping of uniform size may be considered to exist when:
where pipe nominal diameter, in. resultant of movements to be absorbed by pipe, in. L = developed length of pipe, ft U = straight-line distance betl\-een anchors, ft y
= =
Flexibility should be -Iacorporated into a piping system by the use of piping bends, loops, or offsets or the use of flexible joints. The use of expansion bends is simple and reliable. The most simple form of an expansion bend is U-shaped as illustrated by Fig. 5. The departure from the shortest run of piping between the two anchor points is obvious. The points shown as anchored must be considered to be anchored not only in position but also
PIPING SYSTEMS
MARINE ENGINEERING
680
With Restraining Forces
No Restraining Forces Fig. 5
Exponsion U-bend
Anchor
Point
J.-
7Rg. 6
---------
No Restraining Forces S h o w n
Three-dimension01 expansion bend
so that the piping is held without angular deflection. In order to return the pipe to the position it had before thermal expansion, only a single. force is necessary. However, the action of this force results in the pipe having an angular deflection a t this point and to compensate for this a restraining moment must be added. The expansion stresses in the pipe in this case are entirely bending stresses. The deeper the U-bend between the anchored points, the less the stresses for a given temperature change. In actual practice expansion bends frequently are not in a single plane as in the case of the U-bend but occur as three-dimensional bends as shown in Fig. 6. The flexibility is increased greatly over that for the single plane bend by the fact that, for movement in any direction, one of the three legs is always in torsion. Usually the additional thermal expansion in the third dimension does not offset the gain in flexibility which the torsional leg develops. Three-dimensional bends can be used advantageously whenever space restrictions do not permit sufficiently large single-plane bends to reduce the stresses to the desired values. Piping should be adequately supported by hangers or guides, so that the weight of the piping is not excessively transmitted to valves and fittings, and the inertia load of the piping due to vibration and pitching and rolling of the vessel are within acceptable limits. Pipe supports should be designed and arranged so as not to restrain thermal expansion of the piping and to insure that excessive loads are not imposed on connected equipment. Anchors, pivots, and restraints should be fabricated and installed so as to secure the desired points of piping i ~ relatively fixed positions and permit expansion and contraction in opposite directions. . Supports should be installed for heavy valves and
fittings to prevent their weight from overloadilig tlie attached piping or causir~gvibration problems. Tlie number of supports installed, the type selected, alld tlieir location should prevent excessive vibration of piping under all operating conditions; but they should not restrain the piping to such an extent as to cause an exccssive transfer of load from support to piping or from support to support. The location and type of support selected should. prevent excessive stress from being transmitted from the piping to machinery, equipment, or ship structure. Stress calculations should be made to determine the magnitude and direction of the forces and momerlts a t all terminal connections and anchor and junctiod points, as well as the resultant bending stress, and longitudinal pressure stress, torsional stress, and combined expansion stress a t all such points. Definitive requirements regarding the calculations of pipe expansion stresses and criteria for ensuring that adequate flexibility is provided are enumerated in the Coast Guard rules [I]. 2.6 Piping Joints. In order to provide piping systems of maximum reliability, welded and brazed joints should be used to the maximum extent practicable. I t is especially important that the joints in areas inaccessible for maintenance be welded or brazed where practicable. Additionally, the number of joints should be miliimized through the maximum use of pipe bends. Components welded in the piping should be accessible for repair, reseating and overhaul in place and should be located to permit removal, rewelding, preheating, and stress relieving in the event major repair or replacement is necessary. Complex assemblies, such as groups of valves, strainers, and traps in a high-pressure drain system; that cannot be satisfactorily repaired in place and ordinarily require periodic removal, should be demountable as assemblies by using a minimum number of flanged joints between the root valves. Cutout valves, other than root valves a t drain connections, may have a flanged end next to the machinery or equipment and the other end urelded, provided such valves can be repaired in place, and the breaking of the one flanged joint provides sufficient portability for overhaul of the machinery or equipment. Similarly, cutout valves in header and other fabricated assemblies may have one flanged and one welded end for attachment to flanged valves or takedown joints and welded components. , On other systems, takedown joints should be used where necessary for portability. On systems that havc brazing components, such as heating coils in tanks, \vhcre systcm tightness is more important than optimum portability, talicdoi~na t locations other than a t flarlgcd valvcs is pcrmittcd by heating a t brazcd joints for disasscmbly and brazing upori reassembly. Ho\vcvcr, wlicrc acccss is insufJicicljt for disasscrnbly by hcating and rcasscmbly by brasirrg, r takedoivn joints should bc installctl. Flanged and ulliorl joints sllould be locatcd \\ Ircrc t1rc.y are lcast affcctcd by pipclitlc forces (lu(: to tllc1-111:d expalision or othcr causcs. Gor~cr:illy,this rc.cl~iirc.stllc
(0) SCREWED
(b) SCREWED
(cT SLIP-ON
AND WELDED
(f) LAP JOINT
WELDED
(9) WELDED NECK Fig. 7
(h) FLARED
(dl
SOCKET WELDED
( i ) BRONZE HIGH-HUB
(e) PLATE
(J) FLANGED PIPE
Flange attachments
joints to be located away from bends or offsets. In no lines exceeding 8 in. nominal pipe size may be designed case should a welding or brazing flange be located on a for a pressure of 25 psi. pipe or fitting radius. Stop-check type cutout valves or combinations of gate, Mitre joints should be used only in piping such as air globe, or angle and check valves should be installed escapes and overflows where their use would not cause where a reversal of flow is possible and would be detrian objectionable pressure drop or turbulence in the fluid mental to proper functioning of the system. Where a flow. Branch connections should be located to minimize gate, globe, or angle valve and a check valve are required turbulent flow, and the type used (crosses, sweep tees, to be located near each other in a pipeline, one stop-check Y's, and lateral fittings) should be selected as required valve may be installed instead. The type of check for the flow characteristics. valve selected (swing, lift, ball, or special) should be the Valves such as safety, relief, regulating, and governing one best suited for the service*condition. Special check valves which require frequent removal for mainte~ance, valves having springs, hold-open gear, or other special should be flanged. Flanged connections in piping devices are sometimes required by the system design. systems containing combustible liquids under pressure Horizontally mounted swing-check valves should be in the vicinity of hot exposed surfaces having a tempera- installed in a fore-and-aft position. ture over 450 l? should be provided with protective All automatic operating valves such as reducing, shielding around the flanged joint. governing, temperature and pressure regulating, back Figure 7 illustrates some of the common alternatives pressure, and unloading valves should be sized on the when selecting the type of flange attachment for a basis of capacity requirements. They should have the particular application. For a detailed description of the control sensitivity and adjustment necessary for all limitations of the various types of flange attachments, operating conditions. Where extreme sensitivity is refer to the Coast Guard rules [I]. required, air diaphragm operated control valves may be 2.7 Valves and Fittings. It is suggested that fittings installed. The design of manually operated throttle such as elbows and tees-be designed to the requirements valves and their operating mechanisms should be set forth in reference [7]. reviewed to ensure that they have the necessary sensiValves should close with a right-hand (clockwise) tivity of control. motion of the handwheels when facing the end of the Gate valves should not be installed for throttling valve stem. Valves should be of the rising stem type, service. preferably with the stem threads external to the valve Cutout valves should be installed to isolate machinery, body. Where conditions do not permit the use of a equipment, and piping components as necessary for rising stem valve, a. nonrising stem valve may be used operating conditions, maing~anceand overhaul, and if it is provided with an indicator which clearly shows damage control. The type of cutout valve selected whether the valve is open or closed. However, it is not (gate, butterfly, globe, ,or angle) should be the one bestnecessary for valves installed in tanks and operated by suited for the service condition. Generally a gate type reach rods to be fitted with indicators. should be installed where pressure drop or turbulent Valves, flanges, and fittings should be designed for the flow characteristics of globe and angle types would be maximum pressure to which they may be subjected, but detrimental to the system. Cocks should not be in no case should the design pressure be less than 50 psi. installed in place of cutout valves unless they are necesHowever, large fabricated ballast manifolds connecting sary for proper operation of the system.
682
MARINE ENGINEERING
Forged or cast steel may be used for the construction of valves and fittings for any system without pressure limihtions. but there are temperature limitations as ------ ....indicated 'by Table 1. ~ressure and temperature limitations for other common valve and fitting materials are also given in Table 1. 2.8 Sea Connections. Suction chests should be located where they are not prone to pick up fluid discharged by an overboard connection. Fire pump sea suction chests should be located where there is no practicable possibility of their picking up fhmmable fluid from an overboard discharge connection. Suction chests should be located to avoid high entrance losses and suction difliculties with the ship underway. The possibility of suction chests becoming airbound under conditions of roll, pitch, or astern operation and the intake of entrained air carried along the submerged hull are also design considerations. Sea chests and overboard fittings should be clear of bilge keels and other hull projections. Their location should minimize interference with docking blocks. They should alm be located clear of butts in shell and innerbottom plating, and clear of welded joints insofar as practicable. Where intersections between fittings and welded structural joints are unavoidable, such as may be the case with main condenser aea chests, the line of intersection of the welded joint should be normal to the structural joint. Suction chests should have perforated strainer plates a t their junction with the shell. Where practicable, the size of perforations should be smaller than the diameter of the heat exchanger tubes to minimize the possibility of fouling of flow passages. The clear area through strainers should be a t least 1.5 times the area of the inlet port of the sea valve. Burrs and sharp edges should be removed from perforations. Suction chests should be provided with a means for clearing the strainer. Steam may be used where it is readily available; otherwise, compressed air can be used.
For this purpose, a connection to the sea chest should be installed outboard of the sea chest valve and so arranged that the chest can be blown out in the most effective manner. Nozzles on sea chests and overboard fittings should be as short as practicable and should have gusset plates welded on when necessary to prevent excessive stress a t the welded connection to the plating. I n way of the innerbottom, sea chests should extend between the shell plating and the innerbottom plating with the mnge for attachment of the sea valve as close as practicable to the innerbottom plating. A sea valve should be installed adjacent to each suction sea, chest and overboard discharge connection. Where more than one pump is connected to a sea suction line or to an overboard discharge line, a valve should be installed in each branch to permit isolation of each pump. Where a pump is located in a compartment other than that in which the suction sea chest or overboard discharge fitting is located, an additiondl valve should be installed in the compartment in which the pump is installed. A sea valve should be installed adjacent to a sea chest which serves manifolds for ballast tanks from the sea. If more than one manifold is served by a sea chest, an additional valve should be installed .- - in each branch to permit isolation of each manifold. Insofar as practicable, sea valves should be of the gate type. Sea valves of the gate type should have stems of one-piece construction and the stem should not be attached to the disk by use of pins. Butterfly valves may also be used. An angle valve should be used where the installation of a gate valve or a butterfly valve is impracticable. Sea valves in spaces not normally manned should have a remote operating system with control from the deck above. Malleable iron and nonductile cast iron are not suitable materials for connections to the shell plating below the freeboard deck. Ductile materials such as steel, bronze, or nodular cast iron should be used instead. Steel plate is the material commonly selected.
Section 3 Piping Systems 3.2 Main Steam System. The purpose of the Steam piping for ships propelled main steam piping system is to deliver steam from the by steam turbine propulsion plants consists of two principal systems, main steam and auxiliiry steam. boilers to the main propulsion turbines for both ahead The latter system is further subdivided into secondary and astern operation. The turbogenerators and the s u ~ ~-l l v-i nauxiliaries, g heat exchangers, deck boiler drum desuperheaters are also served by this svstems -" machinery, heating systems, laundries, etc. Both the system. With the location of the propelling engine or engines main and auxiliary steam systems, and particularly those capabilities essential to the safe operation of the ship, as well as that of the boiler (or boilers) having been must be able to operate individually or collectively determined, the next task is to arrange the main steam without interference. For this purpose, the systems pipes properly. This is comparatively simple in a vessel are provided with the capability of being segregated where there is only one boiler and one engine, but it becomes more complicated as the number of boilers through the use of cutout valves.
3.1
Steam Piping.
I I
i I
I
PIPING SYSTEMS
increases and still more so where there are also two or more propulsion units. I n order to take care of expansion in the pipes, a sufficient number of bends must be used between anchor points such as fixed supports and attachments to boilers and machinery. Lines should be arranged with a definite slope either toward the boiler or toward the throttle; in any case, provisions must be made to drain the line. Pockets in the lihe must be avoided, if possible; if unavoidable, they must be adequately drained. After the pipe sizes have been determined on the basis of capacity, the thickness of pipes should be computed for the pressure involved. The pressure calculations consider the hoop stresses in the pipe and it is also necessary, particularly when high steam pressure is used in combination with high temperatures, to consider the stresses due to the expansion of the pipes. It should be noted that increasing the pipe thickness, while reducing the hoop stresses, will, on account of added stiffness, increase the anchor loads due to expansion and it may become necessary to rearrange the piping in order that the total reaction loads may be kept within the allowable limit. Valves and fittings should be located, insofar as practicable, in accessible positions. When it is necessary to fit a valve in a place where it cannot readily be reached from the floor or grating, access must be provided or an operating extension should be fitted. I n addition to the stop valves a t the boilers, another valve should be fitted in the steam main ahead of the throttle. Any branch line emanating from the main should have a valve a t the branch from the main as well as a t the unit it supplies, in order to provide double valve protection and to facilitate overhauling. Hangers or other types of supports must be provided not only to carry the weight of the piping but also to restrain lateral movement where such is not desired. Fixed supports should be located in such a manner that bends in the piping between them will provide sufficient flexibility to permit movements due to expansion. I n each case the locations of supports depend on the general arrangement; where practicable, the pipes should be placed so as to facilitate simple supports from adjacent ship structure. Figure 8 is a typical main steam diagrarp for a merchant ship. The system shown has two boilers; consequently, as required by the USCG rules, boiler cutout valves of the stop-check type are provided to prevent crossflow from an operating boiler to one not on the line. Each superheater outlet valve is 6 in. nominal size and the turbine root valve is 8 in. nominal size. For all steam valves larger than 6 in., the classification society rules [3] require that the valves be fitted with a bypass. The purpose of a bypass is to equalize the pressure on both sides of the main valve so that it will be easier to open. The bypass valve also permits the pressure and temperature to be raised slowly in the downstream piping when corning up to operating conditions.
683
For this particular design, it was specified that the throttle pressure for the main turbines and the turbogenerator turbines be 850 psig. With a boiler superheater outlet pressure of 875 psig, this allows only 25psi pressure loss from the boiler to the turbine throttles. The 6-in. boiler line from each superheater outlet and the &in. combined boiler flow line to the main throttle satisfy the condition of 25-psi maximum pressure loss to the main turbine throttle. For the turbogenerators, the 3-in. line from the 6-in. superheater outlet line to each turbogenerator meets the requirement of 25-psi maximum pressure loss t o the turbogenerator turbine throttle. 3.3 Boiler Pressure Auxiliary Steam System. The purpose of the boiler pressure auxiliary steam system is to supply steam directly or by way of reducing stations to all the steam-driven auxiliaries and ship services not served by the main steam system. A typical diagram of such a system is shown in Fig. 9. Boiler pressure steam is supplied from the boiler desuperheater directly to the boiler soot blowers and the turbines driving the main feed pumps. Through either reducing station regulating valves or a thermostatic control valve, the auxiliary steam system also supplies the following components: 1. Main air ejectors, turbogenerator air ejectors, and distilling plant air ejectors. 2. Main turbine and turbogenerator gland sealing steam regulators. 3. Steam atomization for the boiler fuel oil burners. 4. Steam to the whistle. 5. Steam to the contamifhted steam system heat exchanger for the evaporation of steam for contaminated services. The normal steam source of this system when under way is from the H P turbine bleed steam connection; however, for slow-speed conditions, port operation, maneuvering, and astern operation, live steam must be supplied to this heat exchanger. When the heat exchanger cannot be used, live steam may be used to serve the contaminated steam system via a bypass around the heat exchanger. 6. Steam for heating seawater for the tank cleaning system via a thermostatic temperature control valve. 7. Steam to the auxiliary exhaust system. The normal source of steam for the auxiliary exhaust system is the exhaust steam from the w i n feed pumps and bleed steam from the main turbines. For conditions when bleed stehm is not available (i.e., slow ship speed, maneuvering, or when in port), live steam must be used to supplement the supply of exhaust steam from the main feed pump. Due to the fact that $he auxiliary steam lines are smaller in size and a t legst in part subject to lower pressures than the main steam, some of the difficulties encountered in the proper arrangement of the latter are eliminated. The inherent flexibility of the smaller piping makes it more convenient to take care of expansion and the necessary supports can be provided more readily. However, the detail design is important, and in arranging the distribution of the steam to the various auxiliary
MARINE ENGINEERING
PlPlNG SYSTEMS
OM DRUM PILOT' SAFETY VALVE
BOILER N 0 . 2
AUX. STM. SYS.
LEGEND OF SYMBOLS W
-11
I-
GLOBE STOP VALVE GLOeE STOP CHECK VALVE SUPHTRSAFETYVALVE PILOTACT. ANGLE STOP-CHECK VALVE RINIG (L BLIND FLANGE
@
REMOTE READING THERMOMETER PRESSURE GAGE -I;+ ORIFICE D THERMOMETER " ~ " T V P ESTRAINER
@
NOTES
AHEADTHROTTLE
T O -+ -,
STM. FLOW TRANSMITTER
Cc
TELL-TALE
'
4. 5.
ATMOS 10. I I.
Fig. 8
A L L GAGES ARE TO BE F I T T E D WlTH A VALVE AT T H E MAlN AND ONE AT THE GAGE EXCEPT WHERE GAGE I S LOCATED ON MAIN. THEN ONE VALVE WILL BE USED. A L L PlPE SIZES SPECIFIED ARE NOMINAL, WlTH 1.P.S OUTSIDE DIAMETERS. A L L PIPING. VALVES AND FITTINGS TO MEET THE REQUIREMENTS OF THE U S C G ANDABS. PlPlNG TO BE BENT T O A RADIUS OF 5 DIAMETERS. PROVISION FOR EXPANSION TO BE MADE IN A L L PIPES WITHOUT THE USE OF SLIP JOINTS OR CREASED OR CORREGATED PlPE BENDS. CONTROL AND INSTRUMENT PlPlNG L E D FROM SUPERHEATED STEAM LINES SHALL BE SOARRANGED AS TO PRECLUDE CONDENSATE FROM ENTERING THE SUPERHEATED STEAM L I N E S IN ORDER TO GUARD AGAINST THE POSSIBILITY OF CRACKING DUE TO THERMAL SHOCK. FLANGES FOR MAlN STEAM TO BE USED ONLY AS REOUIREO AND WHERE ATTACHEO TO TURBINES AND EOUIPMENT. LINE JOINTS TO BE WELDED USING BACKING RINGS. OPERATING CONDITIONS AT THE BOILER SUPERHEATER OUTLET ARE 8 7 5 P S l G AT 930.F TOTAL TEMPERATURE. ALL PIPE LINES A R E TO BE PROVIDED WITH DRAINS WHERE NECESSARY. LOCATE O N M A I N CONSOLE. FOR EMERGENCY OPERATION ON H.P. TURBINE ONLY, TEMPORARY PlPlNG I S INSTALLED AS SHOWN BY DOTTED LINES. FOR OPERATION WlTH THE L.P. TURBINE ONLY, THE BELLOWS I S REMOVED FROM THE CROSSOVER PlPlNG AND DESUPERHEATED STEAM IS INTRODUCED TO THE L.P.TURBlNE BY THE TEMPORARY PIPING.
I
FLOWS AND VELOCITIES
NOTES
m MEET REDUIREMENTS OF THE USCG ANDABS. PlPlNG ARRANGEMENTS SHALL B E DESIGNED WITH DUE REGARD FOR THERMAL EXPaNSION AND ALLOWABLE LOADS ON PUMPS AND OTHER ALL PIPING. VALVES AND FITTINGS
..
EOLI~PMENT
FLANGES FOR AUXILIARY STEAM ARE TO BE USED ONLY WERE ATTACHEO TO VALVES FITTINGS OR EQUIPMENT. BACKING RINGS. LINE JOINTS ARE T O ~ WELDEDUSINO E
Diagram of main steam for a 24,000-shp ship
PlPlNG f0BE-~ENTTO A RADIUSOF 5DIAMETERS. A L L PlPE LINES ARE TO BE PROVIDED WITH DRAINS WERE NECESSARY.
machines, the pipe size must be determined as required for efficient service and the design details must be selected to suit the pressure involved. Some of the auxiliaries are usually designed for full boiler pressure while for others a reduced pressure is required. Where this is the case, a reducing valve is fitted in the branch line and is set to deliver steam a t the required pressure. A stop valve is installed on each side of the reducing valve and a bypass is provided so that steam may be supplied and regulated by hand in the event that the automatic valve should require overhauling. A relief valve must be fitted in the low-
pressure lines to protect both p$ing and machinery which are not designed for the higher pressure. A pressure gage is also provided to indicate the reduced steam pressure. 3.4 Turbine Bleed Steam. As discussed in Chapter 2, the heat available from the steam is progressively reduced as it passes through the main turbines, and only a limited portion of the steam's available energy can be converted to useful work by the turbine. Ultimately, after the steam leaves the LP turbine, a large fraction of the heat in the steam is exhausted into the main condenser. By extracting steam from the turbine,
Fig. 9
Diagram of auxiliary and high-pressure bleed steam
wetern heat exchanger via a regulating valve. Highpressure bleed s t a m is the least economical of the bleed atesm dbmatives since it contains a considerable
amount of heat available to the turbine; nevertheless, u&g it is an economic preference to the use of live
a*m. 2. I n t e r m e h t e bleed connection. There may be mveral intermediate bleed connections as indicated by Fig. 10. S h r n from this connection can be used to supplement the auxiliary exhaust for boiler air heaters, f d w a t e r heating, and to provide for ship services such as qusl4ers heating, hobwater heating, galley and other hotel services. The boiler &r heaters are served directly f r ~ m&is bleed connection. On the other hand, the skmn pressure to the other services is not allowed to iu&uate; et pressure regulating valve is provided to admiti demperheated steam into the system as required tb maintain the pressure at a set level. Intermediate ateern can be used at about one half the cost of live steam. 3. Low-pressure bleed connection. The use of lowpressurn bleed atearn is an effective means of improving the t h d efficiency of a cycle. At this point very little of the h a t remaining in the steam is available to the turbine and &&ti remaining would be largely exbauded ik~hthe condenser. However, the use of sham at &is ~ Q R pressure is limited; its major applicatbm w e to wrvice the low-pressure feedwater heater md the distilling plant. Pn ge.nera1, the design of the turbine bleed system &auld be: sueh that the heat available in the steam is utilimd t~ he& advantage f o m the time it leaves the bailer until it, is returned to the boiler as feedwater. Eat bftlanees, web as: described in Chapter 2, in ~ h i c h the: h a t req~rementsd the various shipboard services me andyzea, w e the design t o d s used to establish the mast+appropriate turbine bleed steam system. b5. AwdTmry Exhaust System. The auxiliary exbus%rsystem wlbcts the steam exhausted by the vwiaus au~nirt~g machinery wmponents and deIivers the &earn ta EL number of services. The design concepts w d ta establish the optimum auxiliary exhaust system amatgemen&ape much akin to t h e used in the design af the t~rhinebleed system. Services typically supplied b y the auxiliary exhaplat system include the ship's space besting system, hotr~vaterheaters, galby, and other hotd equipme~t. The U.S. Coast Guard rules [ll xequire that the seam pressure far space heating not exmd 45 psig; mptsequently, by utifiaing steam from the t$ux&ry exhaust iyrstern, this service is provided in the most emnomieal manner powible. Qmd the primmy objectives of the suxilirtry exhaust sy&em is to m p p I ~stmm to the dmerating feed tank. In the b e r a t h g f e d tank, the exhaust steam comes in& direct csnbeti wikh the candensate which is sprayed (we Chapter 14). ahia mixing of the steam and wadensate b t a ekre fedwater to its saturation temperatare, thus causing the BXimldSir md gases (particuhrly ctqge~ which ~ is highly cmsive) to be nIessed from the Pdmkr.
The exhaust system may d e o t s h m fmm a number vessels9 B a t pctg sysknas, zilaehiy, of a&ry turbines in &B m a a d o&er ieg em & q e due to i e x ~ v e pttrticular design &own in exhaust steam is the fed ee% att a p e s w e not exceeding that fap uhich the 3.6 M y Vdve a d js &signed The reg&tssy bodies require h i or ~ R t valves y must be relief v o r 1 m b e ~ h p 8 . e n S ~ h ~
688
MARINE ENGINEERING
pressure of more than 3 percent above the set pressure of the valve for steam, 10 percent for gases or vapors, and 10 to 25 percent for liquids. A safety valve is broadly defined as a valve designed to open automatically to relieve excess pressure in the pressure vessel or element to which it is attached. Safety valves can, however, be further defined according to their specific type. The more common types of safety valves are defined briefly as follows: Spring-loaded safety value: A valve fitted with a spring which normally holds the valve disk in a closed position and allows it to open and close a t predetermined pressures. Pressure-loaded pilot-actuated safety valve: One which is held in the closed position by steam pressure and controlled in operation by a pilot-actuator valve. Spriny-loaded pilotactuated safety valve: One in which a spring is used in the conventional way to hold the disk against the seat; however, it has a piston attached to.the spindle and enclosed within a cylinder which, when subjected to a limiting or set pressure, unbalances the spring load, thereby opening the valve. Spriny-loaded pilot valve: A conventional safety valve designed to actuate another spring-loaded safety valve through a pressure transmitting line led from the body of the pilot valve. Lever safety valve: A valve having the disk held on its seat by the action of a weight hung on a lever which is pivoted on a fulcrum. Relief valve: Valves having the same general design characteristics as safety valves, provided to relieve excess pressure. A relief valve may, however, be used for a service where an approved safety valve is not required by the regulatory body requirements. Sentinel valve: A small valve that is used to warn the operator of a malfunction and serves as a telltale. The set pressure is usually above the normal working pressure but lower than the pressure the system can withstand. A relief valve, or similar feature, is installed in the outlet piping from each pressure reducing station. The relief valve is located downstream of both the reducing valve and the bypass valve and is sized on the assumption that the reducing valve could stick wide open. Relief valves for steam or air service are provided with a substantial lifting device so that the disk can be lifted from its seat when the pressure below the seat is 75 percent of that a t which the valve is set to blow. If the design of a relief valve is such that liquid can collect on the discharge side of the disk, the valve should be equipped with a drain a t the lowest point where liauid can collect. ~ s c a piping ~ e from the boiler drum and superheater safety valves should be designed to have an area of not less than that of the combined areas of the outlets of all valves discharging into the escape piping and should be led as near vertical as practicable to the atmosphere. The piping should be adequately supported and installed so that little stress is transmitted to the safety valve body. Safety or relief valve discharges, when permitted to
terminate in the machinery space, should be led below the floor plates or to a remote position to minimize personnel hazards. 3.7 Main Feed System. The purpose of the main feed system is to provide a continuous supply of heated and deaerated water to the boilers. The pump in the main feed system takes suction from the deaerating feed tank. Boiler feedwater comprises condensate from the main and auxiliary systems, condensed heating steam, and miscellaneous freshwater drains. These sources of feedwater are collected in the deaerating feed tank. As discussed in connection with the auxiliary exhaust system, auxiliary exhaust steam is used to heat the feedwater in the deaeratisg feed tank. The air and dissolved gases in the feedwater are thereby released and are either vented to the machinery space or piped to the gland exhaust condenser. The heated and deaerated feedwater collects in the bottom portion of the deaerating feed tank where it is a source of feedwater for the main feed svstem. The main feed pumps take suction from the deaerating feed tank and discharge the feedwater to the boilers. I n the system illustrated by Fig. 12 at least two feed pumps must be provided in compliance with the regulatory body requirements. The regulatory body requirements further specify that the only function these pumps are permitted to perform is to supply feedwater into the boilers. The feed discharge piping must be designed for the pump relief valve setting or the pump shutoff head if a relief valve is not provided. To comply with the regulatory body requirements, two feed lines must be used to serve the boilers. In the particular design shown, feedwatgr can be delivered through the main feed line or through the auxiliary feed line. Each pump discharge is fitted with a stop-check valve in the main feed line and in the auxiliary feed line. The main feed line is provided with a feedwater regulator, a feed stop valve, and a feed stop-check valve. The auxiliary feed line is provided with a feed stop valve and a feed stop-check valve. Each boiler in the design illustrated by Fig. 12 is fitted with an economizer; therefore, the feed is circulated through the economizer before being discharged into the boiler drum. If the design had included more than two stages of feedwater heating, the subsequent heating stages would also have been in the feed pump discharge % line. Under normal operating conditions, the level of the water in the boiler drums is automatically maintained, usually a t mid-level of the boiler drum, by the feedwater regulator. Since the auxiliary feed line does not contain a feedwater regulator, the level in the boiler drum must be manually regulated by throttling the feed stop-check valve when the boiler is being supplied via the auxiliary feed line. The auxiliary feed stop valve should not be used for this pdrpose; i t is essential that wear on this t a minimum so that i t can serve as a valve be k e ~ to tight stop ialye in the event positiGe isolation of the boiler is required.
PIPING SYSTEMS
j - - j
L .J
D!ti5 RUB BE^ BILGE.
'
HOSE
CHEMICAL FEE0 TK 15 GAL.)
~ E S CABINET T
NOTES 1. ALL PIPING. VALVES AN0 FITTINGSARE TOMEET T H E REOUIREMENTSOFTHEUSCG AND A B S ALL PIPE SIZES SPECIFIED ARE WITH IPS OUTSIDE
z. nlaucTcmc
NOMINAL
5. VALVES REFERRING TO THIS NOTE SHALL BE LOCATEOONOR OPERATEOFROM THE MACHINERY FLAT. 6. THE MOTOR OPERATED VALVE AT THE MAIN FEED PUMP SUCTION I S CONTROLLED FROM THE MAIN CONSOLE 7. STOP THE VACUUM VALVES UNDER SUBJECT THE TOVALVE VACUUM SEAT S H A L L BE WITH
INSTALLED
8. EACH DECK FILLING CONNECTION S W ~BEL FITTED WlTH HOSE CONNECTIONS AND PROVIDED WITH.LOCK CAP AND CHAIN. EACH CONNECTION SHALL BE MARKED WlTH AN ENGRAVED LABEL PLATE INDICATING THE SERVICE 9. VENT PIPE THERMINAUS SHALL b E FITTED WlTH DOUBLE SCREENS WHICH SHALL BE REMOVABLE FOR CLEANING. M E INNER SCREEN SHALL HAVE I I I Z " S O U A R E OPENINGS T H E OUTER SCREEN SHALL BE FITTED IIP~~FROMTHE INN& SCREEN A N 0 SHALL BE OFSTAINLESS STEEL WIRE WlTH II~~SOUARE
-. -...,. "".
nDmmr.c
10. CONTROL VALVE SENSING LINES SHALL BE A S S H o R b A S POSSIBLE. I I. LOCATE O N MAINCONSOLE 12. LOCATE GAGE TO BE READ ~TCOMBUSTION CONTROL CONS&. '
Y Fig. 12 Diagram-feed, condensate boiler blow, and boiler water sampling
~CUNNEL
PIPING SYSTEMS
Even when 6;Re flaw of @~hf& fe$Twa6er is h w ar h m h d sufE&ently hi& ailhove Bhe M x a t e r r$dn dQQ-~~QVZexkb for s h o ~h t tmah4wohr muse oakQhu8 hantiag tank &at it em drain hy gr&vity to this h k . ts b9, &rodate& t h ~ u g bOPre w i n fwd pump in ~pdw A is m a E y desimMa ta amid a pwphg amgement, as ea, pump and motar end tihe required p i e g and eon~EB cod the p m p Fw thiw puqmse?wdwlating t the system is have h e n pr~vkled. Them lines fishre flaw bmm the W s em eEmh~kd,md the initial e ~ s d pump" dheharge m d rwira$a&e the fiow back to the F & W ~ as are maintenance snd space requirements. A rYypiml wmngement, for a 60ndensate sys.tem is detaamting f cedi tank. High md bw lexgls am mainhined in 8he deaemtbg ilkusbated by Kg. 12. En &e sys4enr &own there is Qne feedvaa$~Beatm 6 a the d e - b spill an4 the m k t main emdenser semd by Cw~amain condensate pump; up f e d & e ~ * mpeeiTFely, each of f i c h ere mta- the re&at~ry bedies q u i r e &at two means be prayid& mpItim& mntded by BigB4e;vd md law-led mnW far dl6&twging the condensatefmm Ohe maah wndewem. Mker being pumped fmm the main cmdem~). the caw pilob. The high-level spill d y e bps the &emate in the h e feeding a h fimbsfaga fedwater B&ex BQ the EEmmte st.tcceely passe8 tbmugh the main stir ejector &tilled water tank; #e ~QW-level mrtKeup f e d valve m a d e a ~ r s % the $and Eed'c-off condenser, and the firstdumps W&&F fmm the disblled water tank to &e m& stage feed heater enroute ta the deaemting feed tank. eormdenser; see Fi.12. 3-9 S a w d e ~Sysams. The pwpse of smwatex As indicated in Fig. 12, water r x t m can ~ ~ be $&en q & e m e m ~arn-dfiuenship is tm pmvide seawater fram the W e r water wa.110, deaexating feed b k , and d m t , ta the main and d i a r y mn4ensers and the c~lhtamin~txd skwn pnemhr amd the mm$es a n he vmiaus heat exchangers (e-g., the main Eubriating-aiE cooled aind tmted in cmvelrient access &aa sink. matera, lubricating-02 d e r s for the turbine-dkea TOf&ciE&aQe the rnaintenaaw of W e r water &emis- aw5Eari.e~~ turbcrgeneretor 8ir and Iwtb~ioaGng-ailw'ters, try, a & e m i d mixing tank and a chemical feed tank m a i ~feed pump IubPicating wler, and ship service and eargQmfG@:e&im~~PEdensem].In most instances, are a h provided w shown in Fig. 12. During perids of shiplay-up, €hebailew &remzmtimee independenti mawater systems are pmvidd ta supply filled with watep, Wdez far t h k purpose can be drown the: emfanant im e a q ~and shig service refrigerating &am distilled wabr tank. A apecid esnxamthn mndemem and ot:h~rmiscellanmus semiem tm %wt~ate$ feam &e distilled w a t e ~tmk ta Me f e d pump s,uetiaaz by Fig. 13. The regdatwy bodies mquiPe that a mala drculating lines Erm been p d e d far this puqms~. (Tbm are dm same bailer Iray-wp t e ~ h ofn the ~ *dry" ~ type in pump and rn emergenoy meam d eireulathg water which the wahepsitde G*comp1eteBy &abed and t3k&ceanb t h u & the main condenser be, provided. When dud me i n t m d u d ;see Chapter 3 for additional ~ m d m -in eireulakjngMmps are ins+dE& (with b t k required for Pull-power design conditians), the emergency requim wx t h i ~mhj&-) 8.8. eZPPbtb a d CrndemseCe Systems. The p q o w of men& is m n & d to he satisfied; QP the emergency the drain and mndenmte sysfmn~fiis ta milmt the d ' ] : s j ! ~ provbiwbns,may e m i s &~f a cannection fmm an iElcIepe* md txwxkmat4 fmm th& vtwhu~sset~mes~ deliver deab, power pump &at is fitted hetween the m& %hemta the deztera&g feed tank. clreulating pump an& the m~denser. Dud main are cHm&m&iw af drnim: hq+mm ejircdrrting pump8 and various cxms wmeetiona fzr nd hw-pressurn dmba IEigh-pmessu~edrains nomd and emergency semiee b $hevaio\rs system are i&& &me fmm the main ateam fin% ba:ie~preswure mmetimw provided M iindiesbed i~ Fig-13. With some ships, a smap system forces &he seawater auxih~ a t a m lines, m d other high-pressure pockets. These Mils me piped b m high-pm~e.cfrtah main bhugb the aondemr [82. The ppewure bead that @a Yre deveEaped:by o snoop ia limited sueh &a,& 10is generally whi& diwhqps $ i t b%~z the dew~3imgf d tan& use samps with t m p a ~candenst3~8. ~ The law-pscssme (or:freshwat,er) &&in mi& OQ$~@& not feasible the &mias from the main turbine, OE~-EUB~&OP~ A d d i t i ~ d sinw ~ , the pressme h e 4 d ~ d ~ by p d tihihe s ~the ~ d ship'^ q - ~ ~ se dq ~ aemga ~ & ~ dLt3W $w& g l d exhawt m&hsert low-pewre W W IS~ p r ~ p s ~ t i t~ earnat b% employed w&h ~QW-SPEB~ship as the sim 4 M n t piping S L W V+dm ~ ~s p m hafiepa, b ~ b d heaters; a ~ g d q p md hun&y. The E o w - m ~ edrain fhe piping and wden~ernemss8ry t a m%intaia a, cm&im inb the freshwater &&a edl-gg tank, and s & i c i d g low ~ X I Y I T dmp I . ~ wnseu'id b m e grrohibd ~ &am t h e it can either k ~ U E QfbOl ~t&. dtw~sti~g t i d y ba~rge. Therebare, 4Ene &p@imtictn~f W X I Q k ' g ~ e ~ dconfined ly ta the higher-spmd ship8 with &&efwd tank w "~a~tl.w-dra@&" to the cxmdenser. En addlthn b the law-p~essurechin mak8there are paes cmdenms (e.g., paersenger &ipsz carga ship* msnd m v e d ather dxains which dimbarge into khe f d w & k r acc&o~d tankers), At very Iw forwad speeds and when manewex+ a drain &mting W k . Dmbs fpm pockets in &e audimy exhaust main, h w - p m w e steam, and bleed m o p does not prgduce a sufEciently large pressure head &earn piping dm &charge Into the freshwater: drain ta maintain am &equate seawater AQW. Therefo~e, when, a scoop i~xjee~an sy&m is install&, it is also mllfecting tank. The low-prwure feed heatex drains can be either n e c v ba pELPOvide a w i n ckdat;lng gpmp far IQW pumped itktD the eondensate .system, wwaumdsagged. to spe& md ahen msneuvering. the condenser, ar the law-presswe. feed heater mqy be Normally, addltktn %i tbesa. mean5 of p r ~ v i h g
I
NOTES
I. REOUIREMENTS ALL, PIPNVG. VALVES OF THEUSCG AM FITTINGS +WOABS ARE w ,WT T* 2. ADJACENT A~~ MOTORTOUN~T OPERATED ANOVALVES ALSO AT.RE THE ID EN= ;*= mOPERATING CONSOLE
3. CONDENSER HEAD V E ~ T SSHAU SLOPE
I
w TOT=-
CHEST VENT. 4. THE SEA CHEST VENT SHALLTE~WA~E OECX, 5. SWING CHECK VALVES SHALL BE ISTAL~OIN N WE AN0 AFT OIRECTION. 6. STRAINERS ARE TO HAVE PERFWTKINS~IA~R~W SEA WATER PASSAGES IN EOUIPMUYT S E ~ ~ D U~~~~~ THAT EOUPMENT ISSERWO W D ~ ~ I T I ~ STRAINERS.
~ ~ r ~ n m ,
am=
z
EMERGENCY ONE HOSE VALVE COOLING L O COF A TLINE E ~~IOWAY SHAFT IN wamu6~ SH~~FTIIL~X -L BE PRW~OEO THE MOST WITHREMOTE A HOSEBEARING. OF SUFFKIENT
~
~~d~~
.
P'PING IS TO BE BENT TO A RADIUS OF S DI~*ETER~9. ALL SEA CHEST WNNECT~ONSSHDXL BE I N S T ~ L L ~ ~ ~ THE SIDE OF THE SEA CHEST EXCEPT V E U wcw ~ -E MOUNTEO ON THE TOP OF THE CHEST. 10. NEEDLE VALVE M U S T S OPENED WHEN 7 ~ ~ ~ c . m ~ SE~VICE T ~ APUMR ~ IS IN OPERATION AM RL-TEO~CH THE TEMPERATURE OFTHE COUTAUINATEO ~ ( U M SUCTION DOES NOT EXCEED laO-F.
692
MARINE ENGINEERING
seawater to the main condenser, yet another supply of located beneath the reduction gear. The sump should seawater is usually provided by arranging the piping have a total capacity sufficient to contain the oil in the such that either the auxiliary condenser seawater pump gravity tank, the service piping, and the normal operating or the general service pump, if such is provided, can capacity without overflowing through the main gear bearings. The main lubricating-oil pumps pump the discharge through the main condensers. Circulating water is taken from the sea by way of oil back to the gravity tank. In order to provide a satisfactory lubricating-oil drain either of two suctions. One suction, usually located near the turn of the bilge, is called the high suction and piping design, exacting attention must be paid to design the other, located very low in the ship, is referred to as details. To minimize foaming within the sump tank, all the low suction. With this arrangement clean water return lines to the sump tank should discharge below can be taken in even when operating in shallow water the normal working level. Since this arrangement precludes a flow of air through the piping, it is essential with a muddy or silty bottom. Seawater piping should be designed such that air that the drain piping be designed so as to avoid trapping pockets will not form, and the design should be such that pockets of air in the piping and thereby restricting the entrained air remains entrained as it passes through the flow of lube oil. This is accomplished by (a) liberally piping, condenser or heat exchanger, and overboard. sizing drain piping and (b) sloping all drain piping but Condenser heads should be vented, and the vent piping ensuring that the slope does not exceed about 60 deg should be sloped upward. It is generally advantageous from the horizontal (in spme cases this may. require the to combine the piping vents with the sea chest vents and piping to follow a spiral path). Additionally, all branch route them within the vessel well above the load water- lines should connect from the side of the main drain lines. For additional discussion regarding the design of line. Protective strainers should be provided to protect the lubricating-oil tanks, see Chapter 20. Two (one a standby) lubricating-oil service pumps tubes of heat exchangers provided with a seawater coolant. The mesh or perforations of the strainer each take suction from the oil sump through a strainer. baskets should be equal to or smaller than the heat Each pump discharge is fitted with a relief valve relieving exchanger tubes which they serve. Duplex strainers to the sump tank. A stop-check valve may be installed are shown for the seawater piping to the turbogenerator in each pump discharge to prevent backflowjhrough oil and air coolers in Fig. 13. Simplex strainers may be the idle pump; or, alternatively, a check valve may be used when services are supplied by multiple pumps, as located on the suction side of the pump. Check valves is the case with the seawater service pump suction piping on the suction side offer the advantage of preventing air from entering the system through the pump seals. The in Pig. 13. 3.10 Lubricating-Oil System. The propulsion plant pumps are preferably located lower than the operating lubricating-oil system provides the means to lubricate level of the oil in the sump in order to ensure that the and cool bearings and working parts within the main standby pump suction is always flooded with oil. The pumps discharge through a duplex magnetic propulsion turbines, main reduction gears, and main thrust bearing. A detailed discussion of lubrication strainer via the lubricating-oil cooler to the gravity tank. theory and lubricating systems is contained in Chapter The temperature of the lubricating oil leaving the coolers 20 and therefore will not be repeated here. Instead, the is regulated by a thermal element which senses the oil following discussion is primarily directed towards the temperature, and the oil temperature may be controlled in either of two ways: (a) a bypass valve may route the piping of lubricating systems. The lubricating-oil piping must be independent of lubricating oil around the cooler when the aooler disother piping systems. The lubricating-oil system should charge is excessively cold (a control valve bypasses oil be designed to operate satisfactorily when the vessel is so as to maintain the temperature of the oil in the range permanently inclined to an angle of 15 deg athwartship for proper lubrication), or (b) the quantity of cooling and 5 deg fore and aft, and when rolling up to 30 deg or water to the cooler may be varied to maintain the oil pitching 10 deg. While this requirement also applies to discharge temperature in the proper range. The gravity tank is provided wit6 an overflow line to other piping systems, it is of particular importance in connection with the lubricating-oil system (e.g., the slope the sump tank with an illuminated sight flow indicator of drain piping must be checked at all inclined positions, located on the main operating level. The capacity of and the oil free surface in sumps must be investigated each lubricating-oil pump is in excess of the turbine and gear requirements, thus providing a continuous flow of to ensure that the oil will be properly contained). The main turbines and reduction gears are usually oil in the overflow line. The overflow line is equipped lubricated by a gravity-pressure system in which oil is with an air separator to free the oil of entrained air supplied to the main propulsion unit from a gravity tank before being returned to the sump. The gravity tank located on the upper level of the machinery space; such is also provided with a low-level alarm switch set to a machinery arrangement is illustrated by Fig. 1, and a sound when the tank level drops below a predetermined diagram of a typical lubricating-oil system is shown in level. A low-oil-pressure indication and alarm is provided at the main gageboard. Usually, about six Fig. 14. The oil drains from the propulsion unit to a sump minutes supply of oil is contained in the gravity tank.
PIPING SYSTEMS
MARINE ENGINEERING
The capacity of the settling tank is usually sufficiently large to hold one complete change of oil. The lubricating-oil service pump motors are fitted with a control system arranged to automatically start the standby pump in case the operating pump fails. An sounds when the pressure developed by alarm the operating pump drops below a set value. The redundancy provided by the standby pump results in a highly reliable pumping capability. Nevertheless, steam-turbine-driven propulsion machinery depending on forced lubrication must be designed to shut down automaticaUy upon complete failure of the lubricating system. To accomplish this, the pump discharge main is often arranged to supply oil for the operating cylinders of the main turbine ahead and astern throttle valves. Filling of the lubricating-oil system is accomplished from the deck to the storage tank. Makeup oil to the main sump may be taken from the storage tank via a locked-valve line to the gravity tank overflow. Makeup oil to the turbogenerator sump may also be taken from the storage tank. Oil may be drawn from the settling s
For batch purification the contents of the entire system are purified. This is accomplished by draining the gravity tank to the sump tank. Then the oil in the sump tank is transferred from the sump to the settling tank with the main lubricating-oil pump. Next the oil in the settling tank is heated and allowed to settle. The purifier suction pump then takes suction from the settling tank and discharges through the lubricating-oil heater to the purifier, and the purifier discharge pump delivers the oil to the gravity tank and fills the sump tank by overflowing the gravity tank. Iror ships which have oil-lubricated stern tube bearings (as opposed to water-lubricated rubber or wood bearings), a separate lubricating-oil system is provided for the stern tube bearings. The system incorporates a gravity tank which maintains a pressure in the stern tube slightly in excess of the seawater pressure. The tank is provided with a fill connection, gravity pressure line to the stern tube bearing, sight level indicator, vent, and high-low level alarm. The system a t the stern tube itself has been designed with many variations. It may consist of a pump taking suction from a drain tank and discharging through a filter or coalescer to the aftermost stern tube seal. A thermometer is located in the return line from the stern tube and a pressure gage is provided at the pump discharge. For a more detailed discussion
regarding lubricating systems for oil-lubricated stern tube bearings, see Chapter 11.
I
3.11 Fuel-Oil Filling, Transfer, and Oily Ballast Syrkm. There are two fuel-oil piping systems aboard
ship which are closely related but are clearly distinct. These are the fuel-oil service system, and the fuel-oil transfer, filling, and oily ballast system. The fuel-oil service system includes the piping necessary to convey the oil from the settling or day service tanks to the fueloil burners. The transfer and filling system includes the low-pressure piping required for fining the tanks and for the transfer of oil from one storage tank to the settling tanks for immediate use or another storage tank when necessary for trimming the ship. In order to increase the draft of the ship and improve its seakindliness, i t is often desirable to fill empty fuel-oil tanks with seawater ballast. The piping system used to take on and discharge seawater ballast from fuel-oil and ballast tanks is known as the oily ballast system. I n addition to providing the ability to fill and transfer fuel oil, the same distribution piping to the fuel-oil tanks may
the ship below deck. Filling lines run from the main down to the tanks. The overflows are provided with adequate height to prevent overflow under normal conditions when filling. The fuel-oil transfer system is served by a fuel-oil transfer pump and, in an emergency, it may also be served by a general service pump. Fuel oil carried in the storage tanks (innerbottom or double bottom and deep tanks) is transferred to the fuel-oil service tank by the fuel-oil transfer pump. The stored fuel oil usually must be heated before it can be pumped. 3.12 Fuel-Oil Service System. I'uel-oil service pumps take suction from the service tanks and deliver the fuel oil to the burners a t the desired temperature and pressure to suit the type of oitburned. A diagram of a fuel-oil service system is shown in Fig. 16. Suction lines are provided a t two positions in the service tanks. Normal operation from the low suction is changed to the high suction in the event of excessive water in the bottom of the service tank. When water or sludge is found in the tank, the fuel-oil transfer pump may be connected to the low suction and discharge to a slop tank. The fuel-oil suction main is provided with a duplex strainer to prevent sediment from the fuel-oil tanks entering the service pumps. Oil from the service pump passes through a stop-check valve to the fuel-oil heaters,
SKETCH "A" TYPICAL SECTION THROUGH FUEL OIL TRANSFER AND BALLAST MANIFOLD.
F ~ . O V B O . BF.O.TRANS. PUMP DISCHARGE O"F.O.OVERFLOW STAN 0.8. TKS. NO. I F.O. OR
sW.
ENDOF TAILPIPE LOCATE0 ABOUT 3 " ABOVE TANK TOP P 8 s
SECTION- OVERFLOW PIPI
NOTES I.
.
DATA BASED ON FUEL OIL V l s c O ~ l ~3y0 0 0 ssU loTAL FLOW OF 2 2 0 0 BBLS/HR TO ALL TANKS ~ N OF 3 6 0 BBLSfHR TO NO I DOUBLE BOTTOM T A 3. SPEED THE F PUMP U OIL ~WHICH TRANSFER ~ DEVELOPS PUMPA IS TO A TY A E R~HEAD T ~ C AOF LT ~ O -
C ~
~
~
~
1 0 0 PSI AT 3 0 0 GPM.
'9.
l5
of fuel-oil filling, transfer, oily bollast, and overflowsinmachinery
discharge strainem, meter1 mntrol valves) and the fuelburner header. Bypass valves are provided around the meter and the control valves. Strainers in the System suction and discharge should be of the type Or of a type of being deaned without intempting the fuelmoilsupply. C0amin@) drip pans, etc., should be fitted under the strainers and and wherever necessary to Prevent oil from
spilling onto Piping plates draining to the bilge. The regulatory bodies require that at least tm-0 fueloil sewice pumps be provided and that each have sufficient capacity to supply the boilers a t their rating. Two motor-driven rotary-type fuel-oil service pumps are usually and are valved and intermnnected such that either pump can take suction from
I
696
MARINE El
either service tank. Each pump is installed such that it is capable of being overhauled while the other is in service. A relief valve is located in the discharge of each pump and is arranged to discharge into the pump suction. An automatic pressure-regulating valve is usually installed between the service pump discharge and the burner header. This valve maintains a constant pressure at the inlet side of the burner headers by recirculating oil back to the pump suction. Fuel-oil service pumps must be equipped with means of emergency shutdown from a position outside of the machinery space which will be accessible in the event of a fire in the compartment in which the pumps are located. Two fuel-oil heaters, each capable of supplying the total oil requirements of the boilers and arranged such that one can be overhauled with the other in service, are usually provided. The steam to the heater is supplied through a thermostatically controlled pressure regulator which is capable of maintaining the temperature of the oil leaving the heater at the value necessary for proper atomization of the oil being burned. A relief valve is located on the oil side of the heater to prevent excessive pressure within the heater. The fuel-oil flow to the boiler is controlled by the combustion control system. A recirculating line is provided for each boiler to return the oil to the pump suction when recircu.lating to heat the oil during a cold start-up or during a condition in which all fires in a boiler are temporarily secured. The discharge piping from the fuel-oil service pumps to the boilers must be of steel, seamless, and of schedule80 thickness. Nonferrous material cannot be used in this system except for the short lengths of tubing from the burner header to the individual burners. Piping between service pumps and burners should be located so as to be readily observable. The relief valve located at the pump discharge may discharge back to the suction side of the pump. The relief valve provided for the fuel-oil heaters should discharge to a service tank. The return line from the burners should be so arranged that the pump suction piping cannot be subjected to discharge pressure. If screwed-bonnet valves are used, the valves should be of the union-bonnet type capable of being packed under pressure. Unions are not used in pipe sizes of one-inch nominal size and above. Bushings and street ells should not be used in fuel-oil discharge piping. 3.13 Bilge and Ballast Systems. The bilge and ballast systems, while having two distinct functions to perform in the ship, are usually so interconnected that it is found advantageous to cover the two systems on the same arrangement plan. The same pumps are commonly connected to serve both systems, and the piping with its valves and fittings must be so arranged that each system can function without interference with the other. For this purpose the systems should be provided with more than one pumping capability, as required by the regula-
PIPING SYSTEMS
tory authorities, particularly for the pumping of bilges. All manned vessels must be provided with a satisfactory bilge pumping system which is capable of pumping from and draining any watertight compartment except for ballast, oil, and water tanks which have an acceptable means for fillingland emptying independent of the bilge system. The capacity and number of pumps or their equivalents to be used in the bilge system depends upon the size, type, and service of the ship. Specific requirements regarding the number of bilge pumps and their capacities are delineated in references [I, 31. The bilge pumping system must be capable of dewatering the ship under all practicable conditions after a casualty including extreme conditions of list and trim. For this purpose, wing suctions are generally fitted except in compartments that are narrow in the fore-and-aft direction and at the ends of the vessel, where one suction is sufficient. In compartments of unusual form, additional suctions may be provided. A nonreturn valve should be fitted to the end of the bilge pipe in the compartment which it serves. All bilge suctions must be fitted with suitable strainers. In the machinery spaces these should be placed so that they are easily accessible from the floor plates and should be of a type that can be easily cleaned. At the end of the bilge lines in other compartments, ordinary box strainers or their equivalent are satisfactory provided they have an open area not less than three times the area of the suction pipe. In addition to the suction strainer, easily cleaned strainers should be fitted in the line between the manifold and pumps and so arranged that the pumps will be protected against any debris that may have passed through the individual strainers. The piping design must be such that miscellaneous water in the compartments will drain to the bilge suction pipes. The bilge suctions are arranged so that water from all tank tops, watertight flats, insulated holds, peak tanks, chain lockers, and decks over peak tanks can be drained. Where drainage from particular compartments is considered undesirable, the provisions for such drainage may be omitted, provided the safety of the vessel is not impaired. The pump discharge from the bilge system is always led overboard,-and when the same pump is used for other purposes, as for ballast water, fire main, etc., other discharge connections are pwvided. Each pump may have a separate discharge through the ship's side, but it is desirable that the shell penetrations be as few as practicable; therefore, two or more discharge lines are usually combined in one outlet, each line being fitted with a valve to permit isolating the line from the others. Bilge and ballast pumping systems are arranged so as to prevent oil or water (from the sea or ballast spaces) from passing into cargo or machinery spaces or from passing from one compartment to another. The bilge and ballast mains are fitted with separate suction valves at the pumps. The number, location, and size of the bilge suctions in
RECIRC LINE -F
1
PLAN VlEW
SEE NOTE 7
NOTE NOTE 7
BRIDGE DECK
.L
SEE NOTE 8
1
1
,,
!
\BAFFLE
r---1
F
I
-, '
\SEE
I
T
NOTE B
SECTION VlEW
SHOWING BURNER FUEL OIL AND DIESEL OIL COLD BOILER START PIPING
PUMP TABLE
ROTaRI
COLD BOlLER S1ART
NOTES I. ALL PIPING, VALVES AN0 FITTINGS ARE TO MEET THE REQUIREMENTS OF THE U.S.COASTGUP.ROAN0 ABS. 2. ALL PIPE SIZES SPECIFIEOARE NOMINAL WITH I P S OUTSIDE DIAMETERS. 3. COAMINGS. ORlP PANS. ETC SHP.LL BE FlTTEOUNOERNELOILSERVlCE 4, ~ ~ ~ ~ j ~ $ ~ f : L ' " p ~ ~ ~ C T ~ ~ ~ ~ ~ S ~AytALGE.. ~ E ~ ~ ~ , " ~ M ~ T ~ L ~ ; ~ J & ~ ~ ~ ~ TEMPERATURE STEAM LINES.
5. FUELOIL SERVICE PUMPS SHALL BE CONTROLLEOFROM THE AUXILIARY BOARO. ALSO THE PUMPS SHALL HAVE PROVISIONSOUTSIOE THE MKHINERY SPACE FOR EMERGENCY SHUT OOWN. 6 . THE MAIN SOLENOIO V4LVE IS TOCLOSE IN EVENT OF LOSSOF POWER TO THE
FORCEOORAFT BLOWER. LOW WATEROR TOTALFLAME FAILURE WITHIN THE BOILER A MANUALTRIP SHALL BE PROVIOEOONTHE VALVE
7. HIGH SUCTIONPIPE SHALL TERMINATE IB'ABOVE TANK BOTTOM LOW SUCTION PIPES SHALL TERMINATE BELOW THE LEVELOF THE BOTTOM o ~ T H E T A N K I N A ORAlN WELL. 8. FUEL O h RECIRCJLAT hG L NL IS TOENTER NEAR TOPOF FUEL OIL SETTLING TAN* AN0 TERMINATE NEAR THE WTTOM OF TANK IN A"U"BENO. 9. THE BLlNER SOLENOIO VALVE IS TO CLOSE N EVENT OF BURNER FLAME fAILJRE. A MANUAL1R.P S ~ A L LLIE PROVIOEOONTHE VALVE.
Fig. 16
Diagram of fuel-oil service
system
1
MARINE ENGINEERING
the boiler and the machinery compartments must be determined when the piping plans are submitted for approval by the regulatory bodies and must be based upon the size of the compartments and the drainage arrangements. A bilge and ballast system for a typical dry-cargo ship is illustrated by Fig. 17. For the purpose of determining the required number of bilge pumps, this ship is classed as an oceangoing dry-cargo vessel over 180 f t long. Since centrifugal pumps have been provided in the system illustrated by Fig. 17 (as is the usual case with bilge and ballast systems), a central priming system has been installed. An alternative would have been to provide centrifugal pumps with their own priming system; or, where positive-displacement pumps are used, a priming system is not required. The chain locker sump and the bilge pockets, such as those around the bow thruster, are often drained by an eductor. Actuating water for the eductor is furnished by the fire main. As stated previously, the ballast system is usually. connected with the bilge system. It is usually served, in part, by the same pumps, but the piping between the pumps and the ballast tanks has no connections with the bilge system. In some cases a tank requires connections to both systems, as when the tank is used alternately for ballast or dry cargo. Under these conditions the bilge system and the ballast system must be arranged to serve the tank but so arranged that, when one is needed, the other is blanked off. The arrangement of the piping to the various ballast tanks is in general similar to that for the bilge system, the main difference being that, with the exception of the discharge overboard, the bilge piping is arranged for suction only, while the ballast. piping aonsists of-both suction from'and discharge to each ballast tank. Separate main lipes for suction and discharge should be provided to simplify operation; in any event the two should be so arranged that water may be drawn from any tank or from the sea and discharged into another tank or overboard as may be found necessary to maintain a proper trim of the vessel. The sea suctions are usually so arranged that water can be run into the tanks by gravity if the pumps should be otherwise employed, and valves must be provided wherever necessary for the proper control of the system. I n general the size of the ballast pipes is made large enough to maintain the suction lift rating of the pump down to one half the full level of the tank. A pump serving both bilge and ballast systems should have at least three valves in the suction manifold: one stop-check valve for connection to the bilge main, one stop valve for the ballast, and one stop-check valve for suction from the sea. I n the discharge manifolds, there should be a valve for ballast and a valve for the overboard discharge. When tanks are being filled by pump pressure, provision must be made to protect the tanks against a pressure greater than that for which they are designed.
For this purpose ovedows are provided. In order to avoid the necessity of comparatively large overflow pipes from each tank leading up through other compartments, where they interfere with the arrangement of these spaces, it is common practice to provide a standpipe on the main discharge from the pumps to the tanks. This pipe should be the full size of the discharge main and should have its outlet at a level not exceeding the allowable test head on the tanks. The pump discharge must be so arranged that when the tanks are being filled it will also have an unobstructed discharge overboard. . The standpipe outlet must be separate from the regular overboard discharge. In the arrangement illustrated by Fig, 17 the forepeak tank and the afterpeak tank axe connected to the clean ballast system. As required for trimming the ship, these tanks are filled from the sea or emptied by the fire and clean ballast pump. The centrifugal motor-driven fire and clean ballast pump may have to be primed before emptying the peak tanks (when the head corresponding to the level of the water in the tank is not sufficient to prime the pump) ; for this reason, it is connected to the central priming system. 3.14 Freshwater Systems. As discussed in Chapter 1, a fundamental decision to be made in the design of the freshwater system is the choice between providing sufficient tankage aboard ship so that fresh water can be purchased at the various ports of call and stored until used, and the alternative of producing fresh water aboard ship by distilling seawater in low-pressure evaporators. The most advantageous choice or compromise between the two depends upon particular circumstances;however, low-pressure evaporators are the common choice. An additional decision to be made deals with the number of grades of freshwater systems carried aboard ship. Here there is the choice between designing a system containing two grades of fresh water (potable and wash water) or only a single grade (potable). The appreximate quantity of the two grades of water can be computed from the data given in Table 3. Carrying a single grade of water greatly simplifies storage and handling facilities, and the trend has been toward freshwater systems of a single grade. The freshwater system must provide a continuous supply from the storage tanks to the hot potable water system, the cold potable water outlets throughout the ship, and other freshwater services. Itmust be arranged to receive potable water from shore or from a n ~ t h e ship r and from the distilling plants to fill the storage tanks. Hot potable water must be supplied to all showers, lavatories, sinks, and other services. The arrangement should contain separate mains with branches or independent loops, each serving a washroom or a group of plumbing fixtures. The steam supply for water heaters should be controlled by thermostatically operated valves set to limit the water temperature. Steam valves should close automatically upon failure of the thermostatic element. Each heater should have a relief valve; the discharge
i
699
PIPING SYSTEMS.
DRAIN FROM BOILERS VACUUM SWITCH
osEN Lv.*lL,--
I
MACHINERY SPACE SUCTION B DISCHARGE TO AFT PEAK TANK
TANK
I
DETAIL 'B"
*
TYPICAL SECTION THROUGH BALLAST MANIFOLD
NOTES I. ALL PIPING, VALVES AND FITTINGS ARE MEET THE REQUIREMENTS OF USCG ANDABS. 2. ALL PIPE SIZES SPECIFIED ARE N O ~ W A LWITH IPS OLS~SIDE DIAMETERS: 3. VACUUM SWITCH STARTS ONE PUMP WHEN'THE VACUUM DROPS TO 1 5 " ~ g . STARTSA SECONDPUMP WHEN I T REACHES 12" Hg, AND STOPS THE PUMPS ~ . SWITCH CONTAINSA MANUAL WHEN THE VACUUM REACHES 2 0 " ~ THE OPERATING FEATURE. 4. SPECTACLE FLANGE TO BE POSITIONED m ALLOW ONLY THE BILGE SUCTION TO BE USED WHEN CARRYING DRY CARGO IN THE DEEP TANKS
Wg. 17 Diagram of bilge, clean ballast, and priming syatem in machinery apace
700
All risers exposed to the weather should have means for draining to prevent freezing. WATERCONBUYPIION, Gage glasses should have guards, or be otherwise gallons/personday Potable Wash protected against accidental breakage. Water Water U8E Hose valves and piping for receiving water or Drinking water supplying water overboard should be installed in a Crew, passengers horizontal position about two feet above the deck and Cooking water Cook~ngand food washing in a protected location. Each hose connection should Dishwashing be provided with a cap and chain. A label plate should Washing be installed to indicate the service. Baths Lavatory Locked, verminproof lockers or other suitable Laundry ... 1.0 storage should be provided for the stowage of potable Cleaning in galley and dining area 1.5 ... water hoses; a label plate inscribed "POTABLE WATER 8.0 22.0 HOSE STOWAGE ONLY" should be installed on the lockers. If a hose connection is used to supply potable water from the relief valve should be to a deck drain knd should to a facility or other system, a warning label plate should be provided in a conspicuous location, and inscribed terminate so as not to endanger personnel. Cross-connections between potable and seawater Z E F - S H A L L BE DISCONNECTED WHEN NOT systems, or connections to other systems that could IN US&.'' Fittings for hose connections .should consist of a contaminate the potable water system, are not permitted. stop-check valve and a vacuum breaker, installed in that Valves and blind flanges are not considered a dependable order, upstream to downstream. means of separating potable water from the other Additional requirements regarding freshwater systems systems. Potable water should not be delivered to other systems, tanks, or facilities that could contain are contained in the regulations of the Public Health water of inferior quality unless an air gap of a t least two Service 121; a thorough understanding of the Public supply pipe diameters is provided between the supply Health Service regulations should be obtained before and receiving connections. Examples of facilities which undertaking the design of a freshwater system. 3.15 Air Escapes,Z Overflows, and Sounding must have an air gap are the supply to galley and pantry While air escapes, vents, overflows, Arrangements. sinks, dishwashers and other scullery equipment, and sounding arrangements do not constitute what may vegetable peelers, steam kettles, laundry machines, sterilizers, and plumbing fixtures of all types. Presh be considered a piping system, they are nevertheless of water should be supplied to steam kettles, dough mixer, great importance for the proper operation of the pumping steam tables, and mixing machines via a swing faucet. systems and are subject to special requirements by the The location of faucets above the rims of lavatories, regulatory authorities. Air Escapes. An air escape must be provided for all sinks, and other open receiving vessels constitutes an compartments fitted with filling, suction, flooding, adequate air gap. It may be necessary to provide a sluicing arrangements, or interconnected by cross funnel in the air gap receiving connection. floding ducts unless the compartment is otherwise Distilling plant discharges to both the cold potable water tanlts and the distilled water tanks must do so via open to atmospheric pressure. Additionally, air escapes an interlocked manifold so that flow from the distilling must be installed for all bottled gas stowage compartments. plant can be directed to only one system a t a time. An air escape or overflow should be provided for Design details to be considered when developing a voids or other-unvented spaces which contain pressure freshwater system are as follows: Bolted connections should not be made through the piping. In general, a 1.5-in. line is considered adequate, based on the probability of slow leakage and not comshell of potable water tanks. Pumps, manifolds (filling, transfer, and suction), plete failure of the pressure piping. Compartments having carbon dioxide flooding systems and gages should be grouped. or in which high-pressure air or inert gas are located, Potable water services should be grouped to keep and which may be closed tightly under any condition, the number of branches from the main to a minimum. must also have a vent independent of ventilation ducts, Service branch sizes should be based on the number of unless the test head of the space is in excess of the fixtures installed. Potable water piping should not be run through pressure that can be built up by the discharge of the l.21 flasks when the compartment is secured. ullgaa. Air escapes shouldbe sized in relation to the maximum All branch oonnections to the main freshwater system should be provided with stop valves. A check valve should be installed in the supply to up in water heatem, to prevent hot water from a The terns air escape and vent are used synonymously with regard to tanks or cbmpartments. the cold water piping.
Table 3
A ~ ~ r o x i m aMinimum te Rates of Water Consumption
.
.
70 1
PIPING SYSTEMS
MARINE ENGINEERING
t
filling rate to which the tank may be subjected. In general, sizes may be as follows: Air escapes from tanks, other than fuel oil, having suction and pressure filing connections should be one tenth the area of the suction or filling connections, -but not less than 1.5-in. ips for freshwater tanks and 2 in. for water ballast tanks. Air escapes from fuel-oil tanks should be sized to limit the air velocity through the air escape to 25 fps when the tank is being filled a t its maximum designed rate, but not less than 2.5-in. ips. If two or more air escapes are joined together or to a header, the size of the common air escape or header should be not less than the total area of the air escapes connected thereto and serving a group of tanks which will be filled or drained simultaneously. In addition to the foregoing requirements, all nonstructural tanks containing combustible fluids must have a total venting area not less than that in Table 4, unless equivalent venting is otherwise provided to prevent excessive pressure when the tank is exposed to an external fire. Air escapes should be run with the maximum attainable rising pitch from the tank to the terminal end to facilitate drainage back to the tanks served. Additional air escapes should be installed for tanks or compartments when their shape prevents complete venting by one air escape. If more than one air escape is required, they may be combined and led to an air escape main or to the weather deck. Each branch should be full size. All air escapes should terminate in return bends (i.e., gooseneck bends). Air escapes should terminate in the weather, except those from water tanks, which may terminate within structure open to the atmosphere, and lubrication-oil tank vents which may terminate in the same space in which the tank is located. In order to prevent water from entering an air escape that terminates on deck, a ball check valve may be installed a t the terminus of the air escape. Permanently mounted means should be provided for closing the openings of all vents (see the requirements in reference [I] on this subject). Air escapes from tanks or compartments carrying flammable or toxic volatiles should terminate outside of the ship a t least three-feet clear of air ports, ventilation intakes, or other openings into the ship. Air escapes from reserve feed or potable water tanks should not terminate in unventilated spaces- or spaces assigned primarily for electric equipment. Those from potable water tanks should be fitted with insect screens and should not terminate in toilet or hospital spaces. To prevent compartment flooding under a damage condition, air escapes should terminate a t a height not less than the tightness head of adjacent main transverse bulkheads. Air escapes from tanks carrying different liquids should not be combined. Fuel-Oil Tank Air Escapes. Air escapes should be
Table 4
Suggested Minimum Vent Sizes for Tanks Containing Combustible Fluids
VENT SIZE,in. ips Test Head Test Head 5 psi 10 psi 0.75 1 1.25 1.5 2 2.5 3
*
0.5 0.75 1 1.25 1.5 2 2.5
provided for fuel-oil and ballast tanlts and settling tanks, and should be led to the weather deck. If the check valves in overflow piping are outside the tanks served, air escapes may be connected to overflow piping on the tank side of the check valves, provided the overflow is connected to the highest point on the tank top. Air escapes from fuel-oil tanlts should terminate in return bends, the open ends of which should be enlarged to 1.5 times the diameter of the air escapes and fitted with double screens which should be removable for cleaning. The inner screen should have a clear area through the screen equal to that of the escape pipe. The outer screen should be fitted about 0.5 in. from the inner screen and should have 0.5-in.-square openings. The screens should be bronze, brass, or nickel-copper alloy. Screens should be installed so that they cannot be easily painted over or dainaged. Plumbing System Vents. The protection of plumbing system trap seals from siphonage or back pressure should be accomplished by the appropriate use of vents from the plumbing drains. The vents should be sized and installed to permit the admission or emission of air so that under normal use the seal of any fixture trap is not subjected to a pressure differential of more than one inch of water. The vents should be installed as continuations of the vertical stack with individual fixture drain connections forming branches therefrom. Where a horizontal drain line connects to a vertical line, the vent should be a continuation of the vertical line. All vent and branch vent pipes should be sloped and connected so as to drip back to the soil or waste pipe by gravity. The vhnt pipe opening from a soil or waste pipe, except for water closets and similar fixtures which flush by siphoning of the trap, should not be below the trap weir. All plumbing system vgnts should terminate. in the weather. Those from fixtures above the watertightness level should terminate just below the deck next above that on which the fixture is located. The vents should be welded directly to the structure where they penetrate the shell or deck houseside. Vents terminating above the watertightness level from soil drains should not terminate near hatches, doors, air ports, ventilation intake openings, or galleys.
MARINE ENGINEERING
702
"T" outlets of the same diameter as the vent should be fitted at the outlet ends of the vents. Overflows. Tanks generally are not designed to withstand the shutoff head of the system pumps; therefore, the tanks must be protected by overflows which will prevent the tank pressure from exceeding the design value. An alternative means of protection (by installing a standpipe at the pump discharge) is described under the ballmt system. Overflow pipes discharging through the ship's side should be located as high as practicable and should be provided with nonreturn valves at the shell; where the pipe does not extend above the freeboard deck, there must be provided, in addition, an effective accessible means for preventing water from passing inboard. Reserve feed tanks or other tanks may be fitted with relief valves in the filling system or on the tank itself to protect against pump pressures. Overflow piping should be designed so that the combined static and dynamic head in the overflow line during the most critical overflow condition will not exceed the test head of the tank. The following factors should be considered when designing overflows : Intertank or tank-to-compartment flooding by any combination of damage to piping and tank boundaries should be minimized. Liquids should not be lost when rolling up to 30 deg. Piping runs should be in protected locations, and out of machinery spaces where possible. Piping should be sectionaliaed to minimize fire hazards. A minimum number of shell or oiltight structure penetrations should be required. The overflow overboard discharge should terminate a t least one deck height above the full-load waterline. The size of the overflow from any tank should not be less than the size of the filling ~ i p i n gto the tank and not less than that required to prevent the test head of the tank from being exceeded. A reverse bend should be installed just upstream of the check valve and the valve installed adjacent to the shell. All check valves in overflow lines should have their flap hinge axes athwartship and should not be springloaded. No closure fitting or gagging device should be in any part of the overflow system. All overflows should be installed to ensure gravity drainage of the piping back to the tank or to the overflow tank if one is provided. Overflows should not discharge inside the ship except to an overflow tank provided with an independent overboard discharge. Tanks filled through a funnel need not be fitted with overflow pipes. Overflows and air escapes may be combined where practicabIe. Sounding Arrangements. In order to determine the @
height of liquid in tanks, sounding rods, which directly mearmre the depth of the liquid, are generally provided. If the shape of a tank which requires sounding does not permit dropping a sounding rod in a straight line from the deck, platform, or flat above to the deepest point of the tank, a sounding tube must be provided. Some tanks or compartments which require sounding may be so located that the sounding tube would seriously jsopardize the watertight integrity of the ship or would be unsatisfactory in other respects. In such casei static-head gages or gages of other suitable types may be used. Factors to be considered in the design of sounding arrangements are as follows: Sounding tubes should not have reverse bends. The minimum radius of curvature for sounding tubes should be 10 ft. Sounding tubes, should be located at or as close as practicable to the deepest point of the tank. This is particularly true in connection with fuel-oil tanks so that soundings can be taken at the deepest point of the tank. Sounding tubes for all tanks fitted with suction piping should be located as close as practicable to the terminal of the suction tailpipe. Combining sounding tubes and air escapes from compartments subject to high filling rates, such as fueloil tanks, should be avoided. The sounding tubes from these tanks should be provided with a cap so designed that the accumulated air pressure in the tube will be slowly released and equalized before or as the cap is unscrewed. Protective plates, extra-heavy tees, or similar protection should be fitted at the bottom of sounding tubes to prevent damage to plating and to prevent the sounding device from leaving the tube. Sounding tubes for potable water tanks should terminate with caps. Each tube should have its own sounding rod permanently stowed in the tube. The upper ends of sounding fittings and tubes that are located below the full-load waterline should terminate in gate valves. Where the sounding fitting would otherwise be located in a locked space, a sounding tube extending to an accessible location on the deck above should be provided. The sounding tube should terminate in passageways where practicable. If a soun$mg tube terminates in a compartment, sounding fittings should be readily accessible and should be located so as not to interfere with the function of the compartment. Oil tank sounding tubes should terminate one deck height above the tank top if practicable. Sounding tubes for double-bottom oil tanks and water tanks should extend to a convenient height above the floor plates and in all cases above the highest point of the tank. Upper ends of oil tank sounding tbbes should be located so that oil cannot discharge onto high-pressure
PIPING SYSTEMS
703
steam p i p or e1eMe equipment when the tube ia of the pump being contaminated with oil. This is the pump, as required by the rules, which always should be opened and the tank b being fiBerE. a.16 Fire Pretectlott, A number of meass m y be readily available for fire-fighting purposes. The fire, bilge, and clean ballast pump is used to used for fighting s h i p b r d h.The alternatives include ma,qa&er,w&z fag, carbon dioxide, and foam draw clean seawater or clean ballast and discharge to systems. Eah dtesnative hm limitations (e-g., sea- the clean ballast tanks or to the sea. When serving the water and mtm fog are h d o u s when 4 in the clean ballast system,, there .is no danger of the pump vicinity of electrical equipment, them is danger d being eontaminated with dl. Since there are several personnel asphyxi~dion should c m b n dioxide be other bilge pumps (oily bilge and ballast and the general indvertentlg dimhwged into l i h g qwkrs, and foam service pump) yhich normally are used for bilge service, systems detariomte when subjected ka hihigh tempem- the fire, bilge,' and clean ballast pump is seldom used turn). E ~ m l ~ due er~ t o i b availability and generally exoept for loading and discharging clean ballast. In satisfactary r e d @ seawater has continued to be the remote instances when this pump must be used to draw from the bilges, it normally will pump non-oily bilge. primary rneaeana of fighkhg shipbard fires. The U.& CQ& Guad as well as the elmifieation Therefore, this pump is entirely suitable for fire-fighting societies have apee3c quirementa r e p d h g fire- purposes. Since both d the fire pumps are motor-driven and the extinguishing equipment on ~ h i p b m d d; e t d s of these source of power is the turbogenerators, another power requirementis m oontained in their published d e s and regarE&&m [I, 31. The h-fighting aystem deait with source other than stkam must be provided for one of here is eonfined ta that employing seawater as $he the pumps to satisfy the rules. As a remlt, one of the medium; ifhwin be noted that this system has a number of fire pumps is connected to the emergency diesel generator interfaces 6th @tihershipboard piping ayatems. 0 t h ~as the second source of power. The fire pumps are appmtxls mob as portable extinguishers a d special located in different compartments as required by the eqaipmend required fa^ different type8 d mrviee md d m . (The basis of the rule is to ensure that a fire in a machinery instaI:Iatb~are defined under the generd single:space will not incapacitate all of the fire pumps.) hwdbg of 6re gmfmtion in the W.S.Gaast G u d mlm. As required by the rules, each fire pump discharge is (a) Fhe MirXain System. In order to review the fitted with a pressure gage. dmijp d B fire main sya&m%c ~ m i d wtihe ~ystern The fire main aystem must be designed for a t least the illustrated by Rg. 18. The fire main system shown in shutoff head of the fire pumps. If the piping itself is Fig. 18 ia f ~ an r oeem going dry-carm ship and iEl not so designed, the pump discharge must be fitted with themfore required to be equippad with at Iemt f i w ~fire a relief valve and the system' must be designed for the relief valve setting. pump%. Eaeh of the fire pump muet have a r&*urn There must be provided at least one international eapnity oaf two thirds of the required minimum capacity of an itrdqendent Bilge pump. Each fire pump muat shore connection to the fire main available to each side t a b be cd ~ui%ient eapaclty to deliver water simul- of the vessel and it must be located in an accessible tmmdy hhmagh the two highmt and most remete location. Suitable cut out valves and check valves 2% -is. hydrants connected w%h W ft of h w . Each should be provided. Suitable adapters should also be h e mzde af in. d i m mu& have a t least 60 psi a t provided to ensure that the vessel's shore connection8 will mate with those on the shore fin lines. the ~ d e .(The, capacity mquhd far 50 psi 3 inlet af tm %-in. nozkles i~ generally 1e6s than the AU paste of the fire main located on exposed decks reqi~emenf,d h m %hi& of the ezapacity of one bilge &ou1d either be protected against freezing or be fitted cut oqG valves and drain valves so that the entire ~ump.1 When the fire main i~ use& for washing down the e-d parts of this piping may be shut off and drained mdor &@in, &shing the & a h Eaekar and sirniflar in freeaing weather. Except when closed to prevent d i c e , tihe fim grrmp eclpeitiy mmt be i n m d t o take fmeakg, these valves should be sealed open. oage d this w~~ on the laasia that &he flushing;s e d e ~ ESre Byrkants and Hose. Fire hydrants should mdd 14e in WE &nIrltmnwu~lpwhen water k required be of ~uffioientnumber and so located that any part of for fire exfhqpishing~ U ~ ~ Q S Mowwer, ~ E A the fire pumpa living qumfiers, storerooms, working spaces, and weather an be connected to the deck wash q~ternwithauk de&e acsessible to the crew while a t sea may be reached i n e d n g the mquhd pump m@ty. with twa effective spray patterns of water. At least one As sbawn ie. Fig. 1% the two fire pump ma also usad in other s@ema m their name& imply. One is called spray pttem should b e ' h m a siligle 50-ft length of the fire md eaiEtk c J d n g p m p md &B ~ t h m is referred how. In main maehhery spaces, dl portions of such a~ the frre, bilge* and clean b d a t pump. Neither spaces ~houldbe capable of being reached by a t lemt pump k pemi8ted ta have meotiom ta lie^ or tmks tm effeaive apmy patterns of water, each of which should be fram a single S f t length of hose from separate m~t4ningoil. The &e. and amk d e e n g pump ia us& to wmh down outlet% Ebb fire station h y d t a ~&ouId t be provided with at @heear@ t d w . Qelhn swamter is heah&far bmk Ie& one len& of hose with a noade atkached. A hose e purposes; mnsequentkl~k,there, b na h w
PIPING SYSTEMS
MARINE ENGlNEERlNG
/SEA
ChEST
wvENT
RESERVOIR TANK
INTERCEPTOR
FLOOR LVL IN PUMP RM
SKETCH "A"~ TYPICAL DIAGRAMMATIC ARR OF STRIPPING SYSTEM FOR CARGO OIL PUMPS
BILGE VERTICAL B STRIPP DUPLEX PUMP
PLAN VIEW -PUMP ROOM Fig. 21
Dbgrcm d carga-oil piping in p m p room
Since the operation of large pumps with steam conditions as low as 500 F would not be economical, the turbines are confined to the engine room. Figure 19 illustrates the cargo-oil piping on deck. The discharge from each cargo pump is led to the upper deck through individual lines to the discharge station amidships. Each discharge line is fitted with twin hose connections on the port and starboard sides of the ship for discharging to shore facilities. The loading of cargo oil is generally accomplished through filling drop lines connecting to the lines within the tanks; however, filling may be accomplished by cross connections between the deck piping and suction piping in the pump room. The filling drop lines and their connections are shown in Figs. 19 and 20. The piping main (the fore-and-aft piping extending from the pump room forward) is sized for a low velocity (8-9 fps) so as to avoid excessive pressure drops in the
suction piping. The mains are generally sized larger than the branches to permit simultaneous pumping from either one large tank or two small tanks without restricting the oil flow. For the purpose of flexibility in handling the cargo oil in the tanks, the mains are usually capable of being cross-connected in such a manner that any pump can be connected to any tank in any group. One of the more difficult aspects of the design of a cargo-oil system is the stripping (removal of the last remnants of oil) of the tanks. When the oil level in a tank is pumped down, air's enter the suction piping if the pipe suctions are not completely covered with oil; however, it is not necessary for the suction pipe inlet to become uncovered in order for gaa to enter the suction piping. A major source of gaa in the suction piping is oil vapor produced by the lighter ends of the cargo-oil gasing-off. When the level of the cargo oil in the tank is low, but still covers the suction inlet by several feet,
MARINE EN
i
I
~
gasing-off of the cargo oil in the suction pipe can be a serious problem. When the air or oil vapors subsequently pass through the suction piping and enter the eye of the pump impeller, the cargo-oil pump will lose suction, operate erratically, and possibly be damaged due to excessive speed surges. The high viscosity of the oil is one factor which impedes the uniform flow of the oil into the pipe suction. Additionally, the flow restrictions presented by the necessary strength members in the bottom of the cargo-oil tanks further complicate the stripping of the oil tanks. An important feature not shown in detail in Fig. 20 is the design of the suction fittings at the end of the tailpipes in each tank. To be effective in emptying the tank completely, the suction fitting must be located within to in. of the bottom plating and have a flow area of approximately 1.5 times the cross-sectional area of the attached tailpipe. The different types of suction fittings are described in detail in reference [9]. Several different schemes have been used for stripping oil tanks. One entails the operation of a main cargo-oil pump at a reduced power level such that the gas will not be as damaging when it passes through the pumps. An alternative arrangement is one in which stripping is accomplished by independent reciprocating pumps. Yet a different stripping scheme involves the use of vacuum pumps to remove the air and oil vapor from the main cargo-oil pump suction piping. This scheme is described in references [9, 101; it essentially consists of a separate tank (from which gas is removed by means of a vacuum pump) that is installed in the suction piping just upstream of the main cargoail pump. A diagrammatic sketch of a typical arrangement for removing the air from the suction piping of cargo-oil pumps is shown in Fig. 21. A diagram of the piping arrangement in the pump room is also shown in Fig. 21 and it may be noted that a steam reciprocating pump is provided for bilge and stripping duty. A reciprocating pump is preferred for bilge service due to the fact that maintenance of a suction head is not essential; contrary to the case with centrifugal pumps, reciprocating pumps can pump gasliquid mixtures without difficulty. Additionally, the reciprocating pump is well-suited for stripping the cargo oil left in the tanks and stripping the oil from the main cargo pump and piping. Strippin pumps are sometimes arranged to prime the main argo-oil pumps. A relief valve should be installed in the discharge of each pump and piped back to the suction (except that, when the pumps are designed such that it is impossible to build up a pressure above that for which the system as a whole is constructed, relief valves are not necessary). It is highly desirable to maintain the system pressure below 225 psig (Class 2 piping systems as defined by references [I] and [3]; otherwise, the system must conform to the more rigorous requirements of Class 1 piping systems). As an alternative to the cargo-oil system shown in Figs. 19, 20, and 21, some installations make use of the
/
PIPING SYSTEMS
deep-well type of cargoail pump. This eliminates, in some cases, the cargo-oil pump room and the engine room space for the prime movers. Deep-well pumps (pumps with the impeller located down in the cargo tank and driven through a long shaft by a prime mover located on deck in an enclosure; see Chapter 12) were primarily developed for small special cargo tanks but their use has been expanded. The use of deep-well pumps can reduce the need for the long runs of piping required with the pump rooms located aft. More significant than the reduction of piping is the elimination of the associated suction pressure losses, particularly with the larger units. An important aspect of deep-well pumps is their self-priming feature; when the pump discharge pressure drops (due to gases passing through the impellers), the liquid discharged from the impellers is recycled back to the pump suction. There are, however, practical limits in the application of deep-well pumps. Although size is theoretically not a limit, due to the size of the vertical driveshaft and the prime mover with the higher ratings and the thrust developed, deep-well pump installations are usually limited to about 1000 hp. See references [9, 101 for additional discussion regarding deep-well pumps. All cargo-oil tanks must be cleaned periodically to remove the residue that accumulates on the tanks. The medium used to clean the tanks is seawater that is pressurized to about 200 psig and is either used cold or heated to a temperature of 180 to 200 F (see Section 3.4 of Chapter 14 for a discussion of the heat exchangers employed). Either fixed or portable tank cleaning (Butterworth) systems may be used. With a portable system the spray nozzles are lowered into the tanks a t intervals of about 5 f t by means of their attached hose. The nozzle mechanisms are constructed such that the nozzles slowly revolve around the central gears, thus directing a stream of hot seawater over all parts of the tank. With a fixed tank cleaning system the nozzle mechanisms are permanently mounted in the tanks, with the valves which control the flow of water located on deck. Neither the portable nor fixed system cleans the tanks thoroughly; in order to remove the oil residue from the regions that are not properly covered by the spray patterns, cleaning with a manual nozzle is required. During the tank cleaning operation, the stripping pumps take suction from the tanks and discharge to a slop tank or to shore. This oily mixture is handled in a manner similar to the oily ballast dishssed earlier. All cargo-oil tanks must be vented, and for this purpose each tank should be fitted with a pressure and vacuum relief valve. Alternatively a vent pipe may be led from each tank into a common header, which should be carried to ten or more feet above deck and be fitted with a flame arrester or a vacuum relief valve a t the outlet to the atmosphere. Where vessels are intended only for the carrying of oil having a flash point above 80 F, a venting system consisting of individual return-bend vents fitted with wire gauze and a positive means of closing is satisfactory.
4 "Power Piping," USA Standard Code for Pressure Piping, USAS B31.1, published by ASME. 5 L. F. Moody, "Friction Factors for Pipe Flow," Trans. ASME, 1944. 6 "Flow of Fluids Through Valves, Fittings, and Pipe," Engineering Research Division, Crane Co., b Chicago. 7 "Wrought Steel Buttwelding Fittings," USA Standard Code USAS B16.9, published by ASME, 1964. References 8 E. F. ~ e & n sand J. R. Reilly, "Condenser Scoop - - - - -- - - - - - Design," Trads. SNAME, vol. 48, 1940. 1 Code of Federal Regulations, Title 46-Shipping, 9 A. W. Feck and J. 0. Sommerhalder, "Cargo published by the Office of the Federal Register. 2 "Handbook on Sanitation of Vessel Construction, " Pumping in Modern Tankers and Bulk Carriers, " , 4, no- 3, July 1967. U.S. Department of Health, Education, and Welfare, Marine T e c h ~ h Ivol10 W. M. Morrison, "Control of Cargo Oil HanPublic Health Service. 3 Rules for Building and Classing Steel Vessels, dling," Automation in Merchant Ships, published by Fishing News (Books) Ltd., London, England. American Bureau of Shipping.
The control of the cargo-oil valves within the tanks and in the pump room can be accomplished by various means. Cargo-oil valves on the older ships were manually operated from deck stands with reach rods to the valves. The trend is to fit hydraulic operators to the valves and to control the valves from either a location on deck or from a central control room. Section 2.6 of Chapter 21 contains additional discussion regarding the automation of cargo-oil piping systems.
'
i
709
ENVIRONMENTAL CONTROL
11
CHAPTER XIX
Marked
I En~ironmentalControl
Section 1 Introduction 1.1 Merchant vs. Naval Practice. The terms air conditioning, refrigeration (cooling), ventilating, and heating cany dierent connotations as applied to merchant and naval practice. Accordingly, effort is made in the following discussion to differentiate between the two; but space limitations preclude full treatment of the latter. In general, the ventilating, heating, and cooling facilities on naval construction are the minimum required to accomplish the following: 1 Keep the crew physically fit and mentally alert. 2 Provide ambient temperature conditions suitable for the operation of equipment. 3 Provide suitable ambient conditions for certain ammunition. Preserve perishable products (stores, and
the or explosive atmosphere.
of a contaminated' noxiouq
The weight added, the space occupied, and the power consumed by these facilities are generally undesired; therefore, the minimum of equipment is provided to maintain acceptable conditions. The ability of equipment to withstand shock and the maintenance of the ship's watertight integrity are vital design requirements. Compared to naval construction, merchant ship ventilating, air conditioning, and cooling treatment are, generally speaking, more liberal. Almost all spaces which are not air conditioned are ventilated. I n addition to comfort, heating is provided to prevent the freezing of stores, cargo, and equipment and to facilitate operation of equipment. I n some cases dehumidification facilities are provided for dry cargo preservation. Refrigerated stores are more segregated on merchant ships than naval ships. Also, in many cases, refrigerated cargo compartments can be designed for specific products rather than a "universalncargo. Additionally, merchant ship air-conditioning practice is less standardized. Space and weight considerations are far less restrictive; therefore, accessibility for maintenance and repair operations are much better. Furthermore, many refinements can be provided which are not feasible for naval construction. Summarily, naval ship facilities require more precise application of the various techniques involved to provide
tolerable conditions with minimal equipment, space usage, and energy consumption. I n comparison, merchant construction is less standardized and allows greater latitude for innovation. 1.2 Air Conditioning Deflned. The various techniques utiliied to achieve control of the environment of shipboard spaces are collectively known as air conditioning. By definition, air conditioning is the process of treating air so as to control simultaneously its temperature, hGmidity, cleanliness, and distribution to meet the requirements of the conditioned space. Contrary to widespread belief, odors rather than carbon dioxide concentrations and depletion of oxygen are the most significant cause of discomfort in poorly ventilated spaces. The term replenishment air is used in Navy pratice to denote the air supplied to air-conditioned compartments to maintain an aiceptable atmosphere. I n merchant practice the terms minimum outside air or minimum .freshair have the s a m d a n i n g . A ventilation system serves a dual purpose of removing contaminants as well as heat generated in a space. Its ability to control temperature is limited, however, in that it can only maintain an ambient temperature above that of the outside air temperature. The temperature rise, i.e., the difference between the temperature of the supply air and the space ambient, depends on the ventilation (heat removal) rate. For this reason, temperature rise is one of several criteria for ventilation design. To maintain a space at an ambient temperature below that outside, the supply air temperature must be below the required ambient; thus, some form of refrigeration is employed to cool the supply air. Humidification is the process of adding moisture to a space to prevent excessive dryness, which may cause either discomfort to occupants or deterioration of goods. Dehumidification is the process of removing moisture from a space to achieve one of several objectives. Dehumidification is an integral part of the air-conditioning cooling process; it maintains the relative humidity of the space served within the "comfort" range. In cargo spaces, dehumidification is primarily employed to prevent deterioration of stowed goods. The dehumidification inherent with the cooling of refrigerated compartments, however, is disadvantageous as it causes surface %
deterioration (drying, wilting, etc.) and increases the refrigeration load. Air motion serves to produce a uniform ambient temperature (temperature equalization) throughout an airconditioned space. However, excessive air velocity .has a localized cooling effect and, therefore, every effort must be made to achieve uniform, low-velocity air motion in the occupied zone; i.e., to avoid drafts. Consequently, proper air motion is an important characteristic of an air-conditioning installation. Effective temperature (ET) is an arbitrary index which combines into a single value the resultant physiological effect of temperature, humidity, and air movement. It is obtained from tables or a chart on the basis of dry bulb and wet bulb temperatures and air velocity [1].' Tests have shown that personnel can stand 93 ET for a regular 4-hour watch and still work efficiently without ill effects. Where heavy work must be performed, such as handling rooms and magazines, the effective temperature should not exceed 85. 1.3 Preservation of Perishable Products. The preservation of perishables is commonly associated with refrigeration, which, in the true sense, only involves the extraction of heat. However, all of the processes associated with air conditioning are required to preserve perishable products. All fresh fruits and vegetables, including flowers, are living things and are commonly referred to as "live" products. Even when separated from the tree, vine, or soil they continue to generate and dissipate heat as does the human body. The amount of heat respired (called heat of respiration) varies with the product and its temperature. I n the respiration process, live products absorb oxygen and give off carbon dioxide and water. Respiration and ripening can be controlled by reducing the oxygen supply and increasing the carbon dioxide in the surrounding atmosphere. In some cases carbon dioxide is injected into refrigerated spaces, but not on U. S. ships. Unripened fruits also give off other gases which must be controlled. For instance, during ripening bananas release ethylene and volatile esters. Ethylene accelerates ripening, during which starch in the pulp is converted to sugar. The color of the peel also changes during this process, which is of advantage because it provides a visual indication of the fruit's ripeness. Ventilation is required to control the concentration of the released gases. Refrigeration (cooling) serves to maintain the individual product at the temperature at which it must be carried for proper preservation. Heating is required to prevent freezing of many products, such as fruits and vegetables. libr instance, bananas should not be exposed to a temperature below about 55 degrees for an appreciable length of time. High humidity is required to prevent wilting of leafy Numbers in brackets designate References at end of chapter.
vegetables, loss of moisture, and associated shriveling and loss of flavor. Whiie low relative humidities (40 to 50 percent) are conducive to human comfort, humidities of between 80 and 95 percent, as high as possible in many cases, are required in refrigerated spaces carrying fruits and vegetables. Good air distributi~nis essential to the preservation of perishables. It providesbiform air conditions (temperature, humidity, and air purity) throughout the storage space.,,,'Thus, localized spoilage is prevented. Air distribution is not as critical for frozen products, referred to as "dead" products, as it is for most unfrozen (live) products. 1.4 Psychrometry. Psychrometry is the branch of physics which deals with the properties of air under various conditions, particularly air containing moisture. Psychrometric data are presented in both charts and tabular form; Fig. 1 is a popular form. The terms used in Fig. 1 are defined as follows:
Dry Bulb Temperature. The temperature of air as registered by an ordinary thermometer. Wet Bulb Temperature. The temperature registered by a thermometer whose bulb is covered by a wetted wick and exposed to a current of rapidly moving air. Dew-Point Temperature. The temperature a t which condensation of moisture begins when the air is cooled. Relative Humidity. Ratio of the actual water vapor pressure of the air to the saturated water vapor pressure of the air at the same temperature. Specific Humidity or Moistwe Content. The weight of water vapor in grains or pounds of moisture per pound of dry air. Enthalpy. A thermal property indicating the quantity of heat in the air above an arbitrary datum, in Btu per pound of dry air. Enthalpy Deviation. Enthalpy correction due to the air not being in the saturated state. Enthalpy deviation is applied where extreme accuracy L required; however, for normal air-conditioning estimates it is neglected. Specific Volume. The cubic feet of the mixture (air and moisture) per pound of dry air. Sensible Heat Factor. The ratio of sensible to total heat. Alignment Circle. Located at 80-F DB and 50 percent RH on Fig. 1, it is used in conjunction with the sensible heat factor to plot the various air-conditioning process lines (room and coil slopes). Note that all units are expressed in terms of "per pound of dry air"; the reason for this is that the various properties of the moist air chaqge during an air-conditioning cycle and only the weight of dry air remains constant. It may be seen that if any two psychrometric properties are known, the remaining can be determined. The following relationships also can be observed from Fig. 1: 1 The enthalpy (or total heat content) is determined from the wet bulb temperature alone, and vice versa. 2 The dew point depends only on the moisture content of the air, and vice versa.
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
3 Any air-conditioning process which doea not add or extract moisture is represented by a horizontal line through the original condition point. 4 Any air-conditioning process which does not change the dry bulb temperature is represented by a vertical line through the original condition point. Figure 1 also shows a simple air-conditioning cycle. Points 1, 2, 3, and 4 represent psychrometric conditions of room, outside, mixture of outside and recirculated air, and air leaving cooling coil, respectively. Points 1 and 2 are established by design conditions while point 3 is determined as follows : Ha = (Ql/dl)Hl Q1/d1 where
(Qn/da)Hz ++ QJdz
Q = air quantity, cfm d = air density, cu f t per pound of dry air H = enthalpy, Btu per pound of dry air Subscripts apply to the various points. A much simpler way of obtaining the conditions of the mixture at point 3 is commonly used, except where unusual accuracy is required. This is represented by the equation: Distance 1-3 = Qs Distance 2-3 Q1
(1)
When the dry bulb temperatures are known, this may be expressed, and ts obtained, as follows:
The total heat, HT, removed by a cooling coil consists of the sum of the sensible heat load, Hs, and the latent heat load, HL. The sensible heat load is the transfer of heat which results in a change of temperature as read by an ordinary dry bulb thermometer. The latent heat load is the heat necessary to cause a chaqge of state, e.g., steam to water. These quantities may be computed as follows: Hs = 1.08 QAt (3) HL = 0.68 QAG (4) where Q = air volume entering coil At = dry bulb temperature reduction, deg F AG = moisture removal, grain moisture/pound dry air H T can also be expressed as: HT = 4.45 QAH
(5)
where AH
=
enthalpy reduction of air passing through coil, Btu/pound dry air
Figure 1 shows two sensible heat factor slope lines. Line 1-4 is the "room slope" and line 3-4 represents the "coil slope." By definition, the sensible heat fact0r.i~:
SHF =
Hs Hs
+ HL
Hs or HT
I n order to obtain the proper (simultaneous) balance between room sensible and latent heat removal, the air supplied must be so conditioned that it falls on the room slope line. Any point on this)ine is satisfactory. While the coil slope 3 4 is represented as a straight line, the actual path contour depends on several variables, one of which is the coi1,ebnstruction;in any case, the only significant point is the psychrometric conditions of the air actually leaving the coil, i.e., off-coil conditions. Design off-coil conditions below those established by the room slope frequently are arbitrarily selected for systems with high sensible heat factors. This is done in order to utilize larger terminal temperature differentials [At, equation (3)]; thus, less air is required, which in turn results in smaller systems. Where a system serves several spaces, it may be neither possible nor practical to provide conditioned air to satisfy the lowest individual room slope. I n such cases an arbitrary or average off-coil condition is selected for the system. Spaces having a lower sensible heat factor, theoretically, will be overcooled if the air quantity is selected to meet the latent heat load. Therefore, the supply air to such spaces may require reheating to compensate for the excess removal of sensible heat. In order to obtain an optimum design and avoid excessive air quantities, some marine specifications allow a degree of latitude in the room design conditions of spaces with high latent loads. By slightly reducing the room design dry bulb temperature and increasing the wet bulb temperature (and relative humidity), the sensible heat of occupants is increased and the latent heat is reduced (see Table 8); thus the sensible heat factor is increased and the necessity of reheating may be avoided on occasions. 1.5 The Refrigeration Process. The compression cycle of refrigeration, Fig. 2, is the most common one used today. In this type of cycle, a refrigerant vapor is compressed and, in the compression process, gains energy corresponding to the work of compression. The hot compressed vapor next is cooled by any convenient, inexpensive, plentiful medium, such as water or atmospheric air. This cooling process condenses the hot vapor to # a liquid. The high pressure liquid next is expanded to a lower pressure and becomes a cold mixture of liquid and vapor. This refrigerant mixture is fed into a heat exchanger (evaporator) where it absorbs heat and changes back to a vapor, the same state as the beginning of compression cycle. The principal components of a compression refrigeration system, Fig. 2, are a receiver for storing liquid refrigerant, an expansion valve for controlling the flow of refrigerant (liquid and vapor), an evaporator where the required useful refrigeration is produced, a compressor which simultaneously mairitains the evaporator (suction) pressure and increases the refrigerant vapor temperature and pressure, and a condenser which coola '4-
715
ENVIRONMENTAL CONTROL
MARINE ENGINEERING WATER REGULATING
T VALVE
Fig.
'
RECEIVER LlNE L
m -l EXPANSION VALVE
-- - - - - - - - s - + - s - J
S
A
Actual compression cyde [Allied Chemical]
-
Y EFFECT OF SUPERHEATING SUCTION VAPOR I
I
-
M EVAPORATOR
OlAGRAnWATlC REFRIGERATION CYCLE REFRIGERANT STATE A N 0 ENERGY CYCLE
KEY
- LOW PRESSURE LlOUlD - LOW PRESSURE VAPbR - HlGH PRESSURE VAPOR (HOT GAS) - HlGH PRESSURE LlOUlO Fig. 2
KEY
- L -+- LlOUlD LlNE - S -b - SUCTLON LlNE
-
HG -b
-
-
HOT GAS LlNE
W
+- CONDENSER WATER LlNE
C
--r)
R -b
- CAPILLARY TUBE - REGULATING VALVE TUBE
Graphic presentation of a mechanical refrigeration cycle
(removes both heat of compression and refrigeration load) and condenses the hot refrigerant vapor to its original liquid state. The refrigerant properties of importance in refrigeration thermodynamics are temperature, pressure, volume, enthalpy, and entropy. Reference [I] presents these properties in both tabular and graphic form. Pressureenthalpy diagrams are one form of presentation and comprise three zones as may be seen from Fig. 3. Zone 1, on the left, represents subcooled liquid; Zone 2, the middle zone, represents a liquid-vapor mixture; and Zone 3, on the right, represents superheated vapor. Figure 3 also shows how the actual cycle departs from the theoretically ideal Carnot cycle of Refrigerant 12 in several respects. Due to these departures, published performance data (refrigeration capacity, refrigerant flow rates, horsepower requirements, condensing medium requirements, and efficiencies) are derived from actual code tests and empirical data. Superheating has certain thermodynamic advantages and disadvantages as may be seen from Fig. 3. In general it improves compressor and system performance and is desirable. Some subcooling, about 5 deg F, of the liquid refrigerant occurs in the condenser which increases its refrigeration effect; i.e., it reduces the quantity of refrigerant circulated. Where cooling is provided by a liquid-suction heat interchanger (Pigs. 18 and 19)) the associated superheating of refrigerant vapor also serves a useful purpose.
In addition to the single-stage compression cycle just described, there are many more efficient-but also more complicated-ones [2]. Such systems usually are justified in marine work only for very-low-temperature applications, such as liquefied gas transport and fish processing, where single-stage compression would involve excessive compression ratios. Another reason for using compound compression, though not particularly applicable in shipboard applications, is to increase thermal efficiency. Cascade systems have been installed on a few ships. With this arrangement, two completely independent refrigeration cycles are used; the refrigerant of the high-stage cycle is the condenser cooling medium of the low-stage cycle. By uaing a refrigerant in each cycle which is most suitable to its operating range, very .efficient performance is obtained. Reference [3] contains an excellent discussion of this subject and a comparison of reciprocating versus centrifugal compressors for cargo refrigeration applications. Absorption refrigeration systems have been used a t sea only for certain special applications. The only fundamental difference between an absorption cycle and a compression cycle is the substitution of the absorbergenerator compressor (operated by heat energy) for the mechanical compressor (operated by work energy). (See reference [I] for a complete discussion of absorption cycles.) Thermoelectric cooling has many advantages which are particularly significant with regard to marine applica-
tions. However, thermoelectric cooling is normally not competitive with other forms of cooling and is used only for special applications; e.g., temperature control of electronic components and scientific instruments. 1.6 Refrigerants and Brines. A refrigerant is commonly defined as a substance which produces a refrigerating effect by its absorption of heat while expanding or vaporizing. Additionally, a secondary refrigerant is defined as the volatile refrigerant used in a secondary system, or a nonvolatile liquid refrigerant such as brine. A brine is defined as any liquid cooled by the refrigerant and used for transmission of heat without a change in its stage, having no flash point or a flash point above 150 F. From the foregoing it is evident that water may be clmified as either a brine or a secondary refrigerant. Refrigerant characteristics have a bearing on system design, application, and operation. Theoretically, a refrigerant should be selected for a specific application so as to match the system performance requirements. However, safety, standardization, availability, and other nontechnical factors enter into this selection. The selection of refrigerants for various shipboard applications has become more or less standardized; all are halogen compounds. Refrigerant 41 is used with centrifugal systems for air conditioning and large hightemperature cargo refrigeration installations. Refrigerant 12 is used with reciprocating systems and large centrifugal systems for medium and low-temperature refrigerated cargo installations. Refrigerant 22 is used with small reciprocating factory assembled units for air conditioning, liquid chilling, and refrigeration; this includes water coolers, ice cubers, pantry refrigerators, and freezers. In Europe R-22 is used for certain cargo refrigeration applications. Future developments in this country will undoubtedly include the gradual replacement of R-12 by R-22 for high-temperature.systems and R-502 for low-temperature systems. Water solutions of calcium and sodium chloride are common refrigeration brines; properties of these brines are presented in references [I, 41. Ordinary salt (sodium chloride) is used where contact with calcium chloride could not be tolerated, as for example the brine fog method of freezing fish and other foods. Nevertheless,
the lower freezing point of a calcium chloride solution makes its use advantageous. The presence of magnesium salts in either brine is undesirable, as they tend to form sludge. Neutral brines are comparatively noncorrosive. However, they are subject to contamination from various sources which influence the brine's corrosive effects. Since air (oxygen) and carbon dioxide are among such contaminants, excessive aeration of brine is avoided and closed systems are used whenever possible. Ethylene glycol and propylene glycol brines, inhibited for corrosion control, have been widely used. Ethylene and propylene glycol are colorless, practically odorless liquids, and completely miscible with water. Inhibited ethylene glycol solutions are normally preferred over propylene glycol solutions because of their more desirable physical properties, especially at lower temperatures. However, in applications involving direct contact with food or beverages, toxicity considerations may require the use of propylene glycol. On an equivalent percentby-weight basis, propylene glycol solutions generally have slightly higher values of specific heat than ethylene glycol, though solutions of either are not as efficient as water alone for heat-transfer applications. Information with regard to glycols may be obtained from reference [5]. 1.I Thermal Insulation. Thermal insulation is provided on a ship's structure and equipment to limit heat transfer (gain or loss) and to prevent condensation. Condensation is generally undesirable in that it can damage the insulation itself as well as other materials it contacts. In addition, condensation can cause damage indirectly by creating high-humidity ambients which in turn can cause deterioration of structures and materials. Effective and efficient insulation treatment is particularly important to economical ship design for the following reasons : 1 Most materials use$-in ship construction have very high thermal conductivity. 2 Structural conformations cause discontinuity of insulation applications. 3 The "fin effect" of beams and other stiffening membem appreciably increases heat transfer through structures. 4 There are many heat-producing (hot) spat- which
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
A
must be thermally isolated from cooler surrounding Because of the weight factor, insulation treatment on naval ships is somewhat less than it is on merchant ships. This applies primarily to the basic thickness of the insulation used for comparable applications. Another difference is that on merchant ships, ceilings and linings are fitted in "hotel" areas to conceal structure as well as insulation, piping, ducts, and other services. Therefore, the insulation can be less durable with respect to service abuse, and does not require a surface finish for protection or decorative reasons. Also, as shown in the discussion of insulation theory, the resulting dead-air spaces have an insulating effect.Another significant difference is that on merchant &ips, metal sheathing is fitted to protect insulation which is not concealed by linings and ceilings (joiner work). The high conductivity of this sheathing appreciably reduces the effectiveness of the insulation treatment unleis adequate means are provided to thermally isolate it from the supporting ship
a = thermal conductance of air space, Btu/hr-sq ft-deg F
in the foregoing equations to obtain:
Values for f,, fo, and C are obtained from tests of typical constructions using the "guarded hot box" method. fi and j o vary with surface temperature and direction of heat flow (up, horizontal, or down); also, fi varies with the frame-spacing dimension. Conductance values, C, for a particular material vary according to the mean of H = UAAt ('I) the surface temperatures, (ti f to)/2. For estimating purposes, an outside film coefficient, fo, of 7.0 is commonly of heat transmission, Btu/hr- used in deriving air-to-air U factors; the l/f component is omitted for sides exposed to the sun or liquids, i.e.,
Of the three variables having a n effect on heat transmission, only the overall coefficient of heat transmission
other highly conductive material extending wholly or partly through insulation should, if possible, be determined by test; but approximations can be made as described in references [I, 61.
This is facilitated by using the expression temperature of the air. M~isturecreates a partial pressure in air which is a function of the air temperature; consequently, if the tem~eratures*ofthe two parallel surfaces of a porous or fibrous material are different, a moisture vapor pressure gradient will exist between the two surfaces. Unless a suitable barrier is provided on the warm side, the difference in vapor pressure will force 1 x 1 moisture from the warm air into the colder material. R =-+'+-+3+fi ki a ko f o Even a small increase in moisture content appreciably increases the conductivity (k value) of insulating materials. Therefore, moisture in the insulation and condensation accumulating on surfaces below the ambient f = film or surface conductance, Btu/hr-sq ftdeg F dew point both negate the effectiveness of the insulation, x = material thickness, in. k = material thermal conductivity, Btu-in/hr-sq ft-- even if the moisture does not actually cause deterioration of the insulation.
where R is the total resistance to heat flow; it is equal to the sum of the various individual resistance components. For example, the thermal resistance of a structure con&ting of two homogeneous materials (denoted by i for inside and 0 for outside or weather) having an air space . between them can be expressed as:
OECK
-
-
E ~ P O S ~ ~
CONSTRUCTION
-
== -OR(
11
XXXXX
Heat transfer occurs whenever there is a difference in temperature between two adjacent bodies. It may be accomplished by any one, or a combination of, three modes: conduction, convection, and radiation. The heat, H, transmitted through a flat structure, such as a deck or bulkhead, can be expressed as follows:
= overall
SURFACE TO AIn
The thermal conductance of an air space is a function of the height, depth, position, character, and temperature of the bounding surfaces. The relationships are not linear, and accurate must be determined test. a structure consisting a single homogeneous matedal) the expression reduces to
or In the case of c~nStructiOnshaving irregular sections, such as are common akoard ship, the resistance of the section, 1/C, is substituted for
717
= =
=
.
MARINATE SHEATHING OR LINING SHEET METAL SHEATHING OR LINING WOOD DECK 12.5 INCHES THICK)
ANTI-SWEAT COHPOUNO
UNLESS OTHERWISE NOTED, INSULATION THICKNESS I S 2 ~ ~ C ~ ON I E PLATES S AND I INCH ON BEAMS AND STIFFENERS. THERMAL BREAKS OF NON-METALLIC FURRING AND BLOCKS ARE CONSIDERED TO BE PROVIDED TO PREVENT METAL-TO-METAL CONTACT BETWEEN METAL SHEATHING AND SUPPORTING STRUCTURE.
INSTRUCTIONS
2.
USE "SURFACE TO A;IR"FACTORS FOR CALCULATING SOLAR LOADS AND TRANSMISSION FROMSURFACES D P O S E D TO LIQUIDS ON ONE SIDE. USE "AIR TO AIR" FACTORS FOR ALL OTHER SURFACES. TREAT PROMENADES AND SIMILAR SPACES, WITH OR
3-
WITHOUT WINDOWS, THE SAME AS WEATHER AREAS. FOR CARPETED SURFACES, REDUCE THE u FACTORS
1.
BY 20 PERCENT; NEGLECT THE EFFECT OF DECK SUCH AS, TILE, MAGNESITE, LINOLEUM, AND TERRAZZO. 4. FOR HEAT ABSORBING AND DOUBLE GLASS, REFER TO MANUFACTURER'S CATALOGS FOR SUITABLE u FACTORS
USE0 PROJECTED AREA FOR BUD SOLAR ONLY
OECI
-
INSIDE
718
MARINE ENGINEERING
The term permeance, expressed in perms, is used to classify a material's resistance to the passage of water vaDor. Metals have zero permeance and, therefore. are pe;fect barriers. closed-cell, nonhygroscopic mateAals, such as cellular glass, also have zero permeance. The permeance rating to be specified for a surface depends on the severity of the particular application. For instance, the vapor barrier for the insulation on an air-conditioning duct need not have as low permeance as that on the hot side of a refrigerated compartment boundary. The joints invariably are the "weakest link" in vapor barrier treatment. Even when a barrier has been perfectly applied, expansion and contraction in time' open joints and thus reduce the effectivenessof the installation. It is a matter of record that most applications of insulation to refrigerated spaces deteriorate with time because of water vapor absorption. Accordingly, it is traditional to design such applications conservatively (81. 1.9 -Sound Conditioning. The noise and vibration generated by refrigeration, ventilation, and other airconditioning equipment are only part of that generated within the ship. Obviously for efficient and consistent treatment, all such disturbances must be considered simultaneously ; otherwise overcorrection of one source of annoyance will only serve to emphasize others. As a result of this necessity for coordinated analysis, noise and vibration problems are best handled by an acoustic
ENVIRONMENTAL CONTROL
specialist. Accordingly, no effort is made herein to treat this subject aside from stating the following -general -guidelines:
RECIRCULATION-EXHAUST FROM SPACE
e
1 Locate noise-generating equipment as far as practicable from spaces where quietness is essential. Passageways, or spaces where quiet operation is not as essential, can be used as buffers between noisy and quiet
EXHAUST
e-
-- _ tipaces.
2 Do not mount noise-generating equipment on bulkheads bounding quiet spaces. 3 Use rotating rather than reciprocating equipment where quiet operation is of primary importance. 4 Secure foundations for equipment on rigid structural members, such as deck beams; i.e., do not support equipment on flexible plating. 5 Mount equipment on resilient mounts and provide flexible pipe or duct connections to isolate vibration.
NOTES Main air connections and shuluff eocb are not shown.
OUTSIDE AIR
If the mhimum ouhlde air bv~ossis omitted. a s b ~ . aiiustable for balancinb; is provided on the ouhide alr damper ta prevent full closure and ta permit the passage of minimum outside air when in the closed (minimum open) position.
On merchant ships, acoustical and vibration control serve mainly as a means for comfort and to prevent damage to equipment. On naval and oceanographic ships, however, vibration and noise must be controlled for additional and more vital reasons. These include elimination of interference with detecting and measuring devices, prevention of detection by enemy, prevention of deafness, and elimination of interference with aural communications.
3. TWO-position valver the valve is closed below the change-wer temperature.
CHILLED' WATER
LEGEND HUMIDISTAT
ROOM THERMOSTAT
4 4 1 0 0
DUAL DUCT AIR MIXING UNIT
FAN
PNEUMATIC DAMPER & MOTOR
FILTER
CQOLING COIL
Fig. 5
Section 2 System General Requirements The spaces served by a common system depend upon the usage of the spaces, periods of occupancy, occupancy density, and odor potential; also, in naval construction, damage control classification is a prime consideration. System capacities from 5000 to 8000 cfm are common, with extremes being approximately 1000 cfm and 10,000 cfm. Naval ships are air conditioned by central station reheat systems which are referred to as "recirculation systems." The requirements for reheat zoning is the only basic variable. I n general, where the net heating load is negative, such as an internal space with a yearround cooling load, reheating is not provided. An exception to this is where one or more of the spaces served by a system require control of the relative humidity. Since the humidity controller can override the room thermostat, reheat is required to compensate for overcooliig. Generally, spaces are zoned according to heating requirements. Large spaces, such as mess rooms and "critical" spaces, may have individual reheaters or a separate recirculation system. Air-conditioning systems installed on merchant ships are classified as: (a) those serving small spaces (passenger staterooms, crew's quarters, offices, shops, etc.); and (b) those serving large spaces (dining rooms, lounges, 2.1
AilcConditioning Systems.
messrooms, etc.). Figure 5 illustrates the "Class A" system typical for air conditioning large (public) spaces. The outside and return (recirculated) air mixture is filtered and conditioned (preheated, cooled, dehumidified, and reheated as required) by the central station equipment. Air is recirculated in air conditioning systems for the sake of economy. Automatic dampers (outside air, recirculation, and exhaust) in Fig. 5 modulate the supply of outside air commensurate with the refrigeration system capacity; 100-percent outside air is supplied unless room conditions cannot be maintained with the cooling coil valve fully open. Note that cooling is controlled by both a humidistat and room thermostat. The room thermostat controls both cooling (water) and reheater (steam) valves, which are sequenced to maintain a set room temperature. When the humidity exceeds the humidistat set point, it overrides the room thermostat and opens the cooling coil valve for additional dehumidification. Should this cause overcooling, the room thermostat opens the reheater valve to maintain the set temperature. When a single Class A system serves several spaces, each has its own reheater controlled by only a room thermostat. A Class D (terminal reheat) system is shown in Fig. 6.
Clan A air-conditioning system with humidity control
M
MASTER
P
POSITIVE POSITIONING RELAY
P PREHEATER (STEAM) R REHEATERIW=WATER, S=STEAMI
@DUCT
THERMOSTAT
N.O. NORMALLY OPEN (VALVE OR DAMPER) N.C. NORMALLY CLOSED (VALVE OR DAMPER)
Conditioned air is supplied to each space to accommodate maximum design cooling load requirements. A mixture of outside and return air is filtered, conditioned (preheated or dehumidified and cooled as required) centrally, and distributed to individual reheaters at the spaces served. This system is used for passenger staterooms on most ships now in service, as well as for crews' quarters of many vessels. Note that automatic dampers (two-position) are controlled by two sensors (ductstats) in the air intake; one for the heating cycle and the other for the cooling cycle. The preheater thermostat is set several degrees below the design chilled air (off-coil) temperature to prevent simultaneous operation of the preheater and cooling coil. The Class E system, Fig. 7, is another system provided for small spaces requiring high-quality treatment. It is a primary air-secondary cooling and heating system;
PNEUMATIC CONTROL VALVE DIVERTING RELAY B.P. n l N l n u n OUTSIDE AIR BYPASS, SEE NOTE 2.
the central station equipment conditions onljr the primary air. This primary air is distributed to induction units located in the various air-conditioned spaces which induce ropm (secondary) air to flow through a water coil, an integral pad of the induction unit. The induced air is either heated or cooled by this coil. No r e t h (recirculated) air facilities are required. The plimary air is sufficient for ventilation purposes and to balance exhaust air requirements;-$ is conditiohed to take care of the entire ,latent load (outside air plus room). Primary air is reheated, as necessary, to provide all heating for spaces when the temperature is above the changeover point (described in the following) with chilled (secondary) water being simultaneously distributed to the induction units. Below the change-over point, the water-chilling plant is inoperative; hot water is supplied to the induction units and primary air is preheated
ENVIRONMENTAL CONTROL MARINE ENGINEERING COJD
HOT
RECIRCULATION -EXHAUST FROM PASSAGE
RECIRCULATION-EXHAUST FROM PASSAGE
-TO
-
u
EXHAUST
L TO ROOM
ROOM
STEAM
S E E FIGS. 5 8 6 FOR CONTROLS
OUTSIDE AIR
. ..
/CHILLED
\--WINTER
WATER HOT WATER SYSTEM MASTER THERMOSTAT ISEE FIG. 14)
SEE FIG. 5 FOR LEGEND AND NOTES
W S E E FIG. 6 FOR CONTROLS
Fig. 8
SEE FIG. 5 FOR LEGEND AND NOTES
Clau G air-conditioning system
Fig. 6 Class D air-conditioning system
(TYPICAL EACH ROOM1 TINDUCTION UNIT
TO OTHER UNITS
OUTSIDE AIR
+
I
I
SEE FIG.5 FOR LEGEND AND NOTES
SEE FIG. 6 1 FOR CONTROLS
Fig. 7 Clau E air-conditioning system
(tempered). The primary air is distributed a t high velocity and pressure and thus requires relatively little duct space. However, this space saving is offset to some degree by the space required for additional piping, secondary water pumps, and induction units. The change-over temperature may be defined as the outside temperature which exists when the internal room sensible load equals the transmission loss plus the cooling effect of the tempered primary air. I n central station cooling systems, e.g., Classes A and D, outside air is available up to the full system capacity. Once the out-
side air falls below the design off-coil temperature it can accomplish all of the necessary cooling, and the refrigeration equipment may be secured. Thus, the design offcoil temperature is also the change-over temperature. The change-over temperature of the Class E system, however, must be somewhat below the primary air off-coil dry bulb temperature (approximately 50 F) because it (primary air) provides only part of the sensible cooling load, i.e., about >5. Change-over temperatures usually fall between 30 and 40 F, 35 F being the most common. Note that room thermostats are of the dual pressure
type since it is necessary to change the action from normally open when the water is hot to normally closed when the water is cool. Figure 8 illustrates a varine dual duct or Class G system; it has gained considerable popularity in recent years. The high-pressure fan of a central station unit distributes conditioned air through two parallel systems of ducts (pipes) a t high velocity (approaching 5000 fpm). One duct system carries cold air and the other hot air. The air mixing units automatically proportion the hot and cold air to satisfy room loads. The series arrangement of the cooling coil and reheater shown gives excellent performance; it provides maximum dehumidification which in turn produces low room dew points. Accordingly, the possibility of condensation on cold ducts and equipment and the necessity for perfect vapor sealing of the insulation are minimized. Parallel arrangement of the cooling coil and reheater, which is common'ashore, has proven unsatisfactory. 2.2 Ventilation and Heating Systems. The central station air-conditioning systems discussed in the foregoing also are suitable for ventilation and heating, with cooling and recirculation facilities omitted. Fans amprovided with two-speed motors (100-percent and 50 or 67-percent speeds) with the low speed being used during the heating cycle to reduce the ventilation air heating load and air movement. Systems serving galleys, pantries, shops, laundries, and similar work spaces are provided with preheaters (50 to 60 F leaving temperature); they are referred to as "tempered air" systems. I n merchant practice some systems, such as those serving steering gear rooms, are shut down during the heating cycle, i.e., when not required to limit temperature rise; this permits the omission of preheaters.
I n general, it is most practical to use mechanical supply ventilation systems also for heating. Those spaces not served by air-conditioning or supply ventilation systems are heated by direct radiation. This also applies to those spaces served by supply systems which may be shut down during cold weather. References [9, 101 contain typical ventilating and heating requirements for merchant construction; similar requirements for naval construction are contained in reference [ll]. Galleys, bakeries, and sculleries are ventilated primarily by mechanical exhaust, mainly through hoods located over major heat-producing equipment. Mechanical supply, equal to about 50 to 90 percent of the exhaust, spot-cools working stations. Some supply is drawn (via accesses, grilles, or jumper ducts) from adjacent compartments or passageways to confine heat and odors. Large laundries have a mechanical supply and mechanical exhaust. Exhaust terminals are so located that equipment heat is not drawn across working stations. Some supply is natural to confine heat and moisture. In small deck laundries, the mechanical supply is omitted when a natural supply is from air-conditioned areas; the exhaust is located to prevent short-circuiting. Wherever possible, the exhaust from dryers is discharged directly to the weather with suitable lint-extracting facilities being provided. Mechanical supply ventilation is provided for auxiliary machinery rooms based on a Zmin rate of change or a 15 deg F temperature rise, whichever is greater. The mechanical exhaust should be between 110 and 1'20 percent of supply, where indraft is essential to prevent the dissipation af hot air to cooler adjacent spaces.
MARINE ENGINEERING
Natural exhaust, via other heat-producing spaces, is tration retard heat dissipation to working and watch areas. Most of the supply air is distributed directly to common. Pump rooms handling inflammable cargo have me- watch areas, so that personnel benefit from spot cooling; chanical exhaust systems with a 2- or 3-min rate of air there is a relatively small air temperature rise in the change; high and low terminals are provided, the latter blast area. Exhaust ventilation may be natural, mechanical, or a exhaust from bilges and other low pockets. Exhaust fans are direct-connected, centrifugal type with noil- combination of both. Theoretically, the exhaust equals sparking wheels and watertight fan shaft stuffingboxes. 110 to 120 percent of the supply. The excess exhaust The fan preferably is mounted on the pump room side accomplishes two things: (a) it compensates for the and the motor on the machinery space side of a water- thermal expansion of supply air, and (b) it creates an intight bulkhead; fans are driven by steam turbines draft to confine heat. Exhaust requirements generally where this arrangement is not practicable. Natural are determined without considering combustion air. The supply ventilation helps confine explosive vapors; the combustion air extraction, however, must not impede t g n k s extend down close to the operating level where the normal thermal flow of hot exhaust air up machinery necessary to avoid interference with the natural rise of space casings. Otherwise, the temperature within the air heated by the pumping equipment. Transfer pump casings will increase, which in turn will increase the heat rooms generally are exhausted by steam-ejector vent&* transmitted to surrounding spaces. The introduction of gas turbines has served to reemphasize the importance lators located on the weather deck. Battery rooms and lockers are ventilated to remove of this ~roblem. The annular mace between the outer the hydrogen gas released when charging. Require- staclc and uptake (smokepipe) provides an ideal passage ments with regard to battery room ventilation are for natural exhaust. Natural exhaust produces satiscontained in reference [12]. Acid-resisting paint protects factory results where air paths are short and there is ducts and centrifugal exhaust fans; the latter are located ample space to utilize low duct and terminal velocities (1000 fpm or less) ;where both of these conditions do not outside of the ventilated space. Winch resistor and motor-generator rooms require a exist, mechanical exhaust is essential. Shaft alleys on merchant ships are usually ventilated mechanical exhaust to remove and confine heat produced by the equipment. Heating, required to keep the equip- by natural supply from the machinery space at the forment dry, is accomplished by resistors furnished with the ward end of the shaft alley, and by natural exhaust via equipment. Propeller fans usually are most practical the escape trunk at the after end. This arrangement for these applications. acts as an equalizer when the combustion air consumpEmergency diesel generator rooms may have either tion varies. mechanical or natural exhaust (with natural or mechaniThe ventilation of cargo is by no means simple, as may cal supply). Mechanical supply has the advantages of be noted from reference [13], which was the first major being available for heating and minimizing infiltration. treatise on the subject, and references [l4, 151. Different Supplementary direct radiation is recommended, even cargos present different ventilation problems. Products with tempered mechanical supply, because of the copious such as grain, vegetables, and fruits are ventilated to infiltration through engine cooling facilities. When the cool the cargo and remove liberated gases, moisture, engine is operating, space ventilation is accomplished heat, and odors. Since cooling also reduces the moisture by the engine radiator cooling fan. This fan draws released from hygroscopic materials (wood, grain, paper, outside air (coolant) into the space, through the radiator, etc.), water damage as well as rot and mold growth is and finally discharges it directly to weather. Automatic reduced. It must be noted that dunnage, battens, dampers, at the weather supply and exhaust terminals, packaging, and bilge liquids also are si&ficant moisture are interlocked with engine starting and are open only sources. Regarding metal products, the ventilation when the engine is operating. serves only to equalize temperature and thus prevent Machinery space ventilation provides a tolerable at- condensation on the product, both on board and when mosphere for operating personnel and maintains satis- unloaded. factory ambient conditions for machinery. Judgment, On a typical winter voyage from a cold to warm based on previous experience rather than heat liberation climate, the cargo warms up slo+ly and moisture in calculations, generally dictates air requirements. Knowl- ventilation air will condense on the cold cargo. This is edge of successful treatment of spaces having similar referred to as "cargo sweat." Also under such circumchar teristics and equipment is particularly valuable. stances, hygroscopic cargos cool slowly and, therefore, On me hant ships with turbines or high-speed diesel continue to liberate moisture untii their temperatures propulsion equipment, mechanical supply rlystems usu- have fallen appreciably. Daily temperature changes ally provide one air change per minute. A rate of change due to sun variations can cause a cargo to heat unevenly; of 1.5 or 2 minutes is common for spaces with slow-speed unless removed, the moisture released may condense ondiesel or reciprocating engine rooms which are relatively the colder surfaces of both cargo and ship's structure. spacious. The supply air quantity falls in the range of The latter is referred to as "ship's sweat." Variations 6 to 10 cfm per shaft horsepower. in temperature of seawater, such as encountered by a Exhaust terminals over sources of high heat concen- vessel sailing through the Gulf Stream, have a similar
%
723
ENVIRONMENTAL CONTROL
I
effect. Cargo and ship sweat can cause various forms of damage which can jeopardize both the s h i ~and its cargo. ~ o n s e ~ u e n -t many l~, ships have special dehumidification and ventilation facilities to prevent sweat and to accelerate temperature equalization. Dry cargo spaces generally are ventilated by mechanical supply and natural exhaust, the latter directly to the atmosphere from each compartment. For many years a 30-min rate of change (2 air changes per hour) based on grain cubic, was considered adequate; however, in recent years a 20min rate of change has become more or less standard. Where fruits and vegetables are regularly carried in appreciable quantities, special ventilation is provided to suit the s~ecificreauirements. For further information on cargo 601d ventiation and dehumidification systems, see reference [9]. 2.3 Refrigerated Stores. Changes in food processing, the introduction of new construction materials and techniques, and economic pressures make the design of refrigerated spaces one of the most interesting aspects of shipbuilding. Frozen foods, frozen concentrates, precut and boned meats and similar innovations have materially reduced space requirements; however, improved palatability has increased the demand. Many operators now carry food in frozen form rather than that formally suitable for dry stores stowage. Individual preferences, as well as specific requirements of various trades (class of service, length of voyage, replenishment facilities away from home port, seasonal variation, etc.) are important considerations in planning refrigerated stores spaces. Generally, the owner outlines the segregation of commodities, holding temperatures, capacities (including reserve margins) and other essential design criteria. The bulk of refrigerated stores are stowed in insulated "walk-in" compartments [Fig. lS(a)] which form an integral part of the ship. Each compartment is fitted out to suit the particular produce to be stowed; for instance, unpaclcaged meats require hooks overhead, while those for frozen foods have shelves. Portable stanchions are provided to prevent shifting of commodities. Any general formula for estimating compartment sizes provides only a rough approximation to be used when exact criteria are not available. One method allows 10 pounds per person per day, with gross volume based on a stowage factor of 70 cu f t per ton. This does not include space for the thaw room, ice storage, or frozen foods. Maritime Administration requirements regarding refrigerated stores are listed in reference [8]; Table 1 shows the requirements for passenger ships. A thaw room (Fig. 18) serves several functions: (i) it improves sanitary conditions, as meat must otherwise be thawed in passageways outside the refrigerated spaces where the dripping and other refuse make good housekeeping difEcult: (ii) it is a work space, including a butcher shop where so fitted; and (iii) it is used for overflow stowage. Tangible advantages of a thaw room reflected directly in the refrigeration plant performance are: (i) it reduces infiltration loads; (ii) defrosting is
Table 1
Refrigerated Stores Requirements for Passenger Shi~s (Net Volume cu ft per 100 ~ a e s e n ~ eper r s Day) 1st 2nd 3rd Product Class5 Clad Classe Fresh fruits.. . . . . . . . . . . .... . . . 5.10 3.37 2.55 Fresh vegetables. . . . . . . .. . . . . 7.60 5.02 3.80 Meat.. . . . . . . . . . . . . . . .,. . . . . 17.65 11.65 8.83 Fish. . . . . . . . . . . . . . . . . . . . . . . .1 2.80 1.85 1.40 Poultry.. . . . . . . . . . . . . . . .. . . . 7.20 4.75 3.60 Butter.. . . . . . . . .. . . . . . . . . . . . 0,77 0.51 0.39 Cheese.. . . . . . . . .,::. . . . . . . . . . 0.18 0.12 0.09 Eggs.. . . . . . . . . . . . . . . . . . . . . . . 1.80 1.19 0.90 L d . .. . . . . . . . .. . . . . . . .. . 0.23 . 0.12 0.15 Milk.. . . . . . . . . . . . . . . . . . .'.. . . 2.15 1.42 1.08 Ice cream-. . . . . . . . . . . . . . . . . . . 2.60 1.72 1.30 I?
.I:.
Includes officers. Includes junior officers cadets, etc. Includes members of the crew.
required less frequently; and (iii) temperature fluctuations within compartments are negligible and thus the products are maintained in best condition. On many older ships, refrigeration equipment is located in the passageway serving the refrigerated compartments; the Liberty and Victory ships of World War I1 are examples. However, present merchant practice, fortified by automation, requires that the machinery be located in the main machinery space, along with that for the cargo and air conditioning, where most easily attended. This practice increases costs and requires a more careful design of refrigeration piping. Usually two identical condensing units (compressor, condenser, receivers, and controls) are provided on cargo ships and tankers; see Fig. 18. Both systems operate during pulldown, but only one is required for normal operation (reference [lo] requires each compressor to be adequate to handle the load during a 72-hr pulldown period). All compartments, large or small, have individual temperature control and are entirely automatic in operation. On large passenger liners there usually are several groups of medium-size "short-hold" refrigerators remote from the main stores refrigerated spaces; separate directexpansion systems are provided to serve the walk-in or reach-in refrigerators in each group. Where the individual short-hold groups are widely separated, a refrigerated brine circulating system may be used. Brine systems 4also may serve all of the ship's refrigerated stores compartments. Small miscellaneous (reach-in) refrigerators are provided for pantries, bars, and similar service spaces. These usually are cooled by integral condensing units or separate units located iri;r-or adjacent to, the compartments served. Chilled drinking water may be provided by a scuttlebutt or water cooler connected to the ship's stores. However, Maritime Administration specifications require, and many owners prefer, self-contained water coolers. In Navy prrtctice,,45 F chilled water from the air-conditioning plant is sometimes used to cool potable
724
MARINE Eb
ENVIRONMENTAL CONTROL
725
I
water, with small heat exchangers located near each fountain. Where a large quantity of chilled drinking water is required, it is considered good practice to install an independent refrigeration system for this service only, with the water being constantly recirculated. Cube and flake ice are made by automatic, selfcontained units, which are designed especially for marine use and to comply with U. S. Public Health Service regulations [16]. Ice cube makers may be fitted with attachments for crushing the ice and are arranged to discharge the ice into storage bins. By locating these machines a t the main points of consumption, throughout the ship, very efficient service is possible. Smorgasbord, salad pans, and similar food services are refrigerated; embossed plate type evaporators are used in most cases. 2.4 Refrigerated Cargo. Almost all modern cargo and passenger ships have part of the cargo space fitted out for carrying refrigerated products. This space often comprises 20,000 to 40,000 cu ft, divided into 2 to 6 compartments, each of 4000 to 14,000 cu ft. The size of compartments varies with the trade; much depends on the variety of cargos carried and their compatibility with respect to odors, method of packaging, susceptibility to damage, etc. Some ships are designed primarily to carry a single product, e.g., bananas, in a specific trade; all, or almost all, of the cargo space then is designed to carry refrigerated products. Since the "artJJdeveloped over a long period of years, often by trial and error, the reasons for certain refrigem tion practices may not be self-evident. The industry continually experiments on ways to improve pi-oduct preservations; such achievement can completely revolutionize an industry or create new ones. The United States Department of Agriculture, similar organizations in the United Kingdom and other foreign countries, as well as various fruit growing and transportation companies, all are vitally concerned and sponsor pertinent research in this area. The introduction of frozen foods has substantially reduced the design holding temperature for "universal" (general purpose) refrigerated compartments. Zero F to minus 10 F now are most common. Precooling of cargo appreciably reduces the plant refrigeration requirements. However, shipboard cooling of "hot" produce still is required in certain trades. Shippers frequently specify the pulldown, reduction, and holding temperatures for each product, as well as the maximum allowable variation during holding. One or two degrees F is commonly specified for the latter. To check temperature, locked recorders may be placed within the cargo block. Recent developments in automation, and more stringent regulations requiring highquality cargo out-turn, have greatly increased the use of automatic control, indicating, and recording equipment. Special recording instruments are required for "cold treatment," which is required to destroy certain insects in fruit [17]. For example, the insects imported with Vinifera grapes die when exposed to a su5ciently low
temperature (32 to 35 F) for an extended length of time (11 to 17 days) varying inversely with temperature. The handling of refrigerated cargos must be rapid to minimize temperature variations. Also, it must be carefully coordinated with operations ashore, including transfer to and from the ship. Compartments normally are precooled before loading to remove heat from structure, insulation, and fittings. This accelerates pulldown, and prevents damage or staining of precooled or frozen cargos due to contact with warmer surfaces. Since sweat forms on cold surfaces and the temperature rises when compartments are opened, every effort is made to load and unload an entire compartment at one time. Where this is not possible, curtains or other means are used to minimize these adverse effects. The procedure for handling unfrozen (chilled) beef is unusually painstaking [18, 191; special air circulation facilities are required to minimize temperature variations, i.e., 1 deg 1? and less. Fairly high humidity and exceptional sanitation are essential. Bananas are one of the largest cargos requiring refrigeration. Several recent innovations effect the stowage of this product. Polyethylene bags, provided to protect stemmed bananas, also reduce the deleterious effect of excessive air motion. Humidification facilities, provided in shore-based banana ripening and fruit storage rooms, have not proven practical aboard ship to date. High humidity is instead achieved by careful design of the evaporators (air cooling facilities). By minimizing differences between refrigerant and air temperature, moisture removal also is minimized; thus, the necessary high humidity is maintained. Heating facilities are required to prevent freezing of fruits and other "high-temperature" products. These facilities also are used to warm canned and other nonrefrigerated metal products prior to unloading; this prevents condensation on such cargos when unloaded in humid climates. Ventilation may be either intermittent or continuous; the lBtter is preferred since temperature variations are minimized. Details of ventilation requirements are covered in Section 3. 2.5 Refrigerated Cargo Containers. I n the design of refrigerated cargo containers for transportation by ship, corrosion, ship's motion, stowage stresses, weather tightness, and exceptionally high reliability must be considered, as well as the over-the-rad goals of minimum weight, maximum cubic capacity, and refrigeration facilities which can withstand severe vibration and shock [20,21,22, 231. Since both frozen and chilled cargo are carried, shipboard containers are "universalJJrefrigerated cargo carriers. As such, their capacity must be adequate to pull down "hotJJproduce. A number of different types of refrigerated containers have been used [20]. The regular land type is not suitable since refrigeration equipment is an appendage and thus does not permit economical stowage. Additionally, the materials are readily corroded by salt air, spray, and
a
I
4
stack gases and many commercial containers are not tight. These, plus unusual structural requirements for handling and stowage (stacking six high on some ships), necessitate that containers for shipboard service be specially constructed. Handling and securing attachments must be flush or removable. The minimizing of insulation to increase payload necessitates the elimination of metal attachments, and the associated high heat transfer, between the container's interior, and exterior. Tight external construction is essential to prevent air infiltration due to high wind velocities and vapor migration. Tests have indicated astonishing amounts of leakage; as a result, specifications include pressure and smoke tests to control this factor. Odors can contaminate cargo; therefore, materials are specified to be odor-free, or nearly so [21]. Full-scale road tests have shown appreciable variations in the performance of trailers. Also, theoretical calculations of loads are unreliable and often a 25-percent allowance is added as a margin of safety. The U. S. Department of Agriculture and interested private organizations have developed standards for rating refrigerated containers to assure reliable performance. Tests show (a) infiltration can equal the heat transmission load; (b) the solar load is appreciably reduced when the vehicle is moving and, under this condition, infiltration may be the major component of refrigeration load. Reference [2] discusses the design considerationsinvolved and lists the pertinent references regarding container design. Because of problems associated with operating aircooled cond&nsingunits in confined spaces, such containers are carried only on deck. Containers stowed below decks are provided with water-cooled condensers, preferably utilizing fresh water. Several nonmechanical methods may be employed for refrigerating containers. Liquid carbon dioxide (COz) may be exclusively used to refrigerate containers or to achieve a quick chill of the load in containers fitted with mechanical refrigeration. This system is not to be confused with the solid carbon dioxide or "drv iceJJ method. Liquid nitrogen functions practically th;! same as liquid carbon dioxide; however, it is said to offer greater possibilities for future use. Like carbon dioxide systems, liquid nitrogen systems have no working parts. The liquid nitrogen is carried in a specially insulated cryogenic cylinder which is really a large thermos bottle. There are several disadvantages associated with the use of these methods; one is that operating costs of nitrogen and C02 systems are higher than those of mechanical systems. , Another disadvantage is that a replacement supply of refrigerant may not be readily available. Nevertheless, acceptance by the transportation industry is progressing rapidly. 2.6 Air-Conditioning System Controls. Automatic controls are the brains of air-conditioning, refrigerating, and heating systems. It is essential they be carefully designed, installed and properly serviced; preventive maintenance is essential. Automatic controls fall into
three electric. categories; namely, pneumatic, self-contained, and Pneumatic control systems are the most common type used on merchant construction (see Figs. 5, 6, 7, 8, and 14). Generally, where pneumatic controls are used for any other control service, i.e., boiler combustion and hold dehumidification, it is easy to extend the compressed air system arid utilize this type of control for air conditioning and heating. Air from ship service systems is not suitable,.for this purpose because of its oil and moisture content; this is particularly troublesome on tankers and bulk carriers where there are long runs of weather-exposed piping. The best practice, particularly for passenger ships of appreciable size, is to provide one or more independent compressed air systems, crossconnected to the control or ship service system only for emergency operation. Primary compressed air (usually between 80 and 100 psig) is reduced in two stages, after drying and cleaning, to the 15 to 19 psig required. Heating-cooling thermostats (Fig. 7) require dual-pressure compressed air; the lower pressure is for heating and the higher one is for cooling. Switching of pressures is accomplished automatically or manually. Self-;contained controls are only used on merchant construction where there are few controls or where suitable compressed air is not readily available. On naval vessels, they are used for controlling steam coils; e.g., ~reheaters.reheaters. and combination heaters. Selfcontained controls are limited as to sensitivity, adjustability, and throttling range. They are not suitable where the available pressure drop is small, i.e., singlepipe heating systems,-or for systems requiring sequenced operation. The "metering" poppets of self-contained control valves are designed especially to modulate small steam capacities at relatively high pressures (30 to 35 psig). An integral strainer is provided. Electric controls are not used extensively in merchant construction. They are used primarily for simple applications where (a) compressed air is not available under all circumstances, and (b) suitable self-contained devices are unavailable. A typical example is the control of dampers at weather terminals serving emergency diesel generator radiator cooling. On naval construction, special dual-temperature thermostats are used to control chilled water coils and reheaters (steam and electric). Sensing elements for water coils, either temperature or humidity actuated, control on-off switches. Humidity control is ~rovidedonlv for "critical" maces. The folGwing recomkendations are offered with regard to the design of automatic control systems:
1 The selection of components (size, types, etc.) should be the sole responsibility of the manufacturer. He should design, furnish, and calibrate the system and guarantee its satisfactory operation. He should either install or, as a minimum, supervise the installation. 2 Equipment should have a special marine type corrosion-resistant finish.
726
MARINE ENGINEERING
3 Except for duct bulbs of controllers and thermometers, equipment should not be located in outside air intakes or in the direct path of outside air. 4 Damper motors and linkages should be mounted outside ducts, to minimize corrosion and dirt clogging problems. Outside air dampers should preferably be made of stainless steel; return air and exhaust air dampers may be stainless steel or mild steel, hot-dip galvanized after fabrication. 5 Damper construction should be integrally airtight without the use of nonmetallic seals. They should have simple, accessible position indicators. 6 Sufficient thermometers should be provided to check the performance of all automatic controls. Dials should be mounted on the control panel. The range should be suitable for the temperatures measured with graduations of 1 deg F preferred, but not more than 2 deg F. 7 Control and indicating equipment, with identification label plates, should be mounted on simple, well-
L,,
lighted, accessible panels in equipment (fan) rooms. Panels should also accommodate system label plates, diagrams of systems, and operating instructions (including air and water temperature schedules). 8 Panel-mounted air gages should be provided for checking the performance (branch-line pressures) of controllers located in equipment rooms. Elsewhere, tees or other means should be provided to facilitate testing and troubleshooting. 9 The instruction manual should include (i) schematic diagrams of each type of system with normal valve positions, spring ranges, control pressures, temperature settings, sequences of operation, and master-submaster schedules; (ii) air piping diagram of installation showing isolation valves, compressed air source, and pressure reducing stations to facilitate trouble-shooting; (iii) in- , stallation, operating, and maintenance instructions for each piece of equipment; (iv) spare parts list; (v) instructions for ordering parts; and (vi) information on manufacturer's service facilities.
I
Table 2
Outqide Design Conditions for Environmental Control Systems, deg F
Cycle Maximum cooling.. ........... 90 Intermediate cooling. . . . . . . . . . . . . Heating ..................... 10 Ventilating ................... 90
Wet Bulb 81
...
... ...
Sear rater 85
... 28 85
B";% 90-95" 75 0
Wet Bulb
Seawater
78-82b 75
85-90c 80 30 85
... ...
90
,'
" Maritime Administration standard is 95. Worth Atlantic service only 78; semitropical service 80; tropical service 82; Maritime Administration standard
I
in -- 82.
Inside Design Conditions for Environmental Control Systems, deg F
Cycle ET Maximum cooling.. ........... 74 Intermediate cooling. . . . . . . . . . . . . Heating. ....................... Ventilating. .................. 7 deg F risec
3.1 Air-Conditioning, Heating, and Ventilation Design Criteria. The itinerary, type of ship, and class
I
I
{, !I
,
Maritime Administration standard is 88.
Table 3
Section 3 Design Criteria and load Con~prrnents boundaries exposed to the sun, the maximum simultaneous heat gain may occur a t a time of day other than of service are considered when selecting the outside when the instantaneous load is maximum on either the design conditions for air-conditioning, ventilating, and horizontal or vertical surfaces. Additionally, the ship's heating facilities. Temperatures given in Table 2 may orientation with the sun changes frequently, often a t be used for merchant construction unless specific service short time intervals, which further complicates an accurate analysis of the solar heat gain. experience indicates otherwise. Naval practice is simple and empirical. Surface temThe inside design conditions listed in Table 3 are for "hotel" spaces. Exceptions occur in naval practice peratures of weather boundaries exposed to solar radi% where special equipment requires better control of tion are considered to be as follows: temperature and hu&dity conditions. The higher qualWood decks.. ................... .I20 F ity treatment indicated in Table 3, in general, applies to Steel decks.. . . . . . . . . . . . . . . . . . . . . .I40 F passenger accommodations. Hotel spaces on many Vertical steel boundaries. . . . . . . . . . .I20 F cargo ships are designed for an effective temperature (ET) of 72 F; values above 74 lc are uncommon in new The total heat transfer (solar plus transmission heat gain) construction. An E T of 73 1' is recommended with a resulting from these surface temperatures is considered relative humidity between 40 and 60 percent and an only for that weather boundary having the greatest heat average air motion of 20 fpm. In applying Table 3, all gain. The solar effect of other sun-exposed boundaries combinations of dry bulb, wet pulb, and relative hu- is neglected. Merchant practice varies. In some cases naval pracmidity used must fall in the E T range shown. Ventilating and heating requirements for typical mer- tice is followed; but in others the s8ar load is considered chant ships are enumerated in references [9, 101; similar to be the total of the individual maximum heat transfer information for naval construction may be obtained from through all sun-exposed surfaces, with the calculations being based on the surface temperatures shown for naval reference [ll.]. 3.2 Air-Conditioning Internal Cooling Loads. The practice. A more suitable approach is reflected in Fig. 4 internal cooling load of an air-conditioned space is com- and Table 4. Table 4 contains temperature equivalents, posed of the following elements: solar (sun), transmis- F, of the solar heat gain for spaces with single and multision, lights, equipment (sensible and latent), and person- ple solar exposures. The total heat gain includes the nel (sensible and latent). The instantaneous solar load normal transmission, which is calculated separately. depends upon the sun's azimuth and altitude, cleanliness Note that Table 4 allows for the reflection effect of bulkof the atmosphere, color of impinging structure, and head (or shell) surface color. Except for the heat gain other variables. Where a space has several weather through glass, the solar heat gain through a structure is
727
ENVIRONMENTAL CONTROL
gg
Wet Bulb
80"
68"
...
6SP.d
...
...
ET
g%
Wet Bulb
RH
71-78 71
75-85" 75-77 70-7B-d
66-69" 66-63
60-50 60-45
...
%
...
7 dig F nse~
" This temperature is only approximate since any combination of dry bulb and wet bulb falling on the 74 ET line is acce table. 80 F dry bulb is used for load calculations. b & other than hotel spacesl the tem erature may vary from 40 to 60 Fl the lower value being uaed where heating is provided only to prevent freezing of lquids or to facilitate operation. 60 F generally is suitable for workshops and other spaces where physical activity permits a reduced temperature. c Temperature rise is the maximum allowable difference in dry bulb temperature between air entering and leaving a space. It is the equivalent to At in equation (3). In heat-producing spaces the temperature rise (average) may ap roach 80 deg F. P75 F in hospital a aces and toilet apacsa of paasenger ships; 80 F in operating moms. The Maritime Administration standard is 70 On naval construction, 75 F for shower and toilet qaces. a Maritime Administration standard is 80 F dry bulb and 66 F wet bulb.
8
" -"
obtained by the use of appropriate At values in equation (7). ~r&smission heat gain through surfaces other than those exposed to solar radiation is calculated the same for both naval and merchant practice, with one possible exception. In naval practice, credit is allowed for the cooling effect of adjacent spaces designed for a lower temperature; this usually is not done in merchant practice. Equation (7) is used for calculating transmission gains. The temperature of ventilated spaces equals the outside design (cooling) temperature plus the allowable temperature rise. An assumed temperature must be assigned to nonventilated spaces and those whose ventilation is not based on temperature rise. When assigning such temperatures, consideration is given 'to the temperature of surrounding spaces, heat sources within the space, insulation treatment, and the space's function. Assumed temperatures in Table 5 are applicable to merchant construction; similar data for naval construction are included in reference [ll]. If the wattage of lights is known, the lighting load heat gain equals the total wattage X 3.41; that for fluoresaent lights must include ballast, which can be approximated as 25 percent of the light wattage. Lighting
loads on merchant ships often are so high that use factors similar to those in Table 6 are applied. If actual wattages are not known, the data in Table 7 may be used as an approximation. Equipment heat loads usually are electrical, except in commissary and hospital spaces where steam or water heated equipment also is used. Use factors are applied to allow for intermittent operation of equipment [I]. Hoods and strategically located return and exhaust terminals can effectively minimize the heat dissipated by equipment. Naval practice with' regard to equipment heat loads is described in reference [ll]. The sensible heat and latent heat dissipated by personnel are shown in Table 8. The total personnel heat load of a space is obtained by multiplying the number of persons in each activity classification by the corresponding sensible and latent heat allowances. 3.3 Cooling Coil k d . The fotal system cooling coil (refrigeration) loads are: Room internal sensible Outside air sensible Fan heat Duct loss (supply) and gain (return)
Room internal latent Outside air latent
'I[ 5
4
728 Table 4
MARINE ENGINEERING Table 5
Equivalent Temperatures for Solar Heat Gain
EQUIVAI,ENT TEMPERATURE DIFFERENCE, deg F Bulkhead Solar Exposures Deck or Shell Wood deck only.. . . . . . . . . . . . . . . . . . . . . 35 Steel deck only.. ..................... 50 Bulkhead or shell only--dark color. .... 35 Bulkhead or shell only-light color. .... 20 Wood deck and bulkhead (or shell)dark .............................. 25 25 Wood deck and bulkhead (or shell)light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 15 Steel deck and bulkhead (or shell -dark 35 25 Steel deck and bulkhead (or ahell{-light 35 15 Single glass, deck over space exposed". . 110 Btu/hr-sq ft Single glass, deck over space not exposedb 80 Btu/hr-sq ft a Equivalent temperature for solar heat gain only; transmission gain is not included; bThe shading effect of curtains and blinds is included. See reference [I] for double and special glass. For additional information, see Fig. 4.
When both the fan (or other driven equipment) and motor are in the airstream, or treated space, the heat gain is: Motor brake horsepower X 2545 Htnf = (9) Motor efficiency The efficiency of electric motors may be taken from Table 9. The heat gain when only the fan is in the airstream, i.e., a centrifugal fan or a belt-driven axial-flow fan, is: H j = Motor brake horsepower .X 2545 (10) For merchant construction, the fan heat can be estimated as follows: H j = FSPQ (11) where
Assumed Inside Temperatures of Spaces
ABBUMED TEMPERATURE, deg F Space Cooling Heating, Machinery and boiler spaces, deck. ..... 130 70 Machinery and boiler spaces, bulkhead. . 120 70 Machinery and boiler casings, deck. .... 140 70 Machinery and boiler casings, bulkhead. 130 70 Motor generator and resistor, deck. . . . . 130 30 Motor generator and resistor, bulkhead. . 120 30 Toilet space in air conditioning area. ... b 70 Miscellaneous stowage and lockers in air conditioning area.. ................. L 4 Passages and stairwells.. .............. b 70 Galley and pantry hood areas. ......... 140 70 Miscellaneous stowage and stores spaces outside air conditioning areas. ....... 105 50 Settling tanks.. ...................... 120 70 Fuel oil tanks. ....................... 120 40 Cargo oil tanks. ..................... 2 , 40 Voids above waterline. ............... 120 40 Voids below waterline.. ............... 100 40 Fresh water tanks. ................... 120 30 Cargo spaces above waterline. ......... 110 20 Cargo spaces below waterline. ......... 100 40 Seawater............................ 85 30 Shaft alley. ......................... 105 40 a Depends on ventilation treatment and number of unheated exposures. The following are recommended:
SOURCE SUPPLY Heated Tempered Weather OF
Exposures 0 1 2 over 2
729
ENVIRONMENTAL CONTROL
70 65 55 40
60 50 40 30
50 40 30 20
Air-conditioning design inside dry bulb temperature plus 5 deg F. "Use cargo pumping temperature, t,, except where it may cool during a voyage, in which case use t, - 20, but not less than 110 F. b
culations, the heat gain in the return air path (passageways, stairwells, fans, and ducts) may be assumed to be 5 deg F. The outside (replenishment) air sensible and latent H = fan heat, Btu/hr loads are obtained. from psychrometric data as outlined F = factor: 0.65 for centrifugal fan (outside space) in Section 1. 0.85 for axial and propeller fans 3.4 Heating and Ventilation Loads. The space SP = fan static pressure, in. of water heating load is composed of the heat lost through cold Q = fan capacity, cfm boundaries to the weather and colder adjacent spaces When better information is lacking, the fan air temper* and heat required to raise the outside (replenishment) or ture rise, tj, in deg F may be approximated as follows: infiltration air to the space temperature. No allowance is made for solar heat gain. In merchant practice, no . credit is taken for heat gain from warmer spaces; e.g., The heat gain in the supply ducts of an avergge system uptakes, machinery spaces, and similar hot spaces. with 1-in. board insulation may be assumed to equal a Also, the heat gain from personnel, lights, and equipment 2 deg F rise, except where the longest run of duct after the is disregarded. See reference [Ill for information recooling coil is such that the rise is greater based on a garding naval construction. Spaces which obtain their supply from an adjacent 1.5 deg F rise/100 ft. The heat gain in the recirculation air path (passage- space will have an irrfiltration load if the supply is colder ways, stairwells, and hotter surrounding spaces) may be than the heated space. Where a space, such as the calculated; however, a rise of about 3 deg F is frequently wheelhouse, is adjacent to weather, the heating load assumed. Where return ducts are less than 25 f t long, includes an infiltration allowance for cold air entering the heat gain may be neglected; where longer, the rise when a weather door is opened. A space's ventilation load consists of solar load, transmay be estimated as 0.5 deg F/100 ft. To simplify cal-
mission from warmer .adjacent space boundaries, and Table 6 Use Factors for Electrical Lights, Percentage of heat gain from lights and equipment; only sensible heat Watts Installed is included. Heat from personnel is neglected. Total Watts NUMBER OF FIXTUREB The heat gain through each boundary may be calculated Installed 4 or less 5 and 6 7 and over from equation (7). The solar load is based on the differ- Up to 200 100 ... 301-400 100 90 80 ence between the surface temperature (95 F plus solar 201-300 90 80 75 equivalent temperature from Table 4) and either the 401-600 ... 75 70 design space temperature (outside air design temperature 601 and over ... / 70 60 plus design temperature rise) or an assumed space temperature (Table 5). Loads due to lights and eaui~mentheat dissi~ationare discussed in sections 3.2 and 3.3. table 7 Assumed Lighting Loads The quantity of ventilation air required for a given (Watts/sq ft Deck Area) cooling load is calculated from equation (3). Note that Space Incandescent Fluorescent4 a temperature rise equivalent to supply fan heat, t j in Passenger staterooms and offices. ... 4 3 equation (12)) must be deducted from the allowable Officers staterooms. .............. 3 2 temperature rise to obtain the net rise used in equation Crew staterooms. ................ 3 2 5 4 (3). This consideration may establish whether mechan- Messrooms ...................... ical or natural supply ventilation is selected for certain a Includes heat of ballast. spaces, such as those for refrigerated cargo. 3.5 Air-Conditioning Load Calculations. A form similar td that shown in Fig. 9 is normally used when performing air-conditioning load calculations for the sun-exposed surface are calculated separately to simplify various ship spaces. Instructions for completing the partial load determinations such as those required a t change-over. The solar load is constant irrespective of Fig. 9 form are as follows: the outside dry bulb temperature. As an example, if 1 The appropriate "key" is entered in the "key" To is 35 17, the transmission loss is a negative cooling column to identify the load component computed. load; however, the solar gain is still positive and, being Standard key designations are: the same as previously used to determine maximum cooling load, need not be recalculated. KEY LOAD 2 "T" is the design insideadrybulb temperature and Deck Gver. .......................... 1 is entered for keys 1 through 6 inclusive. It may be Deck under. ......................... 2 taken from Table 3 or as specified. Outboard bulkhead or shell. . . . . . . . . . . . 3 3 "To" refers to .the temperature outside of the Forward bulkhead. ................... 4 space being calculated; it is used, with equation (7)) to After bulkhead. ...................... 5 compute the transmission loads corresponding to keys 1 Inboard bulkhead. .................... 6 through 6. When To is not specified, use 95 F for Lights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 weather boundaries. When the adjacent space is venEquipment ........................... 8 tilated, use 90 F plus the specified temperature rise or an Personnel. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 assumed inside temperature taken from Table 5. Infiltration. . . . . . . . . . . . . . . . . . .1. ...... 10 4 Data regarding insulation treatment, dimensions, The same key legend is used in both cooling and heating and areas are obtained from the ship's drawings. load calculations; however, key 10 is not applicable to 5 The "U key" column is provided to identify the cooling calculations and keys 7, 8, and 9 are not applica- insulation treatment of the boundary in question. U ble to heating calculations. Note that keys 1,3,4, and 5 keys for various construction configurations are obtained are entered twice for each surface exposed to the sun; from Fig. 4. The appropriate overall heat transmission i.e., once for the transmission and once for the solar gain, coefficie~t(U factor) is also obtained from Fig. 4 and is Table 4. The transmission and solar heat gain of each entered in the "U" column.
-
0 . .
Table 8
Activity Classification
Personnel Sensible and Latent Heat Dissipation, ~t$hr
S 75 L
Waiters .................. 350 Persons Eating4.. ........ 280 All Others.. ............. 270 a
510 240 230
a
ROOM DRYBULBTEMPERATURE, Deg F 76
S 320 260 250
L
s
540 260 250
290 250 230
30 Btu/hr is allowed for both the sensible and latent heat of food.
78
L
s
570 270 270
260 240 200
80
L 600 280 300
MARINE ENGINEERING
I I/I/1
Table 9
ENVIRONMENTAL. CON'TROL
The total heating load is simply the sum of the Hw column.
Electric Motor Efficiencies
Motor Size, hp Motor Efficiency Less than gz.. .......... 0.45 342 to %... ............. 0.50 % t o 3);. ............... 0.60 % t o g ................. 0.65 % to I . . . ............... 0.75 I t 0 5 ................... 0.80 5 and larger. ............ 0.85
3.6
6 "TD" denotes the temperature difference across a boundary and is Ta-T for cooling transmission loads, equivalent temperature differentials (Table 4) when computing solar heat gains, and T-To for heating loads. 7 The transmission and solar heat gains through boundaries, HEin cooling calculations and Hw in heating calculations, are determined from equation (7), i.e., area X U X TD. 8 The lighting sensible cooling load is computed as 3.41 times the lighting wattage load. The data presented in Tables 6 and 7 may be used for guidance. 9 Equipment cooling loads are usually sensible and are computed from equation (9) and Table 9; however, latent loads may also be incurred, particularly in commissary and hospital spaces. 10 The latent and sensible heat dissipated from persomel is computed from the data given in Table 8. 11 The infiltration heating load is determined from equation (3), heating TD, and the specified infiltration rate. After the various individual load components are computed for the space in question, the sensible, HE, and latent, HL, heat columns are totaled. The space load, HT, is then determined by summing the sensible and latent heats and the sensible heat factor is computed for each space by using equation (6).
Refrigerated
Compartment
Design
individual requirements of the service, and owner's preference generally preclude the necessity for exacting calculations. This "phiosophy" is further strengthened by the fact that refrigeration equipment is selected on the basis of 18 hr operation per day during the holding ~eriod. The components comprising refrigerated compartment loads are basically the same as those forming air-conditioning cooling loads. The load components are classified as follows:
Criteria,'
Refrigerated compartments are assumed to have boundary temperatures as enumerated in Table 10 for merchant construction. The optimum storage conditions vary with the product to be preserved [2]. Stores which are compatible with respect to storage temperature, humidity, odor generation, and contamination sensitivity may be stowed in the same compartment. common refrigerated stores compartment designations and design storage (holding) stores are dven in temperatures for merchant
1 Transmission (includes allowances for solar heat) 2 Product 3 Container 4 Product respiration heat 5 Ventilation or infiltration 6 Fan, motor, and pump heat
,laulc: _,,_ ,,
11.
The optimum storage conditions (temperature, humidity, and air motion) for a perishable cargo are not static, but depend upon many factors [2, 241. Refrigerated cargo is classified according to its condition when loaded; i.e., frozen, precooled, or hot. As will be seen in Section 3.7, a product's loading condition appreciably affects refrigeration capacity requirements. The refrigeration facilities of most general cargo ships are of the "universal" or "all-purpose" design and are suitable for most perishable products. Such installations are designed for both frozen cargo and chilled produce; the former commonly range from minus 10 to 0 17 and the latter as high as 55 1'. Maritime Administration Standard specifications require three design (holding) temperature conditions; namely, -10, 4-35> and +50 F. The - 10 F holding temperature is particularly costly and, therefore, should only be applied where the vessel's e dshiDJsstores. trade demands it: this also a ~ ~ l i to 3.7 ~ e f r i ~ e r a t e~do m ~ a r i k e cooling nt Loads, The refrigeration system's capacity must be adequate for both "pulldown" (cooling the product from the loading temperature down to the holding temperature) and "holding" (normal) operation. Shelving and other stowage facilities, the great variety of products handled, other
.
The respiration (for live products) is comparable to the personnel load. Personnel and lighting loads, however, are disregarded since they are negligible and intermittent. An analysis of loads always should be made to determine which component (product load, ventilation, etc.) is controlling; possibly a slight relaxation of this item may permit a much more economical and efficient installation. Considerable study is required to obtain the optimum plant design. The procedure for calculating loads and selecting refrigeration equipment for naval construction is fully described in reference [25]. Commonly assumed design temperatures for adjacent spaces are contained in Table 10. Suggested overall transmission coefficients are given in Table 12; these include a margin of safety to cover moisture absorption and the higher conductivity of materials used for furring and supports. Equation (7) is used to calculate the transmission load. The product load which occurs only during pulldown is calculated as follows: H,
=
WSAt/T
73 1
Table 10 Assumed Boundary Temperatures for Merchant Ship Refrigerated Compartments
SURFACE BOUNDARY Sunexposed weather deck, wood. . . . Sun-ex~osedweather deck. steel.. ...
TEMPERATURE, deg F Cooling 120 1 4--0 0 -
Heating
~ ~ $ ~ ~ ;~ : ~ tg ~ Adjacent refrigerated spaces. ....... Surface exposed to seawater. ....... Adjacent to h a ~ sauares.. h ........ Other interior apace;. ..............
Design temp.b 85 100 106
... ... & ... ...
$ ~
10
in 30 -C
Maritime Administration standard is 120 F. Consider onl if adjacent space maximum design temperature is higher than t L t of space being calculated; base heat gain on maximum design (refrigeration) temperature except use 100 F for refrigerated cargo spaces (for cases when the spaces are unrefrigerated). Neglect unless adjoining space may be unheated, then use 35 F. b
Table '1 1
Refrigerated Ship's Stores Temperatures for Merchant Ships
PRODUCT OR SPACE DEBI~NATIONTEMPERATURE, . deg -F
Meat0 ................................ Fish. ................................. Frozen foods and ice cream.. ............ Fruits and vegetables. .................. Dairy ................................. Thaw rooms and vestibules. ............. Scuttlebutt. ........................... Potatoes. ...................; ......... Flowers. ............................... Smoked meat. .......................... Mineral waters, champagne, wkes and liquors. ............................. Beer .................................. Bonvoyage ............................ Butcheshop ...............a.e*....... Morgue ............................... Ice storage. ............................
minus 10 t o 0 minus 10 to 0 35 35
--
60
0 15
a Some cargo ship operators stow both meat and fish in the same compartment.
(13)
where H, = product heat load, Btu/hr W = weight of product stowed, lb S = specific heat of product, ~tu/lb-degF At = cooling range (difference between loading and holding temperature), deg F T = time allowed for pulldown, hr (48 for stores and 72 for cargo)
Fig. 9
Air-conditioning load calculations
Table 12 Overall Transmission Coefficients for Merchant Ship Refrigerated Spaces
CONBTRUCTION
INSULATION U FACTOR THICKNEBB, Btu/q ft-hr-deg F IN.0
Bulkhead with stiffeners on side opposite insulation Bulkhead and deck with stiffenera penetrating insulation Deck-underfoot; cellular glass insulation to ped with 2 to 3 in. minforcecl'concrete mastic Deck underfoot, cellular gWw (1 in.) and polyurethane insulation topped with 2 to 3 in. reinforced concrete mastic Nonmetallic dividing bulkhead (no steel framing)
{a
0.06 0.04 0.10 0.07 . -
0.09 0.07 0.05
A typical refrigerated compartment is shown in Fig. 10. The gross volume of a compartment is the volume included between the finished deck underfoot, inside 0.09' surface of the insulation sheathing overhead, and inside 0.06 insulation sheathing bulkhead to bulkhead (or air cooler space of cargo compartments). The gross volume ex0.08 0.06 cludes the volume of the air cooler (cold diffuser) space, Figs. 37 and 38. The net volume equals the gross volume a Minimum insulation thickness over stiiTeners, beams, and other r stor- structural members is 2 in. for compartments 35 F and higher, and less deductions for space not available f ~ product for those below 35 F. age; i.e., that required for access, air circulation, package 4 in. b Total thickness of cellular glaaa plus polyurethane.
k
MARINE ENGINEERING
,NruunoN
spacingJ ducts, evaporators (cold diffusers, air cooler~J and wall coils), gratings, battens, and temperature re cording and controlling devices. The relationship b e tween the net and gross volumes of typical refrigerated
---
Table 13
Stowage Densities of Common Refrigerated Products
DENSITY,
PRODUCT
Apples (in boxes). ........................... Bacon(inboxes) ............................ B-s ................................... Beef frozen and packed). .................... Beef ~ h m gin quarters). ..................... Beef (boned and compressed). ................ Butter (in kegs or cases). .................... Cheese (in boxes). ........................... Fish in boxes) ............................... ~ i . 6[frozen). ............................... Lemons .................................... Lamb or mutton. ............... ,........... Oranges. ................................... Potatoes in bags). .......................... potatoe. [in b a d ) . ........................
ft/long tono 106 64 120 51 125 67 67 68 70 80
-
Table 15
Product Load Data for Cargo
Specific heat, Btu/lb-deg F . ..... Chilled Frozen
0.90 0.45
Stowage density, p d . . .......... 25 Product temperature when loaded, Frozen 25 deg F . . ..................... Precooled 65 Hot 85 Common holding design tempera- frozen -10 ture, deg F . ................ d . Chilled 30 tot o350 (0 [I/ at055 3 Chilled Pulldown time, hr.. ............ 72 NOTE:Figures in parentheses indicate ventilation air changes per hour based on &OEM volume. Maritime Administration reqmrement for 35 a n t l a F (chilled) cargo is 2 air changm/hr.
Long ton equals 2240 lb.
Respiration rates vary with the product, its temperature, and other factors [24]. For general PurPoses, ever, Table 16 may be used to calculate the product rnspidon heat load. The average of the respiration rates at loading and holding temperatures 'is used for calculating pulldown loads. For ship's stores a rate of 0.09 (0.13 at loading, 0.05 at holding) may be usedThe respiration heat load is calculated as follows:
Fig. 10 Typical refrigerated compartment
Table 14
Product Load Data for Stores
MEAT ITEM
SpecificHeat Btu/lb-deg F . . Stowage denkty, pcf.. ...... Cooling range, deg F . ...... Pulldown time, hr. .........
AND
FISH 0.45 35 15
FRUIT AND
VEGETABLES DAIEY
48
H , = respiration heat load, Btu/hr
W R
= weight of product stowed, lb = respiration rate, Btu/lb-hr
Mechanical ventilation is provided for the refrigerated cargo compartments of most modern ships sufficient to produce the air changes noted in Table 15. For cargo the ventilation is considered only during holding, except for bananas, which require ventilation during pulldoand holding. The normal infiltration'from door oper* tion provides ample ventilation for stores compartments, and the in6ltration load is included in both pulldown and holding calculations. Figure 12 shows reasonable infiltration rates for stores spaces entered from a thaw room. These rates must be multiplied by 1.5 for spaces not protected by a thaw room, and by 2.0 for thaw rooms to allow for heavy door usage. Ventilation and infiltr* tion loads may be calculated from psychrometric data and equation (5); however, charts, e.g., reference [25], or tables, e-g., Table 17, are more practical. Using Table 17 the load, H v (Btu/hr), is determined from the following equation: H v = VoRAH (15)
V O = P O S S volume of compartment, cu ft R = rate of air change (Table 15 for cargo, Fig. 12 for stores) AH = heat removal, Btu/cu ft (Table 17)
z sS0
.45
.40 .%
Fig. 12
Avdrage infiltration rates f a ship's stores c m p a h e n t r entared from a thaw room
Table 16
Product Average Respiration Rates
TEMPERATU~E, deg F 35 45 55 65 75 85
REBPIRATION RATE, Btu/lb-hr 0.04 0.06 0.08 0.11 0.13 0.17
734
ENVIRONMENTAL CONTROL
MARINE ENGINEERING Table 17 Heat Removed in Cooling Air to Storage Conditions, Btu/cu ft
AMBIENTAIR
e
STORAGEROOMTEMPERATURE, Deg F
Values of 100 F and 50-percent RH are commonly specified for the conditions of entering (outside) air. The Maritime Administration Standard specifications require that the ventilation load be baaed on 100 F and 80-percent relative humidity outside air, and that pulldown of the entire cargo be accomplished within 72 hr. Fan and motor heat is introduced into the refrigerated compartment both by the ventilation and recirculation fans. The heat equivalent of pump power must also be included as part of the refrigeration load. (See Section 3.3 for data applicable to fan and motor heat loads.) The difference between motor nameplate and fan brake horsepower may be significant; thus, this should be checked to avoid excessive allowances. Also, since fans add . - heat to the air, natural supply ventilation with mechanical exhaust (fan and mot6; outside the refrigerated space) may be used to avoid this load. Air cooler fans normally operate at full speed only during pulldown; therefore, the slow speed, which is usually 50 percent of maximum, is used during holding and reduces the heat load to ?4 of the maximum value. 3.8 Drinking Water Cooling and Ice Making. Water cooling facilities usually are designed to deliver 50 17 water with 90 F entering water. The water cooling
load, H, (Btulhr), is calculated as follows: H, = 8.33 GR where
(16)
G = cooling capacity, gal/hr R = cooling range, deg F The average cargo ship cooler (scuttlebutt) has a 50gallon storage capacity and a cooling capacity of 10 gallhr. Refrigeration loads for large, extensive drinking-water systems, such as found on a passenger ship, include circulating pump heat, as well as heat gain through the distributing piping. The ice-making load, H i (Btu/hr), is expressed as follows: H i = W[144 (tl - 32) 0.5(32 - tz)] (17)
+
+
where = weight of ice made, lblhr tl = entering water temperature, deg tz = ice temperature, deg F
The common standard is to consider the entering water to be 100 F and the ice temperature to be 15 F.
Section 4 Piping Systems 4.1 Steam Heating Systems. The layout of a steam heating piping system depends upon the vessel arrangement. Cargo ships generally have relatively short fore-and-aft runs of piping; a steam riser in the machinery casing starts a t a pressure-reducing station below the lowest, and extends up to the highest$ deck served. Where areas served are some distance fore and aft, such as on troopships and passenger ships, the steam main is run fore and aft, preferably above the bulkhead
deck and on the deck level having the largest height so as to minimize conflicts with headroom and other services. Risers, taken from the main a t convenient lace tions, feed short laterals on the various decks. Often a horizontal steam main can serve two or even three decks (one or two up and one down). When avoidable, the longitudinal main is not run on a passenger deck. If possible, piping is kept out of cargo holds and other inaccessible (locked) spaces, where leaks may be unde-
tected [261. Also, it is kept out of electrical spaces and is not allowed over electrid equipment. The requixement for a gravity return is a major controlling factor in locating condenmte piping. To achieve an economical and efficient layout, the heating system must be, coordinated with the other hotel services. Structural interferences and limitations also must be considered. The following guidelines are considered to reflect good practice in the design of shipboard steam heating system: 1 Provide constant steam service for equipment operating year around (galley, pantry, water heaters, laundry, converters, air-conditioning reheaters, and filter cleaning stations). Feed the remaining equipment from intermittent steam service. 2 Provide either a common or individual reducing station for each group of units requiring the same operating pressure. Some galley equipment may require lower pressure than heating equipment; also, some units have integral reducing and control facilities. 3 Run all piping (steam and condensate) so it may be drained by gravity. 4 Design condensate piping for gravity return to an atmospheric or contaminated drain tank. Run independent drains from laundry equipment which operates at a higher pressure than other equipment. Where lifting of condensate absolutely cannot be avoided, limit the l i t to one deck height only and fit check valves. Wherever possible, install horizontal runs of condensate piping level or pitch it in the direction of flow. Also, provide drips a t all low points in piping, the base of risers, and ends of mains. 6 Locate drain piping from heating units in exposed compartments, such as gear rooms, below the unit; if located above, the unit can be flooded and damaged by freezing if steam is inadvertently secured in cold weather. 7 Except for converters, do not provide by-passes for traps or automatic valves. provide a gage in each equipment (fan) room, and elsewhere as necessary, to determine the actual working pressure a t heating units. Also, provide plugged test connections a t several points. 9 Provide all units with a stop valve on supply and a trap and cutout valve (gate) on the return. 10 For each air and water heater, provide a strainer and control valve (following stop valve) on the supply and a dirt pocket before the trap on return. 11 Provide a dirt pocket and strainer ahead of the steam trap on a unit heater return. 12 In general, use thermostatic traps only for direct radiation and domestic equipment of low capacity; i.e., less than 20 pounds per hour. 13 Provide an air chamber and petcock in returns from heaters.
735
diameter, in.) are generally accepted and specified. Usually, the steam is slightly superheated and will be dry; for wet steam (as supplied by a waste-heat boiler, ordinary heating boiler, or contaminated evaporator), capacities determined from this equation are reduced about 25 percent. Condensate return lines commonly are sized for maximum velocities, fpm, assuming full pipe flow, derived by the i'ormub V = 30fld. The pressure loss in return lines of 3Gpsig systems should not exceed 4 psi [27]; veloGties calculated from the preceding equation produce between and l psi pef lQO ft of straight pipe. 4.2 Hot-Water Systems. The use of hot water for heating has many advantages, Hot water affordsbetter control of small heating capacities (frequently 5000 Btu/hr and less) than steam. Also, hot watsr eliminates numerous problems associated with condensate return, steam traps, and water hammer. Hobwater heating permits elimination of (a) condensate return pumps and associated problems, (b) condensate cooling facilities, and (c) corrosion of condensate piping. It is to be noted that corrosion in steam heating boilers, particularly those of the flash type, is a serious problem which requires careful feedwater analysis and treatment. (See reference [27] for additional information on the use of hot water for heating services.) The present trend is to use medium-temperature hotwater generators in lieu of auxiliary steam boilers. For various heating services, including distilling plants, ltiundry, and galley equipment, 320 F water replaces 100-psig steam. Water-to-water and water-to-steam heat exchangers are provided where low-temperature heating (water Or is required. This practice has the
1 2 3 4
Less weight and space Lower installed cost Safer and easier to operate Less maintenance and lower operating cost
All hotrwater heating systems on ships have forced circulation. Single-pipe systems, Fig. 15(q), require minimum piping, are ideally suited for shipboard applications, and have been successfully used on many large passenger ships. The true two-pipe reverse-return system is frequently not practical; therefore variations, such as that illustrated in Fig. 15(b), are adopted which are most suitable to the arrangement of the spaces served. (See reference [27] for a further discussion of single and two-pipe 'yStem designs.) On large ships, a numbsr of separate heating systems may be provided with central station equipment, Fig. 14, located in fan rooms or machinery spaces. Single-pipe systems are particularly suited for decentralized service. i.e., service required only in certain portions of t h i Typical equipment hookups are shown in Figs. 13 vessel. Where central station equipment is located in and 14. main or auxiliary machinery spacei, two-pipe arrangeMaximum steam velocities, fpm, in pipes derived from ments are used with two or more pumps and converters the formula V = 4 8 6 0 d (where d is the pipe h i d e being provided to carry the design load; a stand-by pump
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
60
3;
g
83
I0
I-F
do' I3 LJ 4 I3 0
-5@
262
ggg =x-'
2gY2 J
Eg
-Irw
a c m
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
Co's
%a
Lg
P * a y2 U)
I-F ui+ gz
-5 @
8 ,-z
0 K 3
3;
n
ax-',
ZiE acm
4
=-0
", a
ow,
zg
g;k" 52 2' "2
v t p*
2:
ot =L so
2~ aH r w Es SE
73 8
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
cannot be overstressed. Automatic float-type vents with waste connections usually are applied only at high points in mains and those heating units at high points of systems. Systems with open expansion tanks are generally considered to be easier to vent than those with compression (closed expansion) tanks. The following equation shows the relationship between the heating (or cooling) capacity, water temperature drop, and flow rate:
TYPICAL ZONE CIRCUIT FLOW FI r t l u E
EaulPnENT
. .....
OTHER ZONES
(4SINOLE PIPE SYSTEM
where TYPICAL ZONE CIRCUIT
/TO OTHER ZONES
7
L ----
----) &
t
I
CONVECTOR H
O%ER ZONES
E
t/
H TD G
= heating (cooling) capacity, Btu/hr = temperature drop (rise), deg F = flow rate, gprn
Loads are sometimes expressed in "Mbh". One Mbh is equivalent to 1000 Btu per hour; therefore Mbh = 0.5 TD X G
(19)
Plow rates usually are based on temperature drops of 10 to 20 &g F; consequently, it is useful to remember that
1gprn cooled 10 deg F gives up 5000 Btu/hr 1gprn cooled U)deg F gives up 10,000 Btu/hr
Ib) REVERSE RETURN SYSTEM
Fig.
IS .
Hot-water heating systems
is sometimes installed and the equipment i6 cross-connected. Stand:by pumps are not provided for single-pipe systems since they are relatively small (usually between 40 and 80 gprn); instead, complete pump and motor assemblies are carried as spares. It is essential to eliminate air and uncond*nsed eases from closed-circulation water systems because they impair heat transfer, retard circulation, cause pump cavita tioh, and create noise. Air must be purged from a system when it is initially filled; also, air enters with the makeup water and gases are released when the water is heated. Equipment is arranged so that the pump discharges to the converter (Fig. 14), which in turn discharges into an air separator. The latter removes both air introduced with makeup water and that released during the heating process. A water velocity of at least 1 fps is required to keep entrained air and gases moving, and thus prevent air pockets. Velocities in mains are sufficiently high; however. heating unit loads are often so small that minimum velocity, racer than load, determines the design water quantity. It is recommended that at least 1gprn per aii. heater and 0.5 gprn per convector be provided to assure adequate velocity in branches (usually of 0.5-in. pipe size) and units. Air can be confined to the mains; where the velocity is highest, by connecting supply branches to the bottom, or at least the lower half, of mains as indicated in Fig. 16. Accordingly, it is of advantage to keep mains as high as possible. Adequate and easily accessible means for venting air
-
The temperature drop selected for a particular system depends on the magnitude of the heating load. Of importance is the allowable pressure loss through reheaters. Large tempekature drops reduee flow rates and pressure drops, but heat transfer is poor at low water velocities. Accordingly, a compromise must be made. Considerable judgment is required to "juggle" the various design variables so as to achieve the optimum design; to be sure, the design of a shipboard hot-water heating system is not accomplished on the first attempt. The size and number of flow fittings required for individual circuits (shunts) of single-pipe systems (Figs. 15 and 16) depend on their design and the manufacturer's recommendations. Loads are seldom large enough to justify branch sizes over H in. When an appreciable number of larger sizes are indicated, the use of higher temperature drops and/or higher velocities in mains should be considered. Piping arrangements depend largely on the arrangement of spaces served. Where practical, returns from zone circuits (loops) are run independently to the pump suction manifold (Fig. 14); thus,- thennometen (for checking performance) and zone balancing valves are readily aecessible. The relationship of the spaces served controls the number of zone circuits. Where three decks are served, central station equipment, Fig. 14, is preferably located near the center of the middle deck, with a zone circuit serving each deck. The allocation of spaces to zones of single-pipe systems must be such that adequate capacity is carried in each loop to generate the required head at flow fittings; it may be necessary to break up the loops on one or more decks to reduce the pumpiag head. A separate heating system is not required for each air-
740
ENVIRONMENTAL CONTROL
MARINE ENGINEERING
conditioning system; however, to simplify trouble-shooting, only one heating system should serve a particular air handling system. Also, where practical, each zone circuit should serve only a single air-conditioning system. Expansion tanks may be of the open or closed type. The water temperature in open systems is leas than 212 I?; for closed systems, the use of 220 to 240 P water is common. Air elimination and the associated corroaion become greater problems as the temperature increases; thus low water temperatures are preferred. Obviously, the less makeup water used the better, since this minimizes corrosion. sediment contamination, and uncondensable gases ih the system. Converters are selected, from manufacturers' data, for the heating load required plus about 50 percent excess capacity to allow for start-up and scaling. The water pressure drop through converters is seldom critical. Circulating pumps used in hot-water systems are always of the centrifugal type. Generally, mechanical seals are provided, suitable for operation well above the design water temperature. Pump selections should be made from certified performance curves, not tables. The pump selected must not have an excessively flat capacity-head curve. Also, the point of operation should be at or slightly to the left of the maximum efficiency point. The operating point on the capacity-head curve is usually specified at least 10 percent below the peak. Such a selection: (a) provides a margin for errors in estimating pumping head, (b) accommodates changes in system rgquirements (head and volume) occurring during or after construction, and (c) minimizes the possibilitv of motor overload. cocks or valves are fitted in each zone circuit of a system to balance one against the other, Fig. 14. Balancing by temperature drop, using permanent thermometers, is most practical; thermometers also serve to check operation. Piping of hot-water heating systems is arranged to facilitate air removal, venting, and draining. Piping installed to follow the sheer and camber of the ship usually provides adequate pitch and suitable high points for venting. Mains are run as straight as possible to minimize high points (air pockets) requiring automatic vents. Hot-water piping is normally sized for a maxi(where d is the pipe mum velocity, fps, equal to inside diameter, in.), but not in excess of 6 fps. 4.3 Chilled-Water Systems. A typical coolingcoil piping hookup is shown by F5g. 17. Chilled-water systems are designed for direct return since there are relatively few units to balance. Generally, 42 F is considered to be the lowest economical chilled-water temperature; 45 F is specified for naval construction. Water quantities, generally, are based on temperature risos from 8 to 10 deg F, although temperature rises up to 12 deg F have been used for large system (pmsenger ships). Pumping heads of chilled-water systems are estimated the same as for two-pipe hot-water systems. Higher maximum water velocities are used because noise is leas
74 1
critical; also, the heat removal per gpm is smaller (the temperature rise much smaller). Chilled-water piping is normally designed for a velocity, fps, equal to 5 a (where d is the pipe inside diameter, in.), but not in excess of 10 fps. It is normal practice to circulate water continuously through chilled-water systems during cold weather; this tends to overcome the possibility of freezing due to stratification of preheated air. While not recommended, antifreeze liquids have been used. It may be noted that the noncorrosive brine used for other refrigerated services on ship may also be used for air-conditioning purposes. 4.4
2"X 2"COPPER DRAIN TROUGH
Combination Chilled and Hot-Water Systems.
Class E air-conditioning systems, Fig. 7, require twopipe, reverse-return water systems. Facilities for air removal are the same as provided for water heating systems. In general, induction units are bulkhead mounted; thus all p a h (valves, controls, vents, lint screens, flushing by-pass, etc.) are readily accessible for service, maintenance, and repair. The chilled water is first fed through the primary air coil and then is fed, as "secondary water," to the induction units. On a cargo ship, a single pump often handles both semices. The total temperature rise is about 15 deg P (8 deg P primary and 7 deg 1' secondary) ; thus, while dditional pumping head is required (primary and secondary coils are in series), the reduction in water quantity (about 1.6 gpm/ ton of refrigeration) keeps pump horsepower reasonable. On passenger ships where systems are large, the return from the primary water system is metered automatically, as required, into secondary systems serving the induction units. Piping w i n g primary cooling coils is designed to the same criteria as used with chilled-water systems; piping of secondary circuits use the same criteria as water heating systems. Combination water systems serving room fan-coil units are designed the same as Class E systems, when the primary (replenishment) air is dehumidified and cooled. When primary air is only preheated, the system is designed the same as for chilled water since fan-coil units accommodate the entire cooling load, including that of primary air. Drains from induction unit and mom fan-coil unit drain pans usually are N in. IPS. Where a number of 'units are connected to-a common &air?, the size is increased, depending on the number of units served. When practical, these drains are terminated in the corner of shower stalls, about 6 in. above the bottom. Drains direct t o weather are not desired and fortunately are seldom necessary. 4.5 Refrigeration Piping Systems. figures 18 and 19 delineate typical refrigeration systems. The usual balance between power requirements and installation costs applies to such systems, since excessive suction and discharge line losses (see Fig. 3) reduce compressor capacity and increase power requirements. All compressors lose some lubricating oil to the re-
FLOW METER
SUPPLY GLOBE VALVE NOTE: CONNECTIONS TO DRAIN PANS ' I N 0 1 SHOWN1 TO BE I 114" MINIMUM
STRAlNER BLOW OFF
e
Fie. 17
\
a
--
Typical hvo-high d i n g coil piping hook-up
frigerant vapor during the compression process. result, in halogen compound refrigeration systemsAsthea piping must be carefully designed so as to continuously return the lost oil to the compressor crankcase. In parts of the refrigeration cycle, the oil is a mist and the refrigerant is a vapor; here the refrigerant's velocity must be suEciently high to keep entrained oil moving along with it. Often, a riser sized for adequate vapor velocity at minimum load may have a prohibitive pressure drop at maximum load. In such cases a double riser, as shown in Fig. 19(b), ia required. At low capacities, oil collects in the trap and seals off the large riser; thus, the refrigerant vapor is forced to travel up the small riser at high velocity and thus carry oil with it. When the vapor flow increases sufficiently, the oil seal breaks and both risers are utilized. The small riser is sized for minimum design capacity, and the large riser is sized so that the velocity through both is sufficiently high a t maximum load. Figures 18 and 19 incorporate design features required to assure oil return. Liquid refrigeration leaving the condenser is usually
subcooled 2 to 5 deg F. This is advantageous because the friction and loss in static head between the condenser and expansion valve reduce the pressure of the refrigerant; thus, unless subcooled, the refrigerant will flash into a gas to compensate for the pressure drop. Liquid suction heat interchangers are also used to subcool liquid; they are always provided in ship's stores and cargo systems. Locating heat interchangers at receivers minimizes liquid line sizea. Also, they protect compressors by evaporating any liquid mixed in the refrigerant suction vapor. (For complete information regarding refrigeration piping design, including pipe sizing, pressure drop calculations, and subc~oolingrequirements, see references [2,4].) Accessories, valves, controls, and safety devices are shown in Figs. 18 and 19. A comprehensive discussion of these devices (types, operation, applications, details, etc.) and various types of oil separators (not shown) is given in references [2, 281. Materials comprising accessories and valves must be suitable for the refrigerant handled. References [26, ?9,30,31.] delineate merchant
MARINE ENGINEERING
ENVIRONMENTAL CON'TROL SOLENOID VALVE CONTROL SW. BULB BURSTING DISC
(0) Arrangement of comportments
TO COMPARTMENTS 55 F AND ABOVE
Y I I
COMPRESSOR
TO MOTOR CONTROLLER
(c) Condensing unit oirembly
/
-
HOT GAS DEFROSTING
f-
(b) Interconnecting piping
--. .-. J NOTES
1. Refrigerant liquid lines ond copillory tubing for dial (remotareading) thennometen and solenoid valve control switches ore run to refrigeront control panels located by the shipyard.
2.
Meat ond flsh room i s similor except thot two independently controlled refrigerant circuits are provided.
3.
Fruit and vegetable room is similar except thot the cold diffuser hastwo independentlyconholledrefrigerant circuits.
4. Where possible, all freon piping i s bent in lieu of using elbow, to 'n k .i reduce friction and the number of wldered@
5.
All suction line branches are looped into the top of the suction main to prevent oil accumulation in lines not in sewice.
6. Hanger and aupports are instolled w cis to pennit free expamion and contraction between anchorages and to minim'ue vibration.
7. Refrigerant auction lines between the compreuor and refrberated space are insulated with brine thickness insulation. Liquid liner between the heat interchanger and refrigerated space are insulated with ice water thickness. Fig. 18 Ship's stores refrigcwation system diagram [Carricwl
c
POWER
TO SHIP'S
FROZEN FOOD B ICE CREAM RM. INOTE 21 0 F
--
DAIRY (55r1 AND THAW ROOM [&OF1 NOTE 5
(dl Typical pipe coil hook-up
(a) Typical cold diffuser hook-up
Fig. 18 (cd'd)
'
I
ENVIRONMENTAL CONTROL
i
MARINE ENGINEERING
and naval construction requirements for installation, materials, tests, safety devices, and spare parts. The last item, spare parts, is particularly important since the availability of the equipment may depend upon them. Condenser water regulating valves, Fig. 18(c), may be readjusted, periodically, to provide a condensing pressure which is just high enough to balance the operating load. Thus, the most economical operation will be obtained, and compressor repairs will be minimized. This is of particular significance for dual and low-temperature cargo systems which employ compression ratios a t the upper limit of the compressor rating. A dryer is provided in the liquid line near the receiver outlet, Fig. 18(c), or the condenser of systems without receivers. Some dryers include filters, and thus perform a dual function. The type of filter included eliminates much smaller particles than mesh strainerv. With the arrangement in Fig. 18(c), the dryer only is used when (i) charging the system with refrigerant, (ii) adding refrigerant to compensate for the loss through leaks, or (iii) when the presence of moisture in the system is detected by "freeze-up" a t the expansion valve. On naval construction, however, full-flow filter-dryers are provided. Since excessive moisture can cause breakdovn of motor insulation, possibly burnout, and ultimately contaminate the entire refrigeration system, full-flow filter-dryers are provided for hermetic systems. The relief valve in the equalizing line between the condenser and receiver, Fig. 18(c), prevents excessive pressure build-up in the receiver during shutdown when the valves are closed. The condenser relief valve and rupture disk protect the system in case of fire; note that these do not vent to the machinery space. All lines should be plumb and straight, except horizontal suction lines, discharge lines, and those between condensers and receivers which should pitch in the direction of flow. Valves in vapor lines are installed with stems horizontal to avoid pockets. Unions and flanged joints should be minimized, i.e., provided only where necessary for disconnecting equipment, controls, etc. Insulated stuffing tubes are fitted a t penetrations of watertight structures to thermally isolate the cold lines. Where watertightness is not essential, oversize sleeves (properly caulked and ratproofed) are provided, and the insulation and vapor seal are continuous. Sleeves in decks are of ample height to act as coamings; extra care is taken to seal the joint between the pipe and sleeve to
COLD DIFFUSER
10) INTERCONNECTING PIPING
prevent entrance of moisture, which can form ice and crush the pipe. 4.6 Brine (Secondary) Cooling Systems. Brine (secondary) cooling systems are preferred over direct expansion cooling by many operators because of improved reliability, simplicity of operation, greater flexibility, better temperature control, ?nd superior cargo turnout. The small amount of primary refrigerant piping required with brine systems significantly minimizes refrigerant leakaee, oil return problems, and facilities for the return of oil. The initial cost of brine cooling is greater than that for a direct expansion system. The additional heat exchanger, required to produce the necessary secondary cooling, tends to increase power requirements; however, this disadvantage may be overcome by more efficient operation. Generally, brine cooling may be used most advantageously where (a) there are a large number crf refrigerated compartments, (b) compartments served are remote from the refrigeration machinery space, (c) the diversity of simultaneous, individual compartment temperatures is large, and (d) maximum humidity is required to minimize moisture removal from products. A strong conviction of the owner is a t least one, if not the, deciding factor. Note that the brine may also be used for airconditioning purposes. Two-temperature closed brine systems are preferred because they provide optimum control of the supply brine temperature and minimize air leakage which causes air binding and accelerates corrasion [2]. On some ships a primary circulating system provides low-temperature brine to a number of secondary circulating systems, and each secondary system serves one compartment. The primary brine supply to the secondary system is modulated by a space thermostat. Closed systems are preferred to the open type because the cost of pumping facilities is less, and the corrosive effect of aerated brine is greatly reduced. Also, glycols are organic and with open systems some inhibiter is lost to the atmosphere, thus necessitating periodic replacement. References [4,27] contain discussions of the properties of brines and methods of calculating system pumping requirements. All materials in the system, including piping, flange gaskets, valve seats and packing,.pump seals, and other specialitieg must be compatible with the particular brine handled to avoid corrosion. Potential problems in this regard require careful analysis.
Section 5 Air Ha~~dling System Design '
LOCATE BULB I N RETURN AIR STREAM
L-
-
-2
&
TO SHIPS POWER
IblTYPICAL CARGO COLD DIFFUSER HOOK-UP
Fig. 19 Cargo compartment refrigarcrtion system diagram [Carrier]
5.1 Flow and Fan Pressures. The basic relationship between airflow and the pressure (head) required to cause the flow (with standard air density of 0.075 lb/ cu ft) can be expressed as:
h = (V/4005)=
(20)
745
,
where h = velocity pressure, in. of water = velocity of air, fpm There are three interdependent pressures common to air handling systems; static, velocity, and total. Static
MARINEENGINEERING
ENVIRONMENTAL CONTROL
SYSTEM P-V CURVE OPERATING POINT
VOLUME
fig- 21
5.2 System and Pressure Characteristics. The relationship between the system Pressures at any two rates of airflow can be expremed as:
P1/P2 =
references [27, 33, 341.
(Q1/Q2I2
p, = pressure required for airflow &I p2 , pressure required for airflow &z A system pressure characteristic curve, Fig- 21, thus can be constructed once the preasure for one capacity is known, using an arbitrarily selected series of capacities.
DUCT AREA AIR VELOCITY YELOCITY PRESSURE. ST~TICPRESSURE TOTAL PRESSURE
I FTZ 4000 ~ p n
DUCT AREA AIR VELOCITY
l.OOn HZo 1.00" H ~ O 2.00n H ~ O
VELOCITY PRESSURE STATIC PRESSURE TOTAL PRESSURE
0.251' 1.75" Hz0
2.00" Hz0
la) DUCT PRESSURES
F;~.20 glv,~afion of duct and f an presures
lbl FAN PRESSURES
Determination of system operating point
FAN P-V CURVE
MARINE ENGINEERING
ENVIRONMENTAL CONTROL
@)
MARINE ENGINEERING
750 RADIUS RAT10 IRR)= R/W
Fig. 26 Location of splitten in elbows of rectangular duch IMffalo Forgel Fig. 24 R e c t a ~ l a in-run r duct elbow bawa IBufialo Fwd
RADIUS RATIO lRRl
75 1
ENVIRONMENTAL CONTROL
thus, a 45deg elbow is assigned a loss equal to M of a 90-deg elbow. The loss itl an elbow consists of (d) ffiction loss and (b) dynamic loss due to compression of air a t the outside (heel) of the bend. If sufficient straight duct follows, a regain occurs; however, if the air leaving the elbow discharges to atmosphere, no regain occurs and the loss is greater. A straight length of discharge duct of a t least 4 diameters, or 4 times the duct dimension in the plane of the bend, is required before "in run" elbow ldms apply. The straight length of duct required can be reduced by installing splitters or turning vanes; the maximum splitter or vane spacing then becomes the criterion rather than the duct size. Losses for elbows a t the end of a run, Fig. 27, include one velocity head for the discharge to atmosphere. To find the loss for an elbow at the end of a duct with concentric splitter vanes: (i) deduct one velocity head from the loss shown, (ii) multiply the remainder by the ratio of losses in Fig. 24 for elbows in run with and without splitters, and (iii) add one velocity head. To find the loss for an elbow of less than 90 deg a t the end of a run: (i) deduct one velocity head from the loss in Fig. 27, multiply the remainder add by the ratio of the elbow's angle and 90 deg, and (3) one velocity head. Loss* of elbows in run with smallradius turning vanes may be taken as 0.25 and 0.35 for double and single-thickness types, respectively; for elbows a t the end of run, add one velocity head to these values. The pressure losses due to divided flow in round supply ducts (branches and mains) are given in Fig. 28; lacking better data, Fig. 25 may also be used for round or rectangular branches with rectangular mains. Losses in divided-flow exhaust fittings are given in Fig. 29. Figures 30 and 31 show the losses due to velocity or area changes, i.e., abrupt expansion, abrupt contraction, and gradual expansion, in s run of duct. When a connecting duct is less than four diameters downstream of an abrupt expansion, regain is lost and the loss approximates that for an open-ended duct; namely, one small duct velocity head. Abrupt contractions create a vena-con-
tracta st the entrance to the mhder duct; unless there is straight duct following the contraction (for regaih) the loss equah that of a square-edge orifice a t the end of a duct, Fig. 33(a). This point deserves emphasis as it is frequently overlooked in the mechanical execution of ealculatiohs. A loss of 0.05 small duct v~locityhead is sufficient for circular concentric gradual contractions up to an included angle of 45 deg. The loss allowed for symmetric transitions from,one rectangular size to another of equal area is 0.15h+ Where transitions are unsymmetric or one-sided, the &wed loss should correspond to that for the largest angle on one side. This also appliw to the gradual expansion of rectangular ducts.
VELOCITY RATIO-IV DOWNSTREAM- MAIN OR BRANCH / V UPSTREAM1
Fig. 26 Lass-
in divided-flow flltings [ASHRAa
RADIUS WT10. R / W OR R/D
~ g27 .
1 .I0 0
0.1
0.2
I
I
I
I
I
0.S
0.4
0.5
0.6
0.7
CURYE RATIO (CRI
Loues in elbow dischorgiw to ahnosphere
.I
4
~ g 25 . Round in-run duct elbow ImeS IBvffalom e ]
-2
.S
.4 .5 .6 .7 LOSS COEFFICIENT,K, IBRANCH AN0 MAIN1
Fig. 29 L o w s in divided-flow exhaust fltting~
.8
.9
1.0
ENVIRONMENTAL CONTROL MARINE ENGINEERING LOSS COEFFICIENT,K2. I VELOCITY PRESSURE @ AP
754
MARINE ENGINEERING Table 18
Terminal losses
Table 1 9 Terminal and Equipment Air Velocities
.-Loss Com~rcmm,aKClear 0.5 in. Wire' Insectb Mesh Screen Double gooseneck, entrance Fig.35(e)] discharge Single gooseneck, entrance Fig. 35(f)] discharge Mushroom, [Fig. 35(c)]
entrance discharge
ITEM
Louvers. ...................... Mrnhrooms.................... Goosenecks. ................... Heaters (steam). ............... Heatera (hot water). ........... C?o!ing coils.. ................. Dmdioml terminals. .......... Grilles and regidera. ........... Bu ly registers (ventilation). ... ~ g 8 e r a...................... . Ratproof Screens X 35 in. mesh). .......... Insect screen (60 percent clew area). ..................
(x
-
ENVIRONMENTAL CONTROL
3.5
4.0
VELOC~TY RANGE(fpm) 1000 to 1500 (free area) 800 t o 1200 (throat area) 1000 to 1500 ( b t area) 800 to 1200 (face area) 4d0 to 600 (face area) 350 to 500 (face area) 2000 to 3000 (throat area) 800 to 1000 (free area) 760 to 1000 (free area) 500 to 900 (neck area) 500 t o 800 (gross area)
300 t o 500 (gross area)
Weather louverc-d
Losses for the more common terminals and fittings are given in Figs. 32 and 33 and Table 18. Again it is presupposed that there is ample straight duct at entrances to obtain reexpansion of the vena-contracts. Catalogs Door louvern for exhaust registers and grilles usually show only the static pressure loss; therefore, additional data are required to determine the total pmsure loss and avoid misapplication. Catalog losses for registers, diffusers, Hole in bulkhead and other purchased supply terminals usually represent only the pressure imposed by applying the terminal to an open-ended duct; to approximate the total pressure loss, add one h, (based on the connecting duct size) to Short duct in bulkhead the loss given in the catalog. Terminals for merchant (re-entrant) construction are usually sized on a velocity basis (see Table 19). Weather terminals are sized for low velocities where possible, because weatherproof features entail high presOpen end duct sure losses. The data given in Table 18 for 0.5-in. wire mesh and insect screen allow for dirt, lint, and painting, 0.90 1.4 2.0 entrance dAA discharge 1.10 1.6 2.1 which appreciably reduce the effective area. Pressure -7 losses through equipment (registers, diffusers, flexible Duct terminating in plate ducts, cooling coils, and heaters) are preferably obtained from the manufacturers%ertified data. 0.50 ' entrance 5.6 Connections to Equipment. The performance 1.10 discharge 3 . of equipment, including pressure drop, is based on tests performed with a uniform inlet (face) velocity. IncorConver&ng taper rectly designed connecting ducts can invalidate such performance data. The equipment's resistance cannot be depended upon to equaliie the inlet velocity. Unfortunately, axial-flow fan literature seldom emphasizes this point. On shipboaxd, a plgnum with a bellmouth Bellmouth (in lieu of an elbow) is frequently fitted at the fan's inlet. !.zs$ entrance 0.5 - Where elbows are fitted, they are of the small-radius vane type, and connected to the fan by concentric, gradual transformers. Integral fixed inlet vanes and bellmouths NOTES reduce the effect of poor approach conditions; this is a a Losses are based on the aree a t "A" or "D". b Screen losses are baaed on the gross area velocity head, Lo' feature of certain high-quality centrifugal fans 1341. The loss for 0.5in. wire mesh (75% c!ear area) is 0.50 A,; the At least one diameter of straight duct is recommended loss for insect screen (609' clear area) is 1.1 L ;and the loss for at fan discharges before any elbow, split, offset, or transiboth 0.5-in. wire mesh an8 insect screen is 2.0 Lo. Includes one velocity head A) for discharge to plenum. deduct tion. Where divided-flow volumes at a fan's outlet vary 0.5 h when the terminal is .irectly connected to a straight duct. during operation, an expanding duct may be used at the Typically, the totd flow area, A, is 60% 6f the gross area, B. fan outlet which discharges at low velocity (1000 fpm or leas) into a plenum (see Fig. 37). A similar plenum Typically, the total flow area, A, is 50% of the grow area, B.
s t
#
6
@
I
MARINE ENGINEERING 3000
r FILTER 1.31 32x7 92 FAN, 3.25"T.P.
AIRLIFT
J
1l-
COOLING COIL 1.61
PREHEATER 1.2)
Fig. 34
Typical duct system
may be used for air-conditioning systems with several reheaters. 5.7 Fan Total Pressure Calculations. Table 20 contains calculations for the fan total pressure requirement of the typical duct system shown in Fig. 34. The various data discussed in Section 5.5 are used in Table 20. Note that sizes which are based on assumed velocities can be entered directly, but those based on an equal friction loss must be determined from Fig. 23 after an equivalent round size has been selected from Fig. 22. No loss is assigned to the split, a t point.B, because there is no velocity change. Path A-D is assumed to be the longest run (incur the highest loss) in the system. Losses for fittings are determined by ayplying appropriate K and
h, values in equation (22). Losses for straight ducts are determined by multiplying the Loss/100 ft column by the duct length/100. Only the velocity and h, associated with the applicable K factor are shown for fittings involving velocity changes. The column "Pres. Avail. " (Pressure Available) is provided to facilitate the design of submains (those not forming part of the path of highest loss) and branches. It may be noted that 1.804 is available for the 3000-cfm submain from point B to the end, including return losses. Similarly, 1.243 in. is available for the 200-cfm branch from point C to the end. The column "Loss Source" is usually not provided; this column is included in Table 20 for illustrative purposes only.
Section 6 Air Handling Equipment 6.1 Fans. Most marine centrifugal fans are of the backward-curve blade type; generally only small fans (less than 1000 cfm) have forward-curve blades. The principal advantage of a backward-curve fan is that both maximum brake horsepower and fan-set efficiency (at any speed) occur very close to the same capacity point; thus, the motor is selected for a horsepower corresponding to the highest efficiency point. Of equal importance is the fact that if the actual system pressure loss is more or less than estimated, the mator selected for a specific fan speed will not overload. Forward-curve fans have a rising bhp characteristic as the capacity approaches free delivery; thus, a motor selected for maximum fan-set efficiency can be overloaded if the actual system opera& ing pressure is less than estimated. It is common practice to oversize motors driving forward-curve fans; often
small motors are sized to accommodate operation a t free delivery. Beltdrives have advantages. Where the fan speed is less than 1750 rpm, smaller (higher speed) and. cheaper motors can be used. Also, the adjustability of belt drives facilitates system balancing; this is particularly significant for air conditioning, since only very limited overdelivery can be tolerated. Additionally, quieter and more efficient operation can be obtained by reducing the fan speed vice adding resistance iwthe system. Because of the foregoing, belt drives are usually specified for fans of air-conditioning systems and !ens of ventilation (supply and exhaust) systems which either must be in balance with air conditioning or serve other areas requiring quiet operation. Except for those in factory-fabricated units (Fig. 37))
ENVIRONMENTAL CONTROL
centrifugal fans usually are of the single-inlet, singlewidth type, rather than double-inlet, double-width type. Drive arrangements with bearings outside the airstream are specified, except where fans are very large. Axial-flow fans, sometimes called vane-axial fans, have the advantage of compactness, but this is not always significant in the smaller sizes. Also, since motors must be watertight and construction tolerances are smaller, axial-flow fans sometimes cost more than equivalent centrifugal fans. Axial-flow fans have a disadvantage in that motor repair is relatively difficult because of inaccessibility. The static pressure-volume curve of axial-flow fans generally has a dip a t the left of the peak. To ensure stable operation, specifications for marine applications require that fans be selected to operate at a point on the static pressure-volume curve well below, and to the right of, the peak. Motors are selected with a substantial power margin, because of the rising brake horsepower requirements and the reduction in cooling air for the motor as zero delivery (tight shutoff) is approached. Particular care is required when selecting fans for use where quiet operation is essential. While standard code tests indicate that axial-flow fans can be at least a quiet as centrifugal fans, the former are more affected I! y poor duct connections, particularly on the inlet side. Accordingly, bellmouths are fitted a t the fan inlet unless the inlet duct is straight and uniform for some distance ahead of axial-flow fans. Many designers of merchant construction use axial-flow fans only where quiet operation is not essential, although on naval ships practically all fans are of this type. Propeller fans are "simplified" axial-flow fans ; they are not built to the close tolerances required for axialflow fans. Usually these fans are used on systems having static pressures of 0.5 in. of water, or less. Where noise is not important (resistor and motor generator rooms, etc.), static pressures up to 0.75 in. are used. Often, the static pressure-volume curve is such that operation above 0.5 in. is unstable. Of the various relationships dealing with fan performance, the three most essential in the design of airhandling systems are: Qi = (Nl/Nz)Qz (23) Pi = (Ni/Nz)'Yz (24) HP1 = (NiIN2)aHPZ (25) Q, P, H P , and N denote fan delivery, pressure, horsepower, and speed respectively; and subscripts 1 and 2 denote two different operating conditions. There are two commonly used expressions for fan efficiency [32]. One is the mechanical (total) efficiency, ME, expressed as: ME = CFM T P 6356 BHP tind the other is the static efficiency, SE, expressed as: MESP SE=-= TP
CFM SP 6356 B H P
(27)
where: CFM = air volume flow rate at inlet, cfm B H P = power output of drive, hp SP = fan static pressure, in. of water T P = fan total pressure, in. of water Comparisons of performance should be based on mechanical efficiency. ' On merchant construction, flexible connections are provided betwepn fans and ducts to isolate fan vibration from the ducts and to accommodate slight misalignments. Such connections also prevent distortion of fan casings, which can occur when directly connected to heavy coamings, ducts or trunks. This is particularly important for axial-flow fans, because of their small wheel-tip clearances. On naval construction, flexible connections are provided for those fans supported on resilient mounts; both connections and mounts are of a special shockproof design. Fan supports may be rigid or resilient. Where rigid, the fan base is bolted directly to angle foundations which are welded to the heaviest structure available, with the exception of bulkheads adjacent to living spaces or others requiring low noise level, which are avoided. Fans within the living areas of merchant ships are customarily supported on resilient mounts. The fan and motor of belt-driven centrifugal fans are mounted on a common base to preserve alignment (Fig. 39). The base is supported on resilient mounts which in turn are bolted to angles welded to the ship's structure. Flexible duct connections are always fitted to fans on resilient mounts. 6.2 Filters. Roll-type air filters, with a disposable media, are built especially for marine service. Their performance is better than the washable, viscous type. Also, they can be more easily serviced; i.e., they do not require cleaning facilities. Roll filters are selected from the manufacturers' catalogs; about 6 ft-9 in. nominal height is the largest which can be accommodated. Where possible, all filters on a particular ship are selected with the same width. Panel-type air filters with a dry media, similar to that of roll filters, may be used in very small systems where the use of the roll type is not practical or economical. Washable viscous-type air filters are sometimes used on merchant construction. The entire filter cell is fabricated of bronze and/or copper; all steel parts of the holding frames are hot-dipped galvanized (after fabrication). The capacity of a standard 20 by 20 in. unit is 800 to 1050 cfm. Grease filters are selected for about 600 cfm/20 by 20 in. unit. Their construction is the same as the washable, viscous type. Air filters should be so installed that they are protected from the weather and are easily accessible to encourage proper servicing by operating personnel. Filter cleaning stations are provided to facilitate the maintenance of panel air filters. Adequate operating and clqaning instructions, and diagrams showing the location bf filters (posted a t proper locations), are, an essential part of a /
?-
MARINE ENGINEERING
*
good installation. No definite interval of time can be recommended for servicing filters, because the necessity for cleaning or renewing the media depends on numerous individual factors; e.g., port conditions and types of cargo handled. Filters should always be examined after dusty bulk cargos have been handled. Pressure gages, or indicators, are provided to determine the necessity for cleaning or media replacement. On most passenger
objectionable both from the viewpoint of the damage it can cause directly and the deleterious effect it has on the insulation treatment. For these reasons insulation must be properly installed and sealed to prevent moisture penetration. While metal foils are inherently impermeable, their fragility, as well as the difficulties encountered sealing joints, leave their practicability as a vapor (moisture) barrier open to question. Many
6.3 Ducts and Trunks. On most ships main ducts are located in passageways and branches are run to each space. Thorough coordination with piping, wireways, and other hotel services is essential to produce an efficient design and still maintain satisfactory headroom and accessibility. Ducts are usually constructed independent of the surrounding structure. However, where advantageous and where the size is large enough to permit access to the interior for inspection and maintenance, ducts may be "built-inn as trunks using existing bulkheads, decks, beams, or girders to form one or more sides of the air passage. Built-in (trunk) construction is not advocated for air conditioning and other ducts requiring insulation and/or thermal isolation. Kingposts sometimes are utilized for ventilating purposes; this has disadvantages, however, because condensation occurring inside the kingposts can corrode the structure and also damage the cargo. Every effort is made to keep ducts from penetrating watertight and firescreen bulkheads, and the regulatory bodies prohibit their penetration except in very special cases [26, 37, 381. There is seldom a necessity for penetrating such barriers on merchant ships. Fire dampers are also required at penetrations of fire-protected structurw and at certain other locations. Ordinary sheet metal ducts are referred to as "nontight" ducts; this differentiation is necessary to avoid confusion with watertight and gastight structures, which are subject to pressure tests. On merchant construction, most "nontight" ducts are constructed of sheet steel. In order to withstand abuse, ducts in cargo spaces are made of steel plate. Watertight ducts are constructed of the same material as the associated structure; they are made strong enough to withstand the applicable test head requirements [26]. (See references [9, 101 for additional information regarding ducts and trunks.) The cost of air-handling systems can be materially reduced by using factory-fabricated ducts and fittings. These are ordered the same as piping system components; i.e., pipe (duct), couplings, tees, elbows, reducers, etc. The ducts are either smooth tubing or spiral-wound, grooved-seam conduit [4]. The latter is somewhat lighter since the spiral seam affords stiffening; however, the tubing is heavier and confines noise better. 6.4 Duct Insulation. Heat transmission to and from ducts is significant in the design of environmental control systems. Additionally, condensation forms on the surfaces of ducts having a temperature below the ambient air dew point. Condensation can be highly
fie-retardant compounds, reinforced with lagging, have proven most practical. Because of the condensation factor, air-conditioning ducts present a much more difficult insulation problem than those for heating. There are a number of cellular insulation materials which do not absorb an appreciable quantity of moisture; however, those which have the most desirable characteristics from this viewpoint are not incombustible. Insulation material selections are restricted to fibrous glass and mineral wool, which require separate lagging and sealing to prevent vapor penetration. Flanges and all equipment are completely insulated and vapor-sealed. Also, supports and structural penetrations are thermally isolated from the ducts. Any leakage of moisture into the insulation will reduce its effectiveness;even a small defect is serious, because its deleteriouseffect willbecomeprogressively larger. Thus, exceptionally high-quality workmanship is wsential. 6.5 Conditioned-Air Distribution. Shipboard conditioned-air distribution practice is based primarily on experience and tests in mock-ups of state rooms and other critical spaces. Architectural considerations, important for one reason or another, must not be permitted to overrule functional air-distribution requirements. Acceptable air motion is difficult to define; accordingly, specifications regarding air motion should also stipulate the type of measuring instrument to be used. Most merchant ship specifications limit the air velocity at the head end of berths to 35 fpm and elsewhere in the zone of occupancy to 50 fpm, with measurements being made using an Anemotherm. Since the lack of air motion can cause both a sensation of discomfort and stratification (uneven temperatures), a maximum ambient temperature differential of 2 deg F is also stipulated. High entering air temperature differentials (25 to 35 deg F) can cause sweat to form on terminals and their mounting surfaces, particularly during qulldown. Accordingly, most specifications for air-conditioning supply terminals require mock-up tests which demonstrate that sweating will not occur, in addition to demonstrating satisfactory air motion performance. One criterion that has been used is that visible condensation is not permitted to form on the terminals when air is supplied a t the actual design temperature and the room ambient dew point is at least 10 deg F above that corresponding to the inside design conditions. Diffusers provide satisfactory performance for practically all merchant ship air-conditioning applications. Experience shows, however, that the data provided in
ENVIRONMENTAL CONTROL
759
.
4- TRUNK
101
HOOD
f-- COAMJNG
~ b )COWL
('1 MUSHROOM
(dl ROBERTSON EXHAUSTER DECK
4 HOUSE SIDE-
A MINIMUM
INVERTED
(el DOUBLE GOOSENECK
If1SINGLE GOOSENECK
f Fig. 35 Typical weather terminals
DECK 181 AIRLIFT
+
MARlNE ENGINEERING t
returns are located away from sitting areas to prevent drafts on the feet and legs of occupants. An additional advantage of locating returns away from sitting areas is that system noise will not be as noticeable. 6.6 Ventilation Terminals and Dampers. Shipboard ventilation terminals are of either the "weather" or "interior" types, both of which must meet the ratproofing regulations [16]. With the exception of cowl ventilators and natural-exhaust ventilators, weather terminals, [a) ADJUSTABLE BLAST I ~ADJUSTABLE I BLAST Fig. 35, are designed and installed to minimize the effect (DRUM TYPE1 (TYPE E l of wind and ship's movement. Weather-excluding features invariably increase the pressure drop of terminals and, therefore, a compromise may be required. Where possible, terminals are located such that they are shielded from the weather by bulkheads, overhanging structures, and similar barriers. Supply and exhaust terminals are kept well separated in order to prevent contamination of the supply air. The air-lift type of weather terminal is generally preferred. Interior terminals, Fig. 36, are selected to suit their particular application. While controlled drafts are desirable for summer ventilation (evaporative cooling), ICI DIFFUSER [FIXED PATTERNI they are undesirable during heating operations. Therefore, supply, terminals in ventilated spaces are selected I ~ REGISTER I and so located that the occupants may control both the !DOUBLE DEFLECTION TYPE) direction and flow of air. Also, bracket fans are provided in living spaces and certain working spaces. Fig. 36 Typical interior terminals "Spot cooling" of personnel is essential in work areas and watch stations of machinery spaces, galleys, laundries, and other spaces containing large heat sources (including radiant heat). A continuous curtain of supply catalogs are generally not applicable to staterooms. air, discharged vertically downward from a slotted duct, Ceiling berths, soffits, lights, and deck-to-deck furnish- is sometimes employed in front of switchboards and ings complicate air distribution such that mock-up tests adjacent to turbogeneraton. This arrangement preare required to establish a satisfactory arrangement. vents water from entering through the ventilation system Directional pattern ceiling diffusers, Fig. 16, are advan- and being directed toward electrical equipment; this is a tageous, because airflow invariably must be restricted in mandatory requirement for all shipboard terminals. one or more directions.
ENVIRONMENTAL CONTROL
76 1
-
I
VERTICAL SUITS
l I I CARGO SPACE
VERT. SLOTS HERE ONLY IF w IS OVER 10'
.!inWIRE MESH
HORIZONTAL SLOT IDAHPER CLOSED I SECTION B-B
I 1
ENVIRONMENTAL CONTROL 762
MARINE ENGINEERING
763
corrosion-resistant steel, and other parts are hot-dip ing air is distributed over the top of the cargo block. galvanized after fabrication. Liberally sized portable Generally, when frozen cargo is carried, the dampers are plates are provided for inspection and renewing fusible set to provide more horizontal blow, since the colder air will fall through the cargo block; the vertical flow is the links. 6.7 Refrigerated Space Air-Distribution Equipment. minimum required for the transmission load through Air distribution is a vital part of the refrigeration process. the vertical boundaries and deck underfoot. Figure 38 shows an air distribution system suitable for The design of a cooling system entails a compromise between air volume and temperature split (room tempera- fruits and, particularly, cafgos of stemmed bananas. It ture minus supply air temperature), which are related as was developed prior to the introduction of modern methmay be noted from equation (3). There are numerous ods for proteqting and packaging the latter. This arschools of thought regarding the best method of air dis- rangement js' called a "duct bulkheadJJ or "air-duct tribution for a particular produce, and many patents bulkhead" system; many variations of this method of air have been issued relative to this subject. References distribution are patented. While in transit, the duct [2, 18, 22,391 are particularly instructive in this regard. bulkhead passages also provide access to the slide dampFigure 37 typifies the cooling facilities provided on e n (for adjustment), permit inspection of cargo, and cargo ships carrying "universal" refrigerated cargo. provide access to side ports and portable bulkhead Heating facilities and permanent hot-water defrosting sections (for removal). Systems on many foreign ships are included. The fan motor is two speed to permit low also include facilities for periodically reversing the flow air motion and reduce fan heat load during holding (nor- of air so as to achieve a uniform temperature during mal) operations. The drain pan section confines and downpull. Battens and grating facilitate the distribution of air disposes of residue from the defrosting process; sectionalized hinge coven allow ready access to the drain pan. through and around the cargo. Battens also provide The air distribution shown blankets the "hot" surfaces protection for sheathing. Battens are omitted in spaces with cold air and intercepts all transmitted heat; also, carrying stemmed bananas, since only underdeck air minimum and uniform air motion is obtained. Adjust- distribution is traditionally used; also, the battens may mcnt of the air distribution in the compartment is accom- damage the product. Gratings are usually made of plished by means of sliding dampen. The aidow can wood in cargo spaces and aluminum extrusions in ship's be modified to suit the requirements of any particular stores. They are made portable to facilitate removal cargo being carried. The "normalJ' setting of the damp- for cleaning and, in cargo spaces, to allow the use of an, which is established by the shipbuilder, directs about palletized loadings. It may be noted that the pallets 00 percent of the air vertically downward. The remain- used with forklift trucks function also as gratings.
SecEio~~7 Heating and Cooling Equipment 7.1 General. The advantages of hot-water heat- seldom level, return bends may trap condensate, causing ing systems were discussed in Section 4; most of these water hammer and, of greater importance for preheaters, dvantages apply equally to electric heating. Electric freezing. Accordingly, steam coils are preferably of the heating may be used whenever adequate generator ca- header type; i.e., a header is a t each end of straight tubes pacity is available. I n general, the power assigned to with no return bends. Heating elements of electric heaten are usually of the air-conditioning refrigeration may be used for heating. Electric heating is particularly advantageous for elec- tubular type and are monel sheathed. Fins, where retronic and other equipment spaces where leakage from quired, itre monel or corrosion-resistant aluminum alloy wot systems cannot be tolerated. Cost comparisons cast integral with the sheathing. Elements are replacebotween electric and wet heating are so complex that the able without removing the heaters from the ducts. nubject cannot be generalized. Since modulation of the Sometimes corrosion-resistant materials are used for caskouting effect is a significant cost factor, the quality of ings and flanges of smal1,keaters. porformance must be considered in a true comparison of Coil selections may be made from catalog data; howthe various forms of heating. Prior to the introduction ever, the coil construction (fin material, fin thickness, and of ~olid-statecontrols, it was prohibitively expensive to other heat-transfer element details) must be carefully provide a true modulating control of electric heating checked and compared with that specified. Also, final except for special applications. Without modulation, selections must be certified by the manufacturer. ellpacity variation must be accomplished by step control Most preheaters of air-conditioning systems have very of multiple-circuit heating elements. light, or even no, loads at design conditions. However, 7.2 Heating and Cooling Coils. Since a ship is some specifications require a 35 F leaving temperature
MARINE ENGINEERING
765
ENVIRONMENTAL CONTROL
with 100-percent outside air in order to protect the system. The air pressure drop is between 0.10 and 0.20 in. of water. Some specifications limit the combined air pressure drop of preheaters and reheaters in series to 0.50 in. of water. On merchant ships, steam heaters are installed with the tubes vertical if possible so as to minimize the possibility of the condensate freezing. Also, heaters are installed inside deckhouses; this permits better piping arrangements, protects the equipments from freezing, and facilitates maintenance and repair. Hot-water coils are of the serpentine type except where pressure limitations necessitate a header type of construction. The lower tube velocity associated with a header type is disadvantageous in that it adversely affects the transfer of heat. Cooling coils are selected with an even number of rows, usually 6 to 10, so that both piping connections can be made from the same end of the coil. Catalog selections should include a liberal margin; some specifications require at least 6 rows and 25-percent more rows than catalog data indicate. This protects against capacity reduction due to fouling the coil fins. Air pressure drops vary from 0.5 to 1.0 in. of water. 7.3 Direct Radiation and Unit Heaters. Since the capacity of steam heating equipment varies directly with pressure, the high-pressure steam available aboard ship offersthe opportunity of appreciable reductions in space, weight, and cost of the heating equipment. However, small quantities of medium-pressure steam are difficult to modulate (control). Steam working pressures up to 35 psig have been found to perform satisfactorily on merchant ships. Convectors of standardized designs are used for both merchant and naval construction. Fin-pipe elements are made of copper or steel with the latter being hot-dip galvanized after assembly. Fin-pipe elements have the advantages of being low and of diffusing the heating effect to achieve exceptional temperature equalization. Hot-water radiators used on many foreign ships are fabricated of embossed steel plates. They have a superior radiant heating ability combined with compactness, light weight, and simplicity of construction. Convectors and fin-pipe elements are usually selected
on the basis of laboratory testa rather than vendor's catalogs. A great variety of fin-pipe elements is available. Convectors and fin-pipe elements are installed on weather-exposed surfaces as near the deck as possible, Fig. 13. Elevating convectors decreases their capacity (the inlet temperature increases), and results in stratification due t p insufficient air movement. The "wall-hung" type, Fig. lg, facilitates cleaning and installation of piping. Heating elements of steam radiation equipmqat are pitched to facilitate drainage; allowances pre made for camber and sheer as well as for the list and trim of the ship at various conditions of loading. Unit heaters are most commonly used in steering gear rooms; large shop areas, and machinery spaces of diesel ships (to prevent freezing when the main machinery is shut down in port). The capacities listed in catalogs for unit heaters depend on the inlet steam pressure and the entering air temperature; the latter is assumed to be the same as that of the heated space. Where a number of unit heaters are required, they should be directed to circulate the air in the space; i.e., they should not oppose each other. Unit heaters should blow parallel to the shell and bulkheads; they should not blow heated air directly on cold surfaces because the transmission loss will be increased. They are installed about six feet above the deck or working platform. In machinery spaces of diesel ships, unit heaters are preferably located at each end of the space, in diagonally opposite corners; in large engine rooms four heaters may be used, with one in each corner. Small engine rooms, e.g., tugboats, can be heated with a single unit. In working spaces, one unit heater usually handles up to about 40,000 Btu/hr. 7.4 Factory-Assembled Units. Factory-assembled air-conditioning units are commonly used. The unit in Fig. 39 is a special marine high-pressure, highvelocity type for a Class G system as shown by Fig. 8; integral sound treatment attenuates noise from both the fan inlet and discharge. It is the first factory-assembled unit with an integral roll-type filter. In order to prevent injury to personnel and to minimize leakage, it is im~ortantthat the swine of doors on such units be arranged to open against pressure; i.e., close with pressure.
-
1
Section 8 Refrigeration Equipment 8.1 General. Marine refrigeration systems and their individual components are designed to operate with the angles of pitch, roll, list, and trim normally encountered by a ship. Design angles are 10-deg pitch, a momentary roll of 30 deg for merchant construction and 45 for naval, 15-deg permanent list, and 5 deg permanent trim fore and aft. All rotating equipment is normally installed with the shafts fore and aft so as to minimize gyroscopic
effects. Condensers, cooTers, drip pans, troughs, tanks, etc., have their long dimension fore and aft where possible. Swash plates or similar flow retarders are used to minimize the adverse effect of vessel movement. Also, drain connections at both ends of equipment provide free draining regardless of trim. Plans must clearly define the minimum space required for renewal of tubes in shell and tube equipment. Construction and installa-
766
MARINE ENGINEERING
tion requirements for equipment are contained in references [lo, 29, 301 for merchant construction and reference [31] for naval construction. Each component'must be selected to properly interface with all others and to deliver the required performance in the same operating time periods during pulldown and holding. In the past, the procedure was to separately purchase individual components of condensing and liquid chilling cycles and assemble them into a system aboard ship; however, the trend is to purchase these cycles as "packages," including the base, drives, and all controls. Generally this reduces costs and provides improved performance. Since "packages" afford less flexibility of arrangement, space and servicing requirements must be resolved early in the design stage. 8.2 ReciprocatingCompressors. Reciprocating compressors used in marine work are of the multicylinder, single-acting, air-cooled type. Also, they usually are of the open type and are belt- or direct-driven except for those small ones used in water coolers, ice cube makers, self-contained air-conditioning units, and reach-in refrigerabrs. These exceptions are of the welded hermetic type. Large size "semi-hermetic" (accessible) units are not common aboard ship, although greater acceptance may be anticipated for those up to about 20 hp. Small open compressors, 5 tons and less, are of the vertical type; larger ones have V/W cylinder arrangements. Specifications usually require built-in capacity control (cylinder unloading), positive (forced-feed) lubrication, and crankcase oil heaters (energized whenever the compressor stops). Compressors for merchant cbnstruction are selected from manufacturer's data based on standard code rating methods [40,41] or tests specified for naval construction [31]. Refrigeration compressors are selected for a suction temperature, under holding (normal) operation, between 15 and 20 deg 17 below the lowest refrigerated compartment temperature. Since suction temperatures below -20 F involve sub-atmospheric pressures, they are not advocated except during the pulldown of - 10 F compartments. Suction temperatures for air conditioning usually are between 15 and 20 deg F below the leaving air dry bulb temperature. Those for liquid chillers are 7 to 12 deg F below the leaving liquid temperature. Also, a minimum suction temperature of 35 F usually is specified for water chillers to avoid freeze-ups. 8.3 Condensers and Receivers. Refrigeration condensers are of the seawater-cooled type except for very small capacities (usually 1 hp and less), which are aircooled. Seawater-cooled condensers for merchant construction may have cast iron or fabricated (copper nickel) heads and water boxes, welded steel shells, serrated (low) fin or plain tubes (copper nickel or aluminum bronze), copper nickel or Muntz metal tube sheets, and steel or Muntz metal tube support plates. Zinc rods, replaceable without removing the heads, minimize tube corrosion. Condensers are usually sized for a 105 F condensing temperature and a fouling factor of 0.0005, although the latter
is sometimes 0.0010. Catalogs include specific heat rejection data for packaged units; general data may be obtained from reference [2]. (Refer to Chapter XIV for a detailed discussion regarding heat exchangers.) Marine receivers are made of steel, and have sweated or flanged connections. They may be vertical or horizontal; the vertical type, Fig. 18(c), is preferred because there is greater assurance that the outlet will always be sealed with liquid. Horizontal types are level with a bottom connection near each end, or pitched (5 deg when fore and aft and 15 deg when athwartship) with a bottom outlet near the low end. Receivers are fitted with several sight glasses, or a magnetic-actuated mechanical indicator, for determining the liquid level. Mechanical indicators, being subject to wear and malfunction, should be easily replaceable. The receiver must store the full refrigerant charge when the system is pumped down. The capacity should be a t least 120 percent of refrigerant charge to allow for expansion. 8.4 Evaporators. Marine evaporators may be classified as being of the direct or indirect types. The direct type cools the treated space directly; whereas the indirect type cools air or brine, which in turn cools the treated space or another substance. Bare and fin-pipe gravity coils generally are used in merchant ship's stores spaces having temperatures below 35 k', because of frost considerations (see Fig. 18). Embossed and plate coils have been used to a limited extent for cold tables and in refrigerated stores spaces where they also serve as shelving. They have many advantages, including versatility and ease of cleaning, which should make their use much more common in the future. Overhead gravity coils are not recommended for refrigeration applications; however, they are used in ammunition spaces requiring a small cooling capacity. Drain pans sometimes are fitted beneath gravity coils to collect and drain melted frost; electric strip heaters are installed in these pans and around the drain pipes to prevent refreezing. Bare pipe coils are made of steel or copper pipe; those of steel are hot-dip galvanized after fabrication and those of copper are tin-coated, Gravity fin-pipe coils are used exclusively for ship's stores of naval construction; they are of a standardized design. The steel fin-pipe coils, primarily used in merchant construction, consist of 1.25in-OD, 0.083-in-thick seamless tubing on 1.25-in. centers; the fins are 0.025 in. thick and the tube sheets are 0.074 in. thick. Cooling coils for refnieration applications require heavier fins (0.015 in. thick) than those for air conditioning; also all copper surfaces are tin-coated to avoid contamination [16]. Additionally, fin spacing is limited to 3 per in. when the compartment temperature is below 35 F, and 4 per in. for higher temperatures; the wide fin spacing minimizes the pressure drop (frosted coil) and facilitates defrosting. An advantage of fewer, or even no, fins often overlooked is that it necessitates more prime surface; this in turn ~ e r m i t sa higher evaporator temperature, higher (better) space humidity, and less frosting.
767
ENVIRONMENTAL CONTROL
Brine coolers, including water chillers used with reciprocating systems, are of the direct-expansion type; i.e., the refrigerant is expanded in tubes surrounded by brine. Such coolers may have several, independently controlled, refrigerant circuits. Coolers for centrifugal refrigeration plants are of the flooded type; i.e., the refrigerant surrounds tubes through which brine (water) circulates. The materials used for coolers depend on the corrosion characteristics of the brine refrigerant involved as indicated by Table 21. Unit coolers serve as evaporators primarily in ship's stores spaces of merchant ships where the holding temperatures exceed 35 l?,Fig. 18(a). Principally because of automation, unit coolers with time-actuated, electric defrosting have gained acceptance for applications below freezing (in lieu of gravity coils). Except for large ships having brine systems for stores spaces, unit coolers are of the direct-expansion type. When selecting unit coolers, care is required to prevent arbitrary criteria from dictating excessively large sizes, which waste storage space; for the same reason, unit coolers should be carefully located. Coolers should be arranged to blow toward the door so as to minimize infiltration. Cold diffusers fair coolers) may be factory-fabricated units, Fig. 37, or shipyard assemblies. Yard assemblies are mandatory for large systems, to obtain optimum space utilization and because of limited head room. Permanent compressed-air connections facilitate the drainage of heating and defrosting facilities; otherwise trapped water may freeze and cause damage. 8.5 Defrosting Facilities. Moisture forms frost on evaporator surfaces below the freezing temperature; this is most prevalent during pulldown. The accumulation of frost must be removed periodically. "Hot gas" defrosting, which uses the heat of compression to melt the frost, works well for small evaporators, i.e., in ship's stores (see Fig. 18). In this process the compressor discharge gas is fed directly to bare or finned pipe coils, where it releases latent heat and condenses; after this it is re-evaporated in another evaporator. Hot-gas defrosting has not proven successful for large evaporators (cold diffusers), i.e., those serving cargo compartments. Defrosting is very simple for cargo compartments cooled by brine systems [2]. Hot-gas and hot-brine methods have an advantage in that the beating medium is confined to the interior of the evaporator tube; i.e., minimum heat escapes into the refrigerated compartment. The method most commonly used to defrost large direct-expansion cold diffusers consists of spraying hot water (sweet or uncontaminated salt) over the evaporator coils (see Fig. 37). Water is kept below 90 F and the flow is manually controlled; ample drainage facilities are essential to quick defrosting. Automatic electric defrosting has been used for cargo air coolers to a limited extent. 8.6 Centrifugal Liquid Chilling Systems. Centrifugal liquid chilling systems operate on the same basic refrigeration cycle as those with reciprocating compressors. Also materials comprising the various components
Table 21
Materials Used for Brine Coolers Calcium Brine
Ethylene Brine
Shell. ... . .. . . . Bray or Welded steel welded steel Tubes ..... .. . . Copper Steel Tube Sheets. . . Admiralty , Steel metal Tube Supports. Brass Steel Heads. . . . . . . ,.Steel Cast iron
Welded steel
Cooler Part
.
Water
Copper Naval brass Steel Steel
i
are similar. Reference [42] contains an excellent discussion of, and comparison between, reciprocating and centrifugal system characteristics. Centrifugal systems are factory assembled and tested. Except for the larger sizes, they are shipped as single "package" units, which greatly facilitates installation and minimizes the possibility of refrigeration contamination and other construction hazards. "Package" type units are furnished complete, including all interconnecting piping and controls, and are fully automatic. Refrigerant 11is the most common primary refrigerant, being suitable for the high-volume, low-head characteristics of centrifugal compressors. Because the volume of gas is relatively large and the system pressure is below atmospheric, centrifugal systems have integral purge recovery units. Intercoolers (economizers) are provided to increase the efficiency "bf multistage compressor systems. Centrifugal systems are used for air-conditioning and high-temperature refrigerated cargo installations of the larger sizes. Until recently, approximately 250 tons of refrigeration was considered the dividing point below which it was more economical to use reciprocating systems. However, the introduction of single-stage machines and increased production have reduced costs such that centrifugal systems as small as 100 tons of refrigeration are common. Centrifugal compressors may be driven by directconnected steam turbines or by motors and speed increasers; the availability of electrical power often is the deciding factor. A turbine drive has the advantage of greater flexibility to meet partial-load operation without expensit'e controls. Motors are usually single speed and are generally of the open type although some are of the hermetic type. Since capacity reduction cannot be achieved by merely reducing the speed of @e prime - mover, multispeed motors are not used. One method of capacity rbduction employs a butterfly damper built into the compressor suction inlet. Another method utilizes automatic prerotational vanes in the suction inlet. Suction gas throttling is inefficient; thus, prerotational vane control is preferred. Hot-gas by-pass facilities, manually controlled, are provided for loads below the range of the integral compressor capacity controls.
MARINE ENGINEERING ENVIR~NMENTALCONTROL
8.7 Steam-Jet
1
I
and
Absorption
Liquid
Chilling
Systems. Steam-jet, or steam vacuum refrigeration, systems have been used to a rather limited extent aboard ship. Considering the simplicity of the cycle and installation, as well as familiarity of operating personnel with this type of equipment, it is difficult to understand why steam-jet refrigeration has not found greater acceptance. Absorption refrigeration has been used a t sea primarily aboard submarines, where very low sound and vibration are prime requisites. Absorption refrigeration is more efficient than, and has a weight advantage over, steam-jet refrigeration. It has not been applied to merchant vessels because it is not economically competitive and, also, because manufacturers consider the marine market too small to absorb the associated development costs. Exhaust steam and diesel engine jacket cooling water may be used as a heat source for absorption refrigem tion systems. A problem associated with absorption refrigeration is "crystallization"; i.e., solidifying of absorbent salt (normally a solution) within the unit. It may be caused by an abrupt and lengthy power failure without the benefit of shutdown, or a malfunction in the condenser cooling water control. It occurs first in the heat exchanger, and blocks the normal return of solution to the absorber. Decrystallizing involves the application of steam; some machines incorporate automatic decrystallization facilities. References
1 Handbook of Fundamentals, edited by 'c. W. MacPhee, American Society of Heating, Refrigerating, and Air Conditioning Engineers, New York, 1968. 2 Guide and Data Book, Applications, edited by C. W. MacPhee, American Society of Heating, Refrigerating, and Air Conditioning Engineers, New York, 1968. 3 S. W. Brown, "Modern Developments in Marine Refrigeration," Marine Engineering and Shipping Review, June 1948. 4 Carrier Corporation, Handbook of Air Conditioning Design, McGraw-Hill, New York, 1965. 5 Union Carbide Corporation, Glycols, New York, 1958. 6 "Thermal Insulation Report," T & R Bulletin No. 4-7, SNAME, December 1963. 7 "Heat Transfer Coefficients, " NavShips Design Data Sheet DDS 9390-1, Department of the Navy. 8 "Report on Ship's Stores and Installed Cargo Refrigerated Boxes," T & R Bulletia No. 4-4, SNAME, March 1962. 9 D. E. Phillips and John Wiley, Jr., "Environmental Control, " Ship Design and Construction, edited by A. M. DJArchangelo,SNAME, New York, 1969. 10 "Standard Specifications for Cargo Ship Construction, " Maritime Administration, U. S. Department of Commerce, Washington, D. C. 11 "Design Criteria Manual for Surface ships, Air
Conditioning,Ventilating, and Heating, )'NavShips 0938018-0010, Department of the Navy, February 1967. 12 "Electrical Engineering Regulations, Subchapter J," CG-259, U. S. Coast Guard. 13 0. D. Colvin, W. H. E. Hahne, and M. R. Colby, "Care of Cargo at Sea," Trans. SNAME, 1938. 14 0. D. Colvin and W. H. E. Hahne, "Care of Cargo a t Sea-11: Interim Report on Dehumidificationin Ships," Trans. SNAME, 1941. 15 0. D. Colvin and S. J. Duly, "Control of Humidity in the Cargo Spaces on Board Ship," Northeast Coast Institute of Engineers and Shipbuilders, February 1947. 16 "Handbook on Sanitation of Vessel Construction14Public Health Service Publication No. 393, U. S. Department of Health, Education, and Welfare. 17 "Agricultural Research Service," Title 7-Agriculture, Chapter 111, Part 319, U. S. Department of , Agriculture. 18 "Protecting Perishable Foods During Transport by Truck," Agricultural Handbook No. 105, U. S. Department of Agriculture. 19 "Transportation of Hanging Beef by Refrigerated Rail Cars and Piggyback Trailers," Marketing Research Report No. 485, U. S. Department of Agriculture. 20 L. A. Harlander, "Engineering Development of a Container System for the West Coast-Hawaiian Trade," Trans. SNA ME, 1960. 21 L. A. Harlander, "Further DeveIopments of a Container System for the West Coast-Hawaiian Trade," Trans. SNAME, 1961. 22 Eric Rath, "Iso-Thermic Unitized Cargo System, " Trans. SNAME, 1964. 23 J. J. Henry and Henry ICarsch, "Container Ships," Trans. SNAME, 1966. 24 "The Commercial Storage of Fruits, Vegetables, and Florist and Nursery Stock," Agricultural Handbook No. 66, U. S. Department of Agriculture, Market Quality Research Division, September 1954. 25 "Refrigeration Equipment for Storage Comparb ments-Heat Load Calculations and Selections," NavShips Design Data Sheet DDS 9590-1, Department of the Navy. 26 Rules for Building and Classing Steel Vessels, American Bureau of Shipping, New York. 27 Guide and Data Book, Systems and Equipment, edited by C. W. MacPhee, American Society of Heating, Refrigerating, and Air Conditionhg Engineers, New York, 1969. 28 Handbook of Automatic Controls, Alco Valve Co., St. Louis, Mo., 1959. 29 "American Standard, Recommended Practice for Mechanical Refrigeration Installed on Shipboard," American Society of Heating, Refrigerating, and Air Conditioning Engineers Standard 26-56 (also ASA B591-1958). 30 "Marine Engineering, Subchapter F, U. S. Coast Guard. 31 "Refrigeration Plant and Systems-Mechanical and Refrigeration System Components-Naval Ship))
i board," MILR-16743 (SHIPS), Department of the Navy. 32 "Test Code for Air Moving Devices," AMCA Standard 210-67 Air Moving and Conditioning Association, Inc., Park Ridge, Ill., 1967. 33 R. C. Strasser and H. E. Parker, "The Acoustic HabitaEility of Ships," Trans. SNAME, 1964. 34 Fan Engineering, Buffalo Forge Co., Buffalo, N.Y., 1961. 35 "A Method for Determining the Size of Ventilation Ducts," NavShips Design Data Sheet DDS 9380-3, Department of the Navy. 36 "Pressure Losses of Ventilation Fittings," NavShips Design Data Sheet 9380-1, Department of the Navy. 37 "Rules and Regulations for Passenger Ships," Subchapter H, CG-256, U. S. Coast Guard.
769
38 "Rules and Regulations for Cargo and Miscellaneous Vessels," Subchapter I, CG257, U. S. Coast Guard. 39 "Protection of Rail Shipments of Fruits and Vegetables, " Agricultural Handbook No. 195, U. S. Department of Agriculture. 40 "Standard for Refrigerant 12 and Refrigerant 22 ~ 25 Horsepower and Compressors and ~ o n d e d i nUnits, Larger," ARI Standard 516-60 Air Conditioning and Refrigeratio~lnstitute,Arlington, Va., 1960. 41 "Method of Testing for Rating Refrigerant Compressors," ASHRAE Standard 23-59 American Society of Heating, Refrigerating, and Air Conditioning Engineers, New York, 1959. 42 "An Introduction to Centrifugal Air Conditioning Plants," York Corporation, subsidiary of Borg-Warner Corp., York, Pa.
771
BEARINGS AND LUBRICATION
-
CHAPTER XX Watt V. Smith
1. M. Gruber
TYPE
OF BEARINO
DESIRED MINIMUM FILM THICKNESS
WEDOE PROFILE
I Bearings and Lubrication
h, b 0.001'
,
>
hm
TAPERED-LAND THRUST BEARINO
Section 1 Review of Fundamentals Fluid-film lubrication provides the basis for operation of bearings of all types. Lubricants of primary concern aboard ship are liquids which are, for the most part, petroleum oils; however, niany other liquids as well as solids and gases are used as lubricants in modern bearing technology. Fluid-film lubrication depends on the generation of sufficient pressure between adjacent surfaces to transmit most or all of the load. Fluid pressure may be generated either hydrostatically or hydrodynamically. Pressure for hydrostatic bearings is generated externally and is piped to pockets or recesses in the loaded zone of the bearing. Hydrodynamic (self-acting) bearings generate pressure by the relative motion between the surfaces which draws the fluid lubricant into a wedge-shaped space, thus compressing its volume, more rapidly than the pressure can be relieved by leakage. Since virtually all bearings used aboard ship are hydrodynamic, attention is focused on this class. However, a comprehensive treatment of hydrostatic bearing$ is given in references [l-51.' 1.1 Film Pressure Formation. The concept of the wedge, or more generally lubricant compression, is basic to lubrication theory. Figure 1depicts a variety of bearings and the fluid film shapes usually associated with their lubrication. I n Fig. 1, W indicates the bearing load, N the speed of rotation, U the relative velocity of the bearing surfaces, p the pressure in the film, and h, the minimum film thickness. The first three bearings (tapered-land thrust bearing, tilting-pad thrusf bearing, and the journal bearing) are treated analytically as hydrodynamic. The surfaces are considered to be smooth, rigid, and perfectly aligned. The fluid is usually considered to be a totally incompressible liquid and in most solutions its viscosity is considered constant even though all the heat is carried off in the lubricant. The fourth bearing, a rolling-contact bearing, is treated as being elastohydrodynamic. In elastohydrodynamic analysis, the deformation of the surfaces and the effect of changes in lubricant viscosity due to pressure are included in the analysis by solving first the elastic equations for an approximate pressure distribution, then the hydrodynamic equations for a pressure distribution corresponding to the deflected bearing shape with the 1.I
1
Fluid-Film Lubrication.
Numbers in brackete designate References at end of chapter.
lubricant viscosity adjusted to the local pressure. The process is then repeated until an acceptably small error is obtained. It should be noted that fluid films are very thin in elastohydrodynamic lubrication, approximately the same 'magnitude as the deflections of the surfaces. Gear tooth contact and the contact surface of traction drives are other examples where lubrication is elastohydrodynamic. I n inverse hydrodynamic lubrication shown in Fig. 1, the bearing surface is easily deformed many times the film thickness, and the lubricant properties remain un&anged a t the pressures involved. Blok, in reference [3], shows that the pressure profile attains a relatively constant value over the surface in contrast to the other lubrication regimes. Water-lubricated main propulsion shafting bearings are examples of bearings operating with inverse hydrodynamic lubrication. I n thermal wedge bearings, the thermal wedges result 'from differential heating or heat conduction through the bearing wall. Thermal wedges and unintentional mechanical wedges resulting from finishing operations probably provide much more of the load support than would usually be imagined. Flabland thrust washers are examples of bearings which operate with this type of lubrication. The pressure generated between the bearing surfaces is directly proportional to the relative velocities between the surfaces and to the viscosity of the fluid. The pressure is also inversely prowrtional to a power of the film thickness and is a function of the shape of the film. Figure 1 illustrates some of the pressure profiles associated with the wedge shapes shown. The maximum unit pressures in a hydrodynamic bearing are from 3 to 5 times the average unit load. Singe the average unit loads for oil-lubricated bearings are less than 600 psi in most cases, the maximum pressures in the film are less than 3000 psi. The maximum pressures in elastohydrodynamic bearings are less than the Hertzian contact pressures. Since the Hertzian pressure in a rolling-contact bearing may be as much as 300,000 psi in heavily loaded bearings, the film pressures are still very high. Inverse hydrodynamic pressures are low, of the order of 100 psi. The ability of thermal wedge thrust bearings to carry load is considerably less than that of tapered-land or tilting-pad bearings.
I
U
a
P
0 L 0
s
"m TILTINO
-
PAD THRUST B E U I N O -
-
hm, 0.001"
I
"m JOURNAL BEARINO
Fig. 1 Lubrication wedges
h m b P ~ ~ ~ ~ - e m
ROLLINO CONTACT BEARINO hm b I O O X 10
s
-hm
RUBBER BEARINO
b,,,bdOxIO
!
a
?Y Y:
FB
-''
-6 u
hn
FLAT-LAND THRUST WASHER
1.3 Viscous EfFects. The oil leakage out of the lubricant film may be determined from an equation of the form hZ d p = -(1) 121.1d z where
u = mean leakage velocity, ips
h = film thickness, in.
*
p = viscosity, lb-secLsq in.
dz
.
= pressure change per unit length, psi&
Table 1 shows the pressure gradient in psi per inch of length of flow path required to produce a mean out-flow velocity of 1 inch per second from a unit width of film. Table 1 illustrates the very large pressure gradients required to produce a modest flow velocity from the
'
772 Table 1
in.
Pressure gradient, psi/in.
0.001 170
SILICONE DC 500
WATER hmin = (1
16
7.3
0.69
0.31
0.001 0.001 0.001 0.001 0.001 15
28
13
, 1.0
I
Pressure Gradient for Leakage of 1 ips Mean Velocity
TURBINE OIL FLUID NS 2190 TEP Temperature, deg F Viscosity, centipoise 98.1 8.75 Film thckness,
773
BEARINGS AND LUBRICATION
MARINE EN(
1.2
0.54
film. The values of film thickness given are also numerically equal to the discharge in cubic inches per second per inch of film length. Table 1 also shows the effect of temperature on viscosity and that these effects depend on the nature of the fluid. The viscosity of petroleum oil decreases by 11.4: 1over a 100 deg F range of temperature, whereas the silicone fluid and water decrease by a ratio of 2.2: 1. The shear stress in the fluid film for each inch of bearing length of the bearing is the sum of the product of viscosity times the ratio of velocity to film thickness and the product of one half the film thickness times the pressure gradient. For a journal bearing running centered, the surfaces are parallel and the pressure term becomes zero. The frictional resistance in this case is entirely viscous. The analysis of this case was given by Petroff. The Petroff bearing is of interest because all other fluid-film bearings have greater friction. For the Petroff bearing
where
F
= tangential shear force on journal, lb Z = oil viscosity, centipoise N = journal angular velocity, rpm L = bearing length, in. D = journal diameter, in. C = bearing clearance (bearing bore - journal diameter), in. H = bearing power loss, hp f = bearing coefficient of friction P = bearing unit load (i.e., bearing load divided by LD), psi
The oil flow requirements hnd power loss of cylindrical sleeve bearings may be estimated by the following simplified and approximate equations [6] :
- n) C2
where
S
= Sommerfeld number (dimensionless) j = power loss factor (see Fig. 2) Q = oil flow, gpm p = oil density, Ib/gal C, = oil specific heat, Btu/lbdeg F AT = oil temperature rise (outlet-inlet), deg F n = eccentricity ratio (see Fig. 2) hmi, = minimum oil film thickness, in.
The foregoing expressions are among the many that have been developed for bearing characteristics. Hersey developed an analysis of experimental data that showed friction to be proportional to ZN/P. The Hersey number, ZN/P, has been shown to yield an important indication of the region in which bearings pass from boundary and mixed-film lubrication into a zone of fluidfilm lubrication. Figure 3 is a plot of data taken on a bearing of 60deg arc, with an L I D of 0.5, a D/C ratio of 500, and a load of 436 psi using Navy symbol 2190 T turbine oil as a lubricant. The speed was varied from 25 to 211 rpm. The zone to the right of the minimum point of the Hersey curve is the full fluid-film lubrication domain. The region to the left of the minimum is a zone of mixed friction with a decreasing portion of the forces being transferred by hydrodynamic pressure and the balance by solid-solid contact. Since the marine engineer is often confronted by mixed and boundary lubrication regimes, this subject will be discussed further. Before leaving fluid films, however, several of the more important references should be mentioned. The theory of fluid-film lubrication is given its most complete analytical treatment by Pinkus and Sternlicht [2]. Purdy [7] has developed the theory in a somewhat simplified manner, but with exceptional clarity in presentation. Hersey has given the most complete record of the development of lubrication theory in all its aspects in reference [B]. For bearing design and design checking, the presentations in refgrence [2], that by Boyd and Raimondi [3], and Rippel [5,9,10]are especially recommended. The work of Wilcox and Booser [6] is based on a large mass of experimental data and thus it offers the advantage of a verified analysis. Michell's work [ll] is good on the design of large marine and water wheel thrust bearings. 1.4 Boundary and Mixed-Film Lubrication. Boundary lubrication processes are difficult to define with precision since the mechanisms to be included are not well defined. Generallyyhen the film thickness becomes so small that the nature of the surface and the chemistry of the lubricant-surface combination become more impor-
Fig. 2 Eccentricity ratio and power loss factor versus Sommerfeld number for split cylindrical bearings
.o 1
I I SOMMERFELD
tant than the bulk properties of the fluid, boundary lubrication processes predominate. Friction wherein only the properties of the solid surfaces are involved is usually known as dry friction. Some dry-friction processes may be involved in boundary lubrication. Mixed film, as the name implies, is that portion of bearing operation where both boundary and hydrodynamic lubrication processes are effective. The requirement for a shearing velocity and an oil wedge to produce a hydrodynamic pressure was established in the previous section. The time to establish pressure and the rate a t which pressure decays after motion stops were not mentioned. The long retention of low restarting friction for some types of bearings may be interpreted as the holding of a substantial portion of the film pressure trapped between the bearing and shaft surfaces. The squeeze film is known to account for the operation of many bearing surfaces that have a load reversal. Diesel engine connecting-rod bearings are assisted by squeeze film; wrist pin bearings and piston ring flank lubrication are completely squeeze film. The significance of the high friction of dry contact and its accompanying high damage can be appreciated by consideiing that, during any interruption of motion, the machine surfaces operate in the boundary regime. The boundary regime is never quite free of some material transfer and thus the damage is proportional to the amount of time spent operating in this regime. Just as surface roughness can penetrate oil films down to and through the boundary layer, so too can dirt particles that circulate with the oil stream. The effects produced by the passage of dirt through a film can be
I
I
1.0
10
NUMBER, S -
HERSEY NUMBER, Fig. 3
9
Bearing friction versus Heney number
-
quite different depending on the circumstances surrounding the event. The least damaging dirt passage occurs when the dirt is smaller in size than the thickness of the film. The thinnest part of the film through which the particle passes determines the damage. If the particle is too small to contact both solid surfaces, then the force necessary for damage is not available and the dirt passes
774
. Additives are used to change some chemical and some
MARINE ENGINEERING
through, leaving the surfaces unchanged. When the particle size exceeds the film thickness, the least damage occurs when a ploughed track in the soft bearing material is the only result. Damage of this type can raise local temperatures, but the bearing damage is quickly smeared back to the level of the surrounding surface in most cases. The entry of dirt into the bearing has a more serious result when the dirt particle fails to transit the film but, instead, embeds in the bearing surface. Embedment may be complete, in which case the damage is restricted to that done to the journal in the embedment process. When the embedment is incomplete, the consequences may vary from the cutting of a small groove in the shaft, all the way to a catastrophic machining-type failure. In machining (or "steel wool") type failures, the steel surface is continuously cut by particles from the steel journal surface which are embedded in the bearing material and those that weld together to form massive scabs. The articles are hardened bv friction heating and oil quidching. The probability"of a machinin& type failure occurring increases with: dirt size, quantity, and hardness residual hardenability (difference in journal o r . thrust surface hardness and the maximum hardness attainable with the material) oil film temperature (which is primarily related to surface speed) amount of chlorine in the oil (usually an extremepressure additive) composition of rotor materials; carbide-forming elements (high-chrome steels can be especially troublesome)
physical characteristics of an oil. Several additives are usually blended into a single package suitable for converting a specific base oil into the desired product. The commonly used types of additives include: oxidation inhibitors-to reduce the onset and rate of oil oxidation corrosion inhibitoreto reduce or prevent oil attack on alloy bearings and internal rusting of machines antiwear improvers, oiliness agents, and extreme pressure (EP) additives-to improve sliding in the boundary lubrication regime detergents-to improve the cleanliness of surfaces dispersants-to keep carbon and other insolubles dispersed and circulating with the oil; this prevents malfunctions such as piston ring sticking alkaline agents-to neutralize acid from oil oxidation pour depressants-to lower the pour point antifoam additives-to reduce the persistence of foam tackiness agents-to reduce dripping tendencies thickener additive-to convert oil to a solid or semisolid lubricant (grease)
Greases are very important in the lubrication of bearings and gears, where speeds are low and loads are high, and in rolling-contact bearings where oil can bleed rapidly enough from the grease to provide the necessary amounts of lubricant. Grease properties differ greatly from the base oil. The amount and type of thickener additive and the rate of shear of the @pplicationboth affect the grease properties. Nine softness grades are recognized in the NLGI (National Lubricating Grease Institute) system ranging from 000 to 6 in order of increasing stiffness. &eases are formed as a gel of the When designing bearings, consideration must be given base oil and from 4 to 25 percent of a soap of (in decreasto the demands for stable, fully hydrodynamic operation in the upper operating range, the ability of the material ing order of quantity manufactured) calcium, lithium, sodium, aluminum, barium, or other metals or through and lubricant to provide prolonged operation in the the use of finely divided inorganic or organic compounds boundary lubrication regime at slow speed, and the probsuch as modified bentonite, colloidal silica, or arylureas. ability of encountering some boundary operation at A wide range of oils has been used including petroleum, high speed. The great significance of material factors in animal, vegetable, and synthetic oils. The wide variety all types of boundary operation cannot be overem- of startine materials and the combinations that have . . . -a phasized. been made have produced lubricants applicable over a 1.5 Lubricants. Lubricants for marine machinery temperature span from -100 to 450 F. usually have a petroleum base. Crude petroleum oils The factors to be considered in selecting the lubricants are roughly classified as paraffinic, naphthenic, or mixed to be used aboard ship are as follow: base; the differences being associated with the presence of paraffin wax in the paraffinic crude, asphaltic comThe lubricant in each machine must have sufficient ponents in the naphthenic, and both in the mixed-base viscosity to meet the most severe anticipated operaoils. An entire range of crudes exists, ranging from the tional requirements. highly paraffinic to the highly naphthenic; however, The viscosity of the lubricant in each machine should advances in refining and in additive chemistry have be held to the minimum to minimize power losses and greatly reduced the significance of geographic source in cooling requirements. the final application of the oil. Both the wax and the The volume and variety of lubricants carried aboard asphalts are largely removed in the manufacture of the ship should be minimized. lubricating oil. The specific property levels desired are From the known loads, speeds, and temperatures of attained by blending refined, naturally occurring oils and the power plant bearings and gear teeth, the physical by the introduction of additive packages.
775
BEARINGS AND LUBRICATION
i
properties of the lubricants satisfying the maintenance of a safe film thickness can be determined. The lubricant properties of primary concern, the test methods, and the significance of the results are briefly discussed in the following paragraphs. a. Lubricant Viscosity and Viscosity Index. In 1745, Newton measured the torque required to rotate concentric cylinders in a liquid. He determined that the force required on the rim of the cylinders was directly proportional to the speed of rotation and the area of surface in the liquid and inversely proportional to the distance between moving surfaces. The relationship from these results is F = p(AV/h), where p is the proportionality constant. This constant is the absolute viscosity. When A is the surface area in inches squared, V is the surface speed in inches per second, h is the oil film thickness in inches, and F is the force in pounds, the units of p are lb-sec/sq in. This is a reyn. The corresponding unit of absolute viscosity in the metric system is a poise. The units of a poise are dynes-sec/sq cm. The metric unit of absolute viscosity most commonly used is a centipoise, which is 1/100th of a poise. For oils, viscosity is usually measured as the time for a fixed volume to be discharged from a standard container through a fixed orifice at a specified temperature. The viscosity of most oils is nearly independent of the rate of shear and the oils are termed Newtonian fluids. Modern multigrade engine oils are non-Newtonian. At high shear rates, multigrade oils show a decrease in viscosity with increasing rate of shear until the viscosity of the base oil is approached. Grease is strongly non-Newtonian. The apparent viscosity is measured by pumping the grease through each of a set of eight capillary tubes using two pumping rates. From the observed pressures, the dimensions of the capillary tubes, and the pumping rate, the viscosity and shear stress are computed and viscosity in poises is plotted against rate of shear in reciprocal seconds on log-log paper. The viscosity of the base oil at the test temperature provides a lower bound on the grease apparent viscosity. The viscosity index (VI) is a measure of the rate of change of viscosity with temperature. Oils from a l'ennsylvania crude, which had the lowest rate of change of viscosity with temperature existing at the time the index was established (1929), were arbitrarily assigned a VI value of 100. Another series of oils from a Gulf Coast crude with the greatest change were assigned a value of 0. The viscosities of each series of oil were determined for 100 F and 210 F. The viscosity index of an oil is then computed as VI
-7
=
L - u L-H
--- X 100
where
U
= viscosity at 100 F of the oil whose VI is to be
calculated
0
100
PO0
100 1) 0
80Q
80
400
40 PO0
100 a0
10 1)
80
8
40
4 PO P 10 1)
I 0.1)
8
0.8
4
0.4 P 02
I 0.1
o
100
goo
a00
400
TEMPERATURE, DLORLLS FAHRENHEIT
Fig. 4
L
Lubricating-oil viscosity-temperature chart
= viscosity at 100 F of an oil of 0 VI having the
same viscosity at 210 F as the oil whose VI is to be calculated H = viscosity at 100 F of an oil of 100 VI having the same viscosity at 210 F as the oil whose VI is to be calculated In many cases a sufficiently high viscosity will be ensured at the maximum temperature by specifying the minimum acceptable viscosity at each of two temperatures; however, such specifications may permit an excessive viscosity when operating at intermediate temperatures. A comparison of the economic value of lower oil costs for a lower VI oil versus the longer term value of a reduced low-temperature power loss must be made to provide a valid decision. The effectof temperature on oil viscosity is illustrated in Fig. 4. Mercury and'cwater have been included to illustrate low-viscosity fluids. The data given in Fig. 4 may be converted to other standards by employing the following relationships: To obtain Reyns (lb-sec/sq in.), multiply the viscosity in centipoises (dynes-sec/sq cm) by 1.45 X lo-'. To obtain centistokes, divide centipoises by the oil density in grams/cu cm.
776
BEARINGS AND LUBRICATION
MARINE Ek
Above 60 centistokes, the following approximations may be used: Saybolt Universal Seconds = 4.62 X centistokes Redwood No. 1 Seconds = 4.05 X centistokes Degrees Engler = 0.132 X centistokes
b. Oxidation Stability. Oxidation stability provides a measure of the lubricant's resistance to oxidation under an accelerated test. Testing is performed by exposing a small oil sample (300 ml) contaminated with water (60ml), and containing a steel-copper wire catalyst coil, to a flow of 0.5 liters per hour of oxygen at a temperature of 203 F. Neutralization numbers are determined periodically during and at the end of the test. Limits may be set by reaching a maximum neutralization number or by time. The test provides an indication of the oxidation-limited life of a lubricant in service. c. NeutralizationNumber. The neutrdiiation number is a measure of the acidity of an oil. The oil is extracted or mixed in a solvent and the number of milligrams of potassium hydroxide required to neutralize a gram of oil is reported as the neutralization number. The neutralization number can provide a usefd indication of the degree of oxidation of an oil and hence to the need to replace the oil. Care should be used with additive oils since the additive may affect the neutralization number. d. Carbon Residue. Carbon residue is the percentage of carbonaceous material remaining after an oil sample has been burned. The test is best made in a small glass container (Ramsbottom Method) by subjecting a weighed sample of oil to a constant temperature of 1020 F for a fixed time period. The test indicates the carbon-forming tendencies of the oil and is significant for internal combustion engine oils. e. Demulsibility. The demulsibility of an oil is a measure of the separating tendency of oil-water mixtures or emulsions. Equal volumes (40 rnl) of oil and water are stirred for five minutes at 130 F. If the volume of emulsion (unseparated oil and water) remaining is less than 3 ml when the sample is examined at 5-min intervals, the mixture is reported as breaking and the time is noted. If the mixture fails to break after one hour, the volumes of oil, water, and emulsion remaining are reported. The maximum time for demulsification is specified for each particular oil. In used ails, the test is heavily affected by rust, oxidation products, and other less clearly defined contaminants. The test provides an indication of the effectiveness of separators in water removal and of the probable sludge-forming tendencies of the oil. Heavy sludges or emulsions can block the flow in oil return lines and reduce the effectiveness of oil coolers. f. Foaming. The foaming tendency of a lubricating oil indicates the stability of air-oil foams. The test is performed by the formation of a foam by the passage of 94 ml per minute of air for 5 min through a 200-ml oil sample. Foam volumes are noted as formed and also after settling for 10 min. Tests are run at 75 F and at 200 F. Lubricating oils having a tendency to foam are undesired because foam can cause inadequate lubrication,
oil overflow, blockage of oil return lines, loss of cooler effectiveness, air binding of pumps, etc. g. Corrosion Inhibition. The corrosion inhibition capabilities of a lubricating oil or grease are intended to define either the tendency of the product to corrode a component of the system being lubricated or to protect the surfaces in the system from being corroded by fresh or salt water entering the system. A wide variety of tests is used on new oils or greases. Each test is closely related to the intended application. Most of the tests are for surfaces wetted by the lubricant; however, tests have been developed also to measure the vapor space corrosion protection offered by oils that have vapor space corrosion inhibitors included in their formulation. All of the corrosion tests are examples of surface chemistry phenomena and therefore require exceptional care in the preparation of the test surfaces. The material of the specimen must be exactly to specification. The finishing method is critical in each step as to grades and types of abrasive, speeds of finishing, and scratch pattern and depth. Each of these items has an effect on the surface structure and residual stress, and hence on the corrodability of the surface. Finally, the cleaning of the specimen must be performed exactly as specified, if all the ~reviouscare is not to be wasted. The actual corrosion test methods are numerous. ASTM, federal, and military specifications cover most of those tests of concern to the marine engineer. One of the most frequent tests applied to both oils and greases is the copper strip test. Copper is easily corroded and is catalytic to petroleum oils or at least to some of the constituents. The copper strip after proper finishing and cleaning is partially immersed in the lubricant and heated in a bomb pressurized with oxygen for 20 hr at 210 F. After solvent cleaning, the test strip is compared to a reference strip for evidence of discoloration, etching, and corrosion. The test will identify lubricants with sufficient activity to cause corrosion. The corrosion inhibition test for steam turbine oils uses a sample of oil (300 ml) contaminated with 30 ml of seawater for marine turbine oils (or with the same amount of distilled water for oils to be used with landbased turbines). The resistance to rusting of very carefully abraded and cleaned low-carbon steel rods that are suspended in the oil-water mixture for 24 hr at 140 F is used as the test criterion. The Underwood corrosion test ig intended for use with internal combustion engine oil. The test specimehs are connecting-rod shells of bearing materials known to be easily corroded. Jets spray oil on the bearings at 375 F for 30 hr. None of the corrosion tests are expected to yield corrosion rate predictions but they do provide general guidance regarding the corrosive tendency or corrosioninhibiting characteristics of an oil. h. Flash, Fire, and Autogenous Ignition Points. These three points refer, respectively, to the temperature at which the vapor over a heated oil sample will flash when tested with a small flame; to the temperature at
t
1
which sustained burning occurs when similarly tested; and to the temperature at which ignition occurs without oxternal flame. These values have no meaning for lubricating processes but are significant where fire hazards exist. Marine lubricating oil distribution systems are generally considered to constitute such a hazard. i. Pour Point. The pour point defines the lowest temperature at which the oil will pour. An oil sample contained in a tube is cooled in 5-deg F increments. The tube is tilted after the temperature stabilizes and the oil behavior is observed. When the sample shows no motion after 5 sec, it is considered to have become solid. The previous 5 deg F point is reported as the pour point. The significance of the pour point in lubrication lies in the inability of pumps to move the lubricant to the bearings in time to prevent damage, when starting near the pour point. Tank heating coils are provided to avoid difficulties associated with low pour points. j. Grease Penetration. The grease penetration number is a measure of the consistency of a grease. Penetrat i o is ~ determined on both worked and unworlred samples. Worked samples are produced by forcing the grease back and forth through a perforated disk for a specified number of strokes. Normally only a modest number of strokes are required; however, a greater number of strokes can be specified and can reveal shear instability of some greases. The penetration is measured by allowing a standard cone to sink into the grease for a period of 5 sec. The depth of penetration in tenths of a millimeter is reported as the penetration. Grease penetration is useful in selecting greases for ball and roller bearing applications and as a quality control check. Greases that are too soft tend to excessively flow into the path of rollers and cause high heat through churning. Harder greases, especially those displaying channeling oharacteristics, run much cooler. Hardness alone will not ensure grease channeling. Channeling is a complex function of grease hardness, tack, fiber structure and application factors including temperature, shear rate, and vibration. Since no single test suffices to predict channeling, it is often necessary to resort to experiments in critical applications. k. Grease Drop Point. The drop point of a grease is the temperature at which the grease starts to drop out of the end of an inverted cone. The drop point is the nearest approximation to a grease melting point. I t does not provide a safe upper temperature for use of a grease, but it does identify a temperature above the usable range. 1. Service Simulation Testing. A wide variety of service simulation tests has been included in various specificationsfor lubricants. These tests are intended to check the overall suitability of a lubricant for a particular
777
service application. The purpose of the tests is to provide assurance that dervice operation will be free of trouble from unspecified or unspecifiable lubricant characteristics. Many different types of engine tests and the Navy work factor test may be cited as examples of this type of test. Service simulation tests are often very arbitrary. Although petroleum-base$ materials satisfy the great majority of marine lubrication requirements, a number of other lubricants,should be mentioned. Synthetic liquids have been among the most widely used petroleum substitutes. The synthetic fluids in wide use include [3]:
a a
Dibasic acid ester-for high VI, low volatility, and low freezing point Phosphate e s t e r s f o r fire resistance and boundary lubrication Silicones-for the highest VI, thermal and oxidation stability, and low freezing point Polyglycol e t h e r s f o r high VI, and good boundary lubrication
For more than a century, water has been used to lubricate stern tube and strut bearings lined with lignum vitae wood. Plastic and rubber bearings have also been used in conjunction with water lubrication. Pump bearings have of necessity been lubricated with water when water was the fluid being pumped, as is the case with the motors and pumps of pressurized water reactors. Air and other compressible gases and vapors have been studied extensively for bearing service where rotational speeds are high or where the space and weight savings of working fluid machines warrant the added engineering. These bearings have been employed extensively in gascooled nuclear reactor circulating fan bearings, in turboexpanders for cryogenic work, and in control gyros for inertial platforms in missiles. Solid lubricants include the lamellar solids (graphite, molybdenum disulfide, tungsten disulfide, boron nitride, etc.); the soft solids, (white lead, talc, clay, etc.); chemical coatings of oxides, sulfides, and phosphates; waxes (paraffin, beeswax); polymerized olefines (polyethylene, polytetrafluroethylene); thin metal films of tin, lead, silver, or indium; and a wide variety of composite materials. Graphite and molybdenum disulfide are the most widely used of the solid lubricants. They can be applied as rubbed films, bonded coatings, or as dispersions in a carrier such as heavy steam cylinder oil. A number of the solid lubricants are available in convenient aerosol dispenser packages. The solid lubricants are characterized in lubrication technology by requiring feplacement as the films wear out, by being unable to remove frictionally -generated heat, and by satisfactory operation at lower Hersey numbers than any other lubricant system.
I
778
Section 2 Bearings Bearings may be of either the slidingcontact type, where the bearing elements are separated by a film of oil, or of the rolling-contact type such as ball, roller, or needle bearings. Sliding-contact bearings include many varieties of sleeve or journal bearings, and they also include thrust bearings and guide bearings (e.g., bearings which guide linear motions such as crosshead guide bearings in diesel engines). The determination of the type of bearing to use is based upon a study of several characteristics relating to the mechanical requirement, the environmental conditions, and the relative cost. Each type of bearing has its peculiar advantages with respect to each condition of application; Table 2 summarizes the relative advantages of each [6]. 'Generally speaking, sliding bearings are used in propulsion steam turbines, main reduction gears, diesels, industrial-type gas turbines, turbine generator auxiliary power plants, diesel -electric emergency plants, large- - motors or gener.- ators, boiler feed pumps, large fans and blowers, compressors and refrigeration plants, lineshaft bearings, stern tube bearings, steering gears, and rudders. Rollingelement bearings are generally used with electric motors up to 50 hp, aircraft-type gas turbines, small gear sets, some small auxiliary steam turbines, and small mechanical equipment. 2.2 Pressure-Fed Journal Bearings. Many different types of journal bearings have been developed over the years and a rather large variety remains in use. These be roughly classified into the three differentmethods of lubrication: namelv, ~ressure-fed. , nonDressurized.and externally pressurized bearings. A number of designs of pressure-fed bearings are in common use. Variations in journal bearings lie in the shape and location of the oil grooves and in the geometry of the bore. Variations of grooving and bores give rise to the following common types of bearings: a. Cylindrical Bearing. This bearing, Fig. 5(a), has a cylindrical bore. It usually has two oil spreader grooves along the split line. It has a good load-carrying capacity and is used in heavily loaded gear bearings. In designs incorporating LID ratios of 0.5 to 1.0, loadings range from 100 to 500 psi. At light loads (under 100 psi) and higher speeds (over 3600 rpm) it is very susceptible to a bearing instability known as oil whip.2 2.1
General.
may
I
l
1 'I
1 1 11
I li
" , A
.
-
I
Any bearing instability, regardless of its nature, is commonly referred to as 'oil whipn in the marine trade. Actually bearing instabilities fall into two types, half-frequency whirl and resonant whip. Half-frequency whlrl is a vibration of a shaft in a fluid film journal bearing which may occur a t any s eed and a t a frequency of one-half or near one-half journal speed! A resonant whip is a resonant vibration of a shaft in a fluid-film journal bearing which is self-starting a t a shaft speed approximately twice the actual first system critical and remains at a constant frequency as the s eed increases. The fre uency of this vibration is equal to the &st critical frequency of t$e shaft regardle~of running apeed. 2
BEARINGS AND LLlBRlCATlON
MARINE ENGINEERING
SERVICEFACTORS
CHARACTERISTIC
b. Cylindrical Overshot Bearing. This bearing, Fig. 5(b), is the same as that in Fig. 5(a), except a relief groove is added in the unloaded half of the bearing. This is used to reduce the power loss and to provide extra cooling capacity. Oil is usually admitted a t the trailingedge spreader groove and allowed to flow over the shaft within the relief. The excess oil over the bearing requirement is often drained away from the leading-edge spreader groove. c. Pressure-Dam Bearing. Fig. 5(c) shows a pressure-dam or "Newkirk" bearing. I n the pressure-dam bearing a groove over the top half of the bearing is terminated at a sharp-edged dam about 45 deg beyond the top vertical line. Stopping the oil flow in the groove at the dam creates an oil pressure due to viscous pumping and this pressure increases the load on the bearing 1121. This design is quite successful in reducing the oil-whip phenomenon; however, it is not always a cure for bearing instability and it therefore has a limited use. d. Multiple-Groovp Bearing. This is another variation of the cylindrical bearing, but instead of two spreader grooves a t the split line, it has four axial grooves spaced either 35 or 45 deg from the vertical centerline. It has been used successfully in steam turbine rotors at loads of 150 to 250 psi and LID ratios of 0.4 to 0.6. It has some effectiveness in eliminating oil whip. A fourgroove bearing is shown in Fig. 5(d). e. Elliptical Bearing. An elliptical bearing has a bore similar to an ellipse. It is manufactured by machining the bore with shims installed at the split line. After machining, the shims are removed so that when assembled the bearing bore has a larger clearance horizontally than vertically. Usually this clearance ratio is 2 to 1. These bearings are more stable than circular bearings and are used for loads of 100 to 300 psi in LID ratios of 0.5 to 1.0. They are commonly used as steam turbine, industrial gas turbine, and generator bearings. An elliptical-overshot bearing is the same m an elliptical bore bearing but with a relief groove in the upper half, similar to the cylindrical-overshot bearing. It is designed for cooler running and for a decreased power loss. An elliptical-overshot bearing is shown in Fig. 5(e). f. Three-Lobe Bearing. A three-lobe bearing, shown in Fig. 5Cf), is very effective in preventing oil whip a t light-load conditions. Its chief disadvantage is its manufacturing difficulty and awkwardness in handling. It is commonly made in three sections with shims in the joints. g. Tapered-Bore Bearing. As a variation of the three-lobe bearing, a three-groove tapered-bore bearing has 30 deg of arc that is cylindrical and 90 deg of arc that is tapered to a depth of 0.004 to 0.005 in. Unlike the three-lobe bearing, the tapered-bore bearing can be machined in a lathe with a special bore-tapering attach-
Characteristics of Sliding and Rolling Contact Bearings
Table 2
Unidirectional Cyclic Starting Load Unbalance Shock Emergency
Good Good Poor Good Fair Fair
Speed limited by
Turbulence Temp. rise.
{
ROLLING
SLIDING
Excellent Excellent Excellent Excellent Excelle~t Fair ,*
Centrifugal loading Dynamic effects Poor in ball bearings except where designed for a t sacrifice of !oad capacity. Good in spherical roller bearings. Poor m 9ylindrical roller bearings
Mdsaldgnmenl tolerance
Fair
Starting fr-idion
Poor
Good
Small ' to 2 times the shaft dia
Large to f i the shaft dia
Type of failure
Often permits limited emergency operation after failure
Limited operation may continue after fatigue failure but not after lubricant failure
Damping
Poor
Type of lubricant
Good Oil or other fluid,. grease, dry lubricants, air, or gas
hbricatdon, quantity required
Large, except in lowspeed boundarylubrication types
Very swan, except where large amounts of heat must be removed
Noise
Quiet
May be noisy, depending upon quality of bearing and resonance of mounting
Power consumption
N'DaL Varies as -
Varies ing upon widely type de of fubriend-
Poor
cation. Varies directly as speed. Usually !ewer than glider bear1g Good
Space requiremenis (Radial beur-ing) Radial dimension Axial dimension
ENVIRONMENTAL Lowtemp. starting CONDITIONS Highctemp. operation
I
I
Maintenance
I
C
Oil or grease
Limited by lubricant Unlimited, except for cyclic loding
Limited by lubricant Limited by fatigue properties of bearing metal
Clean lubricant required
Clean lubriesnt required. Only oc? sionrtl attention m t h
Very small in massproduction quantities, or simple types
~nt'eh-ediatebut standardize& varying little with quantity
Function of design and installation
Function of t pe of installation. 6sually shaft need not be replaced
780
MARINE EN
(b)
( a ) CYLINDRICAL BEARING
( c ) PRESSURE-DAM BEARING
( 0 ) ELLIPTICAL OVERSHOT BEARING
((I) TAPERED-BORE BEARING
Fig. 5
CYLINDRICAL OVERSHOT BEARING
( d l MULTIPLE- GROOVE BEARING
( f ) THREE-LOBE BEARING
( h ) TILTING-PAD BEARING
Types of pressure-fed bearings
ment. These bearings can be made in two halves. A sketch of this bearing is shown in Fig. 5(g). h. Tilting-Pad Bearing. The most effective bearing to prevent oil whip and dampen shaft vibration is the tilting-pad or multiple-shoe bearing as illustrated by Fig. 5(h). The bearing surface is divided into equal shoes or pads (3,4, or 5 are the numbers most commonly used) and each pad is allowed to tilt freely a ; ~the load and oil-film wedge dictates. Pivot locations are normally centered, but can be shifted downstream to 0.6 the shoe length for increased load capacity. The tilting-pad bearing was designed for the light-load bearing instability problem, but its oil film operates-at a lower temperature than a comparable full-sleeve journal bearing and it is capable of withstanding higher unit loads than a fullsleeve bearing. Tilting-pad bearings are in common use on steam turbines, high-speed reduction gears, and centrifugal compressors. The bearings as listed in items (a) through (h) are generally in the order of increased rotor damping properties. The cylindrical bearing offers the least protection against oil-film whip or rotor unbalance vibration, while the tilting pad is the most satisfactory for these types of problems. Even though widely i
78 1
BEARINGS AND LUBRICATION
accepted, the tilting-pad bearing is often used only as a last resort, because it is more costly than other types of bearings. 2.3 Nonpressurized Journal Bearings. Nonpressurized bearings include bushings, oil-ring bearings, and disk bearings. Bushings may be oil, air, grease, or water lubricated depending upon the application or material used. Solid bronze bushings. are usually grease lubricated while sintered materials are impregnated with lubricants, such as oil or teflon. Some bushes are solid teflon or graphite, or metals coated with molybdenum disulfide. Other nonpressurized bearings include lineshaft (also called pillow block, tunnel, spring, or plummer block) bearings. Lineshaft bearings may be of either the oil-ring type (as in Fig. 13 of Chapter 11) or the disk type (as in Fig. 12 of Chapter 11). The number of oil rings in a bearing should be such that no ring is required to distribute oil for an axial distance greater than 7 in. on either side of the ring. The ratio of the inside ring diameter to the diameter of the shaft should be 1.5 to 1.25, decreasing with increasing shaft diameter. The ring cross section should be such as to have sufficient weight and area to deliver the required flow a t the given speeds and oil viscosities and the oil ring should be submerged a minimum of 1 in. below the oil in the sump under extreme pitch and roll conditions. For low-speed operations, rings for large shafts should be 1 in. wide or more and should be grooved a t the I D for increased oil delivery. There are several disadvantages to oil rings. At low speeds, such as 0.1 rpm while on turning gear, oil delivery is uncertain. Also, rings may tend to hold up on the ring guides and thereby impair oil delivery. I n addition, the oil delivery from rings is proportional to viscosity so that, as a bearing overheats and the oil temperature rises in the sump, the oil delivered to the bearing is decreased a t the very time when more oil is needed. Disk bearings are lubricated by a plate clamped to the shaft a t one end of the bearing. At the OD of the plate a cylindrical section forming a part of the disk rubs against a brass scraper which removes and directs oil into the bearing, generally a t the top centerline of the bearing (as in Fig. 12 of Chapter 11). The cylindrical section of the disk is submerged below the oil sump level, and rotates with the shaft at shaft speed. Disk bearings have a greater and more positive oil flow a t all speeds. Disk oil flow exceeds ?ing oil flow in a given bearing anywhere from 1.5 to 7 times depending on the speed. Tests with disk-lubricated bearings show that a t low speeds the oil delivery can be expressed as:
where Q = oil flow from disk, gal/hr b = disk width, in. D = disk diameter, in.
Fig. 7
SOMMERFELD NUMBER. S.
I I
I
Rg. 6
N
= p =
TZ N( FD)
2
Friction wefflcient versus Sommerfeld number for a 13-in. lineshaft bearing
I
disk mm oil viscosity, r e p s
Sliding bearings of the oil-film type operate on a hydrodynamic film after the surface speed is sufficient to carry the load. Under these conditions an oil-film bearing is extremely efficient and reliable. The problem arises when the speed is too low to maintain a hydrodynamic film. Because steam turbine rotors require a coolingdown period of several hours under turning gear operation, the lineshaft bearings during this period operate on boundary films or metal-to-metal contact even with extremely low loadings (the lineshaft speeds are in the range of so to % rpm). Tests on a 13 in. disk-lubricated lineshaft bearing with 300 SSU oil show that mixed-film lubrication (i.e., the onset of boundary film lubrication) exists a t a Sommerfeld number of about 0.008 (see Fig. 6). It is standard marine practice to limit loadings and rely on large LID'S because of the limitations resulting from continuous operation at turning gear speed. Some thrust bearings and journal bearings of large low-speed shafts have externally pressurized pockets in the bearings as shown in Figs. 7 and 8 for starting and stopping under hydrostatic pressure [I]. The hydrostatic lift in thrust bearings consists of a pocket a t the center of each pad of approximately 2 to 4 in. dia (Fig. 7). The pads are drilled from the pad OD to the center. An orifice and check valve are mounted on the pad OD which in turn is connected to a high-pressure manifold supplied with oil from a high-pressure pump. Figure 8 illustrates a hydrostatic lift arrangement for a journal bearing. The oil pocket at the bottom center2.4
Hydrostatically lubricated thrust bearing
Externally Pressurized Journal Bearings.
Fig. 8
Schematic diagram
of a hydrostatic lift
line has a total area of from 2 to 5 percent of the total projected area. Generally, 1 to 2 gpm of oil is supplied at a pressure of about 1000 psi; however, each application must be designed for the specific flow and pressure required to lift the load free of the bearing. Starting friction with hydrostatic lifts is reduced by a factor of about 10 to 1. For thrust bearings on deep submergence vessels, whIch have large breakaway thrust on starting, a hydrostatic lift may be required in cases where the available starting torque is limited. 2.5 Thrust Bearings. As in the case of journal bearings, there is a variety of thrust bearing designs which have been in use to meet various application requirements. The various types can be classed as follows: flat land, tapered land, tilting pad, step, and pocket thrust. The first three types are in common use and are therefore covered in detail. a. Flat-Land Thrust Bearings. The flat-land bearing is the simplest form of thrust bearing. It consists of an annular flat surface with or without oil grooves, running
MARINE ENGINEERING
Fig. 9
Self-equalizing main thrust bearing
against a rotating thrust collar or shoulder. The loadcarrying ability is derived from the proper number and radiusing of the oil grooves on the flat face. Theoretically the flat-land bearing has no load-carrying capacity. However, experience has shown that a well-designed flat-land bearing will operate successfully a t a unit loading of 100 psi. Their normal application is in the range of 50 to 100 psi. This type of bearing is used on pumps and compressors on the inactive face (or idle side) of the thrust collar, where the reverse load is momentary. Where steady, heavy-thrust loads exist, a tapered-land or tilting-pad bearing is used. The power loss for a flat-land bearing can be determined by the use of empirical data such as presented in reference 161. b. Tapered-Land Thrust Bearings. The taperedlalid thrust bearing has the general appearance of the flat-land thrust bearing. Its surface is divided into a number of pads separated by an equal number of radial oil-feed grooves. I n the tapered-land bearing, each pad is tapered in the circumferential direction so that the motion of the runner will wipe oil into the wedge area to build up load-carrying oil pressures. The taper can be either single, that is, the same amount at the I D and OD, or compound, where the taper a t the I D is larger than that a t the OD. For optimum load-carrying capacity, compound tapers are used. The taper usually extends for 80 to 90 percent of the pad with the rest of the pad remaining flat. For convenience of manufacture and to accommodate splitting in two halves, an even number
of pads is always used. The ratio of pad length to pad height is kept near 1.0 for optimum load capacity. The oil groove width amounts to about 20 percent of the pad length. Pad tapers range from 0.005 to 0.009 in. a t the I D and from 0.003 to 0.006 in. a t the OD, depending on the pad size. As the pad increases in size, the amount of cooling oil required increases and thus tapers are increased to pass a sufficient amount of cooling oil. Detail calculations can be made for this type of bearing by employing the procedure outlined in reference [6] or [131. c. Tilting-Pad Thrust Bearings. The tilting-pad thrust bearing of the Kingsbury or Michell types (both held patents concurrently dating from 1910) differ from the tapered-land bearing in that each pad is an individual shoe that is free to pivot as the oil film dictates. The pivots can be a radial line as in the Michell type or a radiused button support as in the Kingsbury type. It has become common practice in American marine applications for tilting-pad bearings to have plates, known as leveling links, which equalize the load between pads. In Europe and Japan, most marine thrust bearings are built with thrust pads supported directly in housings without leveling linlm. A conventional marine ma& thrust bearing ofthe self-equalizing type is shown in Fig. 9. A standard thrust bearing has 6 pads with the OD equal to twice the ID. The pads, with 51-deg arc, have a length to radial height ratio of 1.33. Bearings of this design have an area equal to one half the square of the bearing OD. They are capable of carrying loads of 300 to 500 psi, depending on the bearing size, speed, and the type of oil. Larger bearings for main propulsion shafting usually have restrictions on the bearing OD; therefore, in order to minimize the bearing size, these bearings quite often have 8,10, or 12 pads. One conventional marine thrust bearing has 16 pads. The most common marine propeller thrust bearing for cargo ships and tankers is an 8-pad bearing ranging in OD from 41 to 61 in. There is no hydrodynamic or manufacturing limit on the size of thrust bearings: The spherically radiused button supports of tiltingpad bearings are generally placed at the center of the pad in the circumferential direction. This is to accommodate ahead and astern rotation. A study by Raimondi and Boyd [14] showed that in theorfa flat pad with a center pivot had no load capacity. In actual practice the pad crowns due to the load over the center pivot and from thermal gradients in the pad itself. Thrust bearings for unidirectional rotation can be designed with a pivot a t 0.58 the shoe length for optimum load-carrying capacity. Navy practice is to apply self-equalizing main thrust bearings at loads of about 500 psi, while the merchant marine practice is to size the bearing at 350 to 400 psi 1151. Most propulsion steam turbines employ self-equalizing pad bearings loaded a t 400 to 500 psi. Nonequalizing
BEARINGS AND L~BRICAT~ON
1
bearings should be limited to a maximum loading of about 300 psi. Tilting-pad thrust bearings quite often have the same bearing elements ahead and astern for interchangeability purposes; a full set of bearing elements may be installed on both the ahead and astern sides or a reduced number may be provided on the astern side (in exceptional cases). Also in exceptional cases, bearings of smaller size may be provided on the astern side than on the ahead side. The bearing design variables of film thickness, power loss, oil flow, and temperature rise can be determined by the hydrodynamic relations given in reference [6]. d. Resonance Changers. An interesting modifica tion to the leveling link (hard) main propeller shaft thrust bearing is the resonance changer or soft thrust bearing. Instead of leveling link supports behind the thrust pads, a piston and bushing is placed beneath each pad, Fig. 10. The pistons are connected to oil manifolds and piped to a high-pressure (2000-3000 psi) oil supply. By varying the quantity of gas in an accumulator, the longitudinal resonant frequency of the shafting system can be changed. This bearing modification has proven extremely useful in multiple-screw ships where excessive longitudinal shaft vibrations are experienced when the ship is in a hard turn [16]. e. Thrust Meters. Electronic thrust meters were first used by the Navy for standardization trials on new ships; however, electronic thrust meters have subsequently been frequently used in pumps, turbines, compressors, and as permanently installed equipment to measure propeller thrust. An electronic thrust meter consists of a standard tilting-pad bearing with strain gage type load cells added to the shoes or leveling links; see Fig. 11. In an &pad leveling-link bearing, load cells can be used in quantities of 2, 4, 6, or 8 depending on the accuracy and reliability desired. Two or four are common for tankers or cargo ships. Standard load cells have been manufactured to fit all thrust bearings from 5 to 65 in., ranging in unit load cell ratings of 1000 to 100,000 lb each. Load cells can be fitted to existing machinery installations. The thrust signal from the load cell is an electronic one, and can be read or recorded from any one of a large variety of instruments ranging from a single readout of strain to a multichannel recorder, which can record directly any number of cells simultaneously. These devices are extremely responsive to thrust variations. 2.6
I
I
Rolling-Contact Bearings
a. General. Rolling-contact bearings are distinguished by the use of a series of rolling elements to position the shaft with respect to the housing of the machine. The rolling elements most frequently employed include; balls, needles, and cylindrical, tapered, and convex rollers. The rolling elements of rollingcontact bearings provide much closer positioning of shafts than can be achieved with the use of self-acting sliding bearings. In addition to the close positioning,
Fig. 10 Resonance changer '
Fig. 1 1
Electronic thrust meter
rolling-contact bearings provide a radially (or axially) stiff bearing that permits heavy loading of machine components with minimum deflection. The lubrication system is usually simpler for rolling-contact bearings, e~pecially~where the size, load, and speed are such that grease lubrication can be used. Rolling-contact bearings have much lower starting friction coefficients (0.002 to 0.006) than self-acting sliding-contact bearings (0.15 to 0.25). The load capacity of rd@ng-contact bearings is fairly well-defined in terms of the cycles of operation to obtain a fatigue failure in a definite percentage of a given population at a single load level. This fatigue failure mode is the normal basis for sizing rolling-contact bearings to satisfy the requirements of a given application. Equations have also been developed to compute the static load to cause surface indentation of a size known to cause rough running. The static load capacity
784
MARINE EN(
provides a design limit for slow-speed, high-load applications. The contact areas in rolling-contact bearings are very small and the cyclic stresses in the load zone are quite large. Useful fatigue lives are achieved by using very hard (Rockwell "C" Scale 58 to 65) high-yield material. The majority of bearings are manufactured from SAE 52100 steel containing 13 percent chromium and 1percent carbon. Heavy-duty bearings, which are used in applications where shock is likely; employ casecarburizing grades Qf steel. Instrument bearings are made from higher alloy steel such as AISI 440C (17percent chromium, l-percent carbon) ; the lower hardness attainable with this grade of steel is more than compensated for by the increase in atmospheric corrosion resistance. Nearly all the steel used in bearings is vacuum-melted or vacuum-degassed to reduce nonmetallic inclusions in the steel. The nonmetallic inclusions act as nucleation sites or stress raisers in the steel that are responsible for many of the fatigue failures in bearings. The size variation of rolling-contact bearing elements must be ltept to very small values. Size variations within an element or between elements in a bearing must be minimized to provide a uniform distribution of load between the elements. Any lack of internal uniformity must be compensated for by compression of the rolls and deflection of the rings arid supporting structure. Rem&kable advances have been made in achieving uniformity of rolling-element diameters. Size variations of from 50 to 300 microin. were not, uncommon less than a decade ago. At least one military specification, MIL-B-17931, restricts ball size variation within a single bearing to 10 microin. Balls with only 3-microin. diametrical variation are available. Unevenness in mounting surfaces also imposes an unequal load distribution on the bearing components and probably accounts for many of the premature bearing failures. b. Mounting. The highly desirable rigidity and close-positioning capability of rolling-contact bearings have implications in the incorporation of these bearings into machinery designs. Careful analysis of the starting and operating temperature gradients of the machine design should be made to prevent internal loading of the bearings as a result of thermal expansion forces. A number of mounting designs have been developed to accommodatethe most frequently encountered situations. Two basic alternative principles underlie most of the mounting arrangements; i.e., fixed-free mounting or opposed-shoulder mounting. These arrangements assume that the shaft will be supported by two bearings, one near each end of the shaft. Axial positioning will be determined by either imposing the constraint at one end, as in the fixed-free mounting, or allowing the shaft to float between the opposed shoulders of the shaft and housing. The free-end play provided in a fixed-free mounting must exceed the sum of the thermal and elastic differential motion between the shaft and the
BEARINGS A N D LLlBRlCATlON
housing. In cases where the free-end bearing is to be preloaded to provide a quieter installation, the spring force should be applied so that it is reduced by differential thermal expansion within the machine. Opposed-shoulder mountings tend to be less expensive than fixed-free mounting in manufacture; however, the axial location is not as close as may be obtained in fixedfree mounting. Some opposed-shoulder mounts are designed to allow adjustmetit of the free play through the use of shims between the cap shoulder and the housing to obtain the degree of axial control sought. c. Ball Bearings. Ball bearings consist of one or two rows of balls contained in grooves having a circular cross section. The grooves form raceways and are normally cut into rings that confine the balls. The radius of the raceway cross section is slightly larger than that of the ball. The largest ball possible, consistent with the other design features of the bearing, is normally used since this gives the largest load capacity. The sphericity of balls is defined by the ball grade, the variation becoming smaller as the grade number becomes smaller. Since the cost of the smaller grade number balls increases sharply as the number decreases below grade 50, the largest grade number consistent with the required performance should normally be used. Ball bearings accept either radial or bidirectional thrust loading. Angular-contact ball bearings provide a very high axial load capacity in one direction. Duplex pairs of angular-contact ball bearings are used for very high bidirectional axial loadings. Ball thrust bearings are designed with a row of balls running in grooved washers placed perpendicular to the axis of rotation. The bearing will accept virtually no radial load and thrust load is liinited to one direction. Two-direction thrust capacity is obtained by adding a second row of balls and a third washer. d. Cylindrical Roller Bearings. Roller bearings are classed as line-contact bearings in contrast to the pointcontact designation of ball bearings. Cylindrical roller bearings consist of right-circular cylindrical rollers between rings of cylindrical inside and outside diameters. The roller length is less than four times its diameter. Rolls are separated by retainers that may be positioned radially by the rolls or by either of the two rings. Tho rolls are restrained in an axial direction by ribs on either of the two rings. The cylindrical roller bearing has very little thrust capacity, and for this reason it is frequently used to provide longitudinal freedom in fixed-free shaft mountings. The fixed-end bearing may be any bearirig providing axial location. Cylindrical roller bearing^ with solid rolls usually have a somewhat larger radial play than ball bearings of the same bore size. By tho use of hollow rollers, cylindrical roller bearings may ba given an internal radial preload. The cylindrical roller bearing has a very high radial load capacity and low friction. e. Needle Bearings. Needle bearings differ from cylindrical roller bearings in having their roll length more than four times the roll diameter. The needlu
bearing is normally restricted to shafts less than three inches in diameter and speeds below 3600 rpm. The needle bearing is qyailable as a full complement of needles and as a bearing with the needles separated by a oage. The cage-type bearing is less subject to skewing of the needles than the full-complement bearing. The noedle bearing occupies the least radial space of any roller bearihg. Both needle and cylindrical roller bearings may be used without an inner race; in such a aese the shaft must be hardened to a Rockwell C hardness of 58-65 and given a fine grind. f. Tapered Roller Bearings. Tapered roller bearings use frustums of a cone as a rolling element. The races have a mating taper. The apex of the tapers on both rings and rolls meet at a single point on the axis of rotation. Cages are used to separate the rollers and a rib is provided on the inner ring to accept the roll thrust oomponent resulting from the small angle of divergence
1
of the conical roller. The large end of the roll and its mating rib are shaped to provide a converging loadcarrying wedge. Pairs of tapered roller bearings used as fked-end locating bearings provide a very rigid highload capacity unit. g. Roller Thrust Bearings. Roller thrust bearings have no radial load capacity and must be used in conjunction with a radial bedring. The radial bearing must be positioned very carefully if internal loading between the radjtL1 and thrust bearing is to be avoided. Roller thrust bearings have a very high thrust capacity. Their speed limits are much lower than radial bearings and more viscous oil is usually employed to prevent smearing of the surfaces. The supporting structure must be very rigid to develop the full capacity of roller thrust bearings. Care must be taken regarding the oil circulatian in large roller thrust bearings to avoid thermal distortions.
Section 3 Lubrication Systems
I
3.1 Obiectives. Lubrication systems have a number of objectives: these objectives mav be: To deliver the required amount of lubricant to each rubbing surface under controlled levels of temperature and contamination. To provide operator assurance that delivery is being accomplished and to provide timely warning of system malfunction. To provide the capability of coping with reasonable contingencies. In many cases, to provide oil to the speed control and to the low oil pressure shutdown device.
not be repeated here; instead, the present chapter will concentrate on the geared steam turbine lubrication system. Reference fig], a publication of the joint ASTM-MEMA-ASME committee on turbine lubrication, provides the most detailed guidance available for marine geared-turbine systems. Related references [20-261 provide additional recommended practices. There are two basic types of lubricating systems: gravity systems and pressure systems. In addition, however, there are two typgs of lubricating systems which combine features of both gravity and pressure systems; these are high-head pressure/gravity systems and low-head ~ressure/mavitv , " svstems. " A gravity system uses one or more head tanks to supply oil to the propulsion plant. This system requires a high head room; reference 1271 suggests a minimum elevation of 30 ft above the machinery. It is the most reliable type of system, if adequate capacity is built into the head tanks. Reference [19] recommends a minimum of four minutes of, oil supply in each tank. The lubricating-oil pumps supply the head tanlts directly in the gravity system. Figure 14 of Chapter 18 illustrates a piping diagram of a typical gravity system. The pressure system supply is direct from the lubricating-oil pump. This system eliminates the space and weight'requirements of the gravity system's head tanks; however, the reserve oil supply must be foregone in order to utilize direct pumping. Two pumps, each capable of supplying the full requirement, are required by reference [17]. Automatic switching of the supply from one pump to the other is required. A gravity/pressure system is a pressure system with one or more head tanks floating on the line to provide a u
1 '
Marine lubricating systems are distihguished by the necessity of including list, trim, roll, and pitch as design criteria. The American Bureau of Shipping 1171 requires satisfactory functisning of lubricating ~ystemswhen the vessel is permanently inclined to an angle of 15-deg athwartship and 5-deg fore and aft. In addition, reference [17] requires that the bearings not epill oil under a momentary roll of 223 deg, for electrical generators. Military vessels are required to accept larger trim, list, roll, and pitch requirements. Reference [18] cites the same permanent trim and list of 5 and 15 deg for surface ships, but adds 45-deg roll and 10-deg pitch requirements. For submarines, reference [18] imposes a requirement of 30-deg trim, 15-deg list, 60deg roll, and l0deg pitch. 3.2 Syslem Types. The lubrication system selection is determined in part by the prime mover selected and, in part, by the layout of the machinery space. The systems for diesels and gas turbines have been covered in the chapters dealing with these prime movers and will
786
MARINE ENGINEERING Table 3
CONTAMINANT ,-----AIR-? Temperature, deg F 70 140 Contaminant density lb/cu. in. 0.000043 0.000038 Contaminant Diameter mlcron8 mib
T
Rising Velocity, ips
BEARINGS AND 'LUBRICATION
-- --
Contaminant Settling Rates in Turbine Oil
--
WATER 70 140
0.03603
SAND 70 140
0.03552
0.094
-.
0.094
sf temperature is recognized in reference [28] by recom- the emergency generator is started automnticnlly niid
IRONOXIDE7,-IRON, 70 140 70 140
0.206
0.206
0.2M
0.284
1.3wk 2.lhr 1.3 m 0.8sec
2.4mo 17.5hr 10.5 m 6 sec
0.9wk 1.5hr 53 sec 0 . 5 sec
Settling Velocitv. ~ D S
Contaminant Diameter mlcrons mils T Time to Rise 12 in. -, , 2 0.08 19 mo 1 . 7 m o 157mo 11.2mo 9.8mo 20 0.80 5.7 d 12 hr 1.6mo 3.4d 3.0d 200 8.0 1 . 4 hr 7.3m 11.4 hr , 4 9 m 43m 2000 80.0 49 sec 4 sec 411 sec 29 sec 26 sec mo = months; wk = weeks; d = days; m = minutes; sec = seconds fi
reserve supply of oil. The high-head version is designed to provide full flow a t rated pressure for four minutes. The low-head system supplies oil for the same period but at a reduced pressure. The various types of lubricating systems have most of the functional elements in common. The recommendations on their design, location, and capacity are virtually the same. Oil is supplied to the ship through a fill line to the storage or settling tank, run by navitv to the main lubricating-oil sum6 up iy suction, and discharged through filters and lubricatingoil coolers to the head tanks or to the bearings, depending on the specific system in use. 3.3 Tanks. The design of storage, settling, sump, and head tanks is largelv based on the same desim criteria. The material recommended for use is restricted to clean steel plate. Nonferrous metals have catalytic effects on the oil which tend to promote oxidation. Many coatings have been attempted for the interior of oil tanks; however, the lack of total success and the high hazard associated with coating failure have led to the recommendation that interior coatings not be used. Tank inlets should be arranged to prevent the introduction of air below the surface by the impingement of oil jets on the surface. At the same time, it is of substantial benefit to introduce oil as near the free surface as possible to minimize the distance that the air entrained in the oil must rise to reach the surface and be released. Suction intakes in lubricating-oil tanks should be maintained a minimum of 10 in. above the tank bottom to avoid picking up solid contaminants from the bottom. The distance from the inlet to the outlet should be arranged to maximize the distance traveled by the oil in passing through the tank. The greater the transit time, the more the particulate matter will settle out; in order to increase the transit time of the oil, baffles are sometimes provided in tanks to good advantage.
--
Time to Settle 12 in. 4
0.8mo 5.9hr 3.6 m 2 sec
3.5mo 1.ld 15.2m 9 sec
Guidance with regard to the necessary transit time (or settling time) of an oil in a tank can be obtained by investigating tbe terminal settling (or rising) velocity of the anticipated contaminants. Stokes showed that the viscous resistance, f, of a small sphere of radius 1.traveling through a fluid, gas or liquid, of viscosity p with a velocity u was given by the equation
f
=
6?rrpu
For a sphere acting under the force of gravity, the gravitational force is (4/3)?rr3(w - w') where w is the weight density of the sphere, and w' is the weight density of the fluid. At the terminal velocity, the gravity and viscous forces are equal; therefore,
-
F
(4/3)?rr3(w - w') = 61rrpu
and thue U =
2r2(w - w') 9~
(10)
where u = terminal velocity of sphere, ips r = sphere radius, in. w = sphere weight density, 1b/cu in. w' = liquid weight density, ib/cu in. p = liquid viscosity, lb-sec/sq in. Table 3 gives the settling rates of the more common contaminants in turbine oil which is presumed to have a viscosity of 47 microreyns at 70 F and 4 microreyns at 140 F with a specific gravity of 0.9 at 60 F. The tablo shows clearly that any chemical agent capable of increasing the size of foreign particles can accelerate corltaminant removal; chemical agents are the responsibility of the oil supplier. No very exhaustive treatment of tank desigp is recognized as being authoritative. The favorable effect
1 I
mording the provinion of heaters in settling tanks. Houters arc dso suggested for sumps but only for heating tho oil to a temperature near that of operation. Referonce [20] contains some explicit warnings on the danger of heating coils with an excessive surface temperature. hmaximum surface temperature of 250P isrecommended with a flow rate of 2-3 fps. If steam is used for heating, the pressure should not exceed 5 psig. Electric heater limitations of 12 watts per square inch for moving l y ~ t e mor s 6 watts per square inch for static systems are recommended. Tanks must be provided with adequately sized vent lines and overflows. Vents must be so located that direct impingement of cold air on the vents is avoided; otherwise, the breathing of the tanks under normal operation may bring moist warm air into contact with the cold metal and result in condensation and rusting. Cure should also be taken to prevent the possibly combustible vapor in the vent lines from coming into contact with hot surfaces. The safe handling of the overflow should be considered for each tank. Every taxik should be provided with a drain system, preferably from a well at the lowest point of a sloping bottom. A oonnection to the oil purifying system should be located to take suction above the top of the well. Oil drained from the well is normally waste oil, too heavily contaminated to be cleaned with a reasonable effort. Tanks should be designed with the necessity of initially cleaning the lubricating oil system and the possibility of subsequent cleaning taken into account. Baffles, dams, t~ndthe like will always make cleaning more difficult but good design practice can relieve this difficulty to some extent. Tank capacity is based on the maximum oil requirement per unit time, and on the minimum allowable residence time in oil return lines and tanks. The oil defoaming requirement exerts considerable influence on the desired ratio of tank capacity to3flow rate. Head tanks in gravity systems are specified to have fourminutes capacity by reference [19]. Sump tanks must be able to accept at least one overhead tank, in gravity systems. Military ships with pressure systems may use sump tanks having as little as one-minute flow capacity. Storage and settling tanks should be designed to accept a complete oil change. 3.4 Pumps. Lubricating-oil pumps may be of either the positive-displacement or centrifugal types. Reference 11'71 requires two pumps, each capable of meeting the full system requirement. Reference [19] recommends that each pump have a capacity equal to 125 percent of the full flow requirement for the system. The use of different types of drives for the two pumps is preferred; if the primary pump or its power source fails, then the resulting low discharge pressure brings the idle pump on the line. However, this arrangemknt is not common in commercial practice. I n commercial practice, both drives are usually electric and come off the emergency switchboard. If the main generator fails,
provides continuity of electric power. Pumps driven by shaft takeoff gearing (usually from the reductioli gears) have been used and are satisfactory if properly applied. I t should be apparent, however, that meeting the oil requirement at very low speeds and at jacking speeds with a ~haft~drive?pump is not practical. Auxiliary motor-driven pumps are installed for lowvspeed and turning-gear operation. Check valve+'inust be provided on pump suctions to prevent pumping through the standby pump to thesump. Pump isolating valves should be provided to permit underway repair of either pump. 3.5 Strainers. Duplex lubricating-oil strainers are customarily installed after the pump discharge. These strainers, equipped with 80- to 100-mesh screens, are capable of passing 150-micron particles [19]. I t should be noted that 150 microns equals 0.006 in., a much larger size than the designed minimum film thickness in marine bearings at the lower speeds of operation. Finer filtration imposes severe penalties in the form of much larger filters and higher pressure drops across the filter. The lubricating oil is provided with its fine filtration by means of a bypass cycle to the purifier system. That portion of the oil used for turbine speed control may be given supplementary filtration. This higher degree of filtration is restricted to the much smaller quantity of control oil. Strainers should be provided with pressure-drop indicators to provide warning of the necessity for cleaning the strainers. Various recommendations have been made for an allowable change in pressure drop across a strainer before shifting and cleaning. "Any noticeable increase, " "5-psi increase, " and "25-percent increase over normal" have been recommended. I n a constanthead system, the pressure increase across the strainer represents a reduction of oil flow and the allowable increase should be related to the system pressure drop. Where head tanks are employed, the allowable pressure drop should be less than that required to overflow the head tank. I n a pressure system, the pressure drop should be less than that required to produce a low-oilpressure alarm at the most remote bearing. Where positivedisplacement pumps are employed, the allowable pressure drop is more dependent upon the effects of overpressure don the filter medium. The strainer collapse pressure is an undesirably high limit since strainer openings are enlarged by pressure, ~frmittinglarger particles to pass. Allowable pressure hmits should be selected that will not endanger the balance of the lubrication '. Te system. Strainer shifting valves should be so arranged that there will be no flow interruption when shifting strainers. Valves should be provided to vent and drain each section of the strainer. Each strainer compartment should be provided with a removable bar magnet system to collect iron s articles in the strainer. Regular observation of the strainer basket and magnets can provide one of the earliest waxnings of trouble in the system.
788
MARINE ENGINEERING
Coolers are provided to remove heat header, some means must be provided to prevent oil from the oil; normally seawater is the cooling medium. starvation of any single point either as the result of The cooler is designed to cool the oil a t full-power excessive design clearance or a bearing casualty allowing operation and at maximum seawater temperature and excessive oil flow to occur. The details of orifice design maximum tube fouling. Single-screw ships are recom- are thoroughly covered in reference [29]. Virtually all mended to have two coolers and multiple-screw ships of the precautions cited as being significant in tho should have one cooler per shaft. Two separate means application of orifices as measuring devices are applicable of circulating water through the coolers are required to their use as flow controls. by reference [17]. Drains, vents, and isolating valves Generally, the pump capacity is about 25 percent in are needed in both the oil and watersides of the coolers to excess of the bearing requirements to allow for changes in permit cleaning, leak detection, and repair. flow and other contingencies. In systems pressurized The coolers are usually operated on a bypass with by a pump, the pressure in the system is controlled by sufficient oil being passed through the cooler to maintain recirculating the oil flow excess back to the sump through the oil supply temperature within the design range. The a back-pressure control valve. This bacl+pressurc temperature-regulating valves should be arranged so that control valve is connected to the system downstream of the oil flow to the system cannot be shut off. Tempera- strainers and filters so that the system pressure a t the ture monitoring should be provided a t the inlet and outlet bearings is maintained at the desired pressure irrespectiuc of the oil cooler. Provisions should be made for supply- of strainer and lube-oil-cooler pressure drops. An excesing steam at 5-psi maximum to the waterside of the cooler sive pressure buildup at the pump is avoided by the usc in order to heat lubricating or cleaning oil for initial or of a pressure relief valve a t the pump. subsequent cleaning of the lubricating-oil system. Suction piping for pumps should be sized to prevent The oil pressure in the cooler should be maintained a t a cavitation at the pump inlet. The oil requirements of higher pressure than the water pressure to minimize bearings increase rapidly at startup and this sudden saltwater intrusion into the lubricating-oil system in the demand can upset the balance of the system. Even event a leak occurs. Corrosion protection to the water- positive-displacement pumps can temporarily disrupt side of the cooler should be provided by the inclusion of a submerged suction supply a t startup if the suctiori zincs or other electrolytic protection. piping has been made too small. Pump suction piping The allowance for pressure drop through both the and pump suctions should be submerged in oil to ensura water and oil sides of the cooler must be established for that the pump will immediately take suction upon the full range of design conditions. The water pressure startup. If the pump suction and piping cannot bo drop should be established for clean and for maximum submerged in oil, an oil bleed from the oil header, that allowable fouling conditions. Oilside losses must be is provided with a check valve, should be installed to known to establish overall system pressure drops at the ensure that the suction piping is maintained full of oil a t maximum and minimum operating oil temperatures. a pressure above atmospheric. See Chapter 14 for additional discussion regarding Drain piping is sized on a different basis than supply coolers. piping. Drain lines are designed to run only about 3.7 Piping and Valves. The lubricating-oil piping one-half full in order to help separate air from the oil. system must be designed to deliver the required amount Booser and Smeaton [3] give the following formula for of lubricant under all reasonable conditions with the the minimum slope of a drain line running half full. minimum available head. In addition the piping should be designed with the 6bjectives of providing some separation of entrained air in the drain piping, simplifying the initial cleaning and subsequent cleaning, preventing the cracking of piping due to the vibration of unsupported where lengths, and preventing inadvertent maloperation of the s = slope of drain line, in./ft system through errors in valve operation. Q = oil flow, gpm The objective of securing the required quantity of oil r = oil kinematic viscosity, Centistokes with the minimum head can only be obtained by the d = pipe inside diameter, in. most careful study of system pressure drops over the full required operating range. Lubrication piping systems 3.8 Lubricating-Oil Purifler Subsystem. The lubriconsist of a number of series and parallel flows through cating-oil purifier subsystem employs the centrifugal pipes of varying diameter and throttling valves that are force in the purifier to remove very small particles [tho changed to meet the required conditions of operation. duplex strainer is expected to pass particles up to 150 The pressure-drop analysis starts from the pressure and microns (0.006in.)] from the lubricating oil. The actual flow required through each bearing and gear spray for particle size passing through a purifier depends on tho the maximum speed conditions. Orifices are usually thickness, longitudinal velocity, and viscosity of the oil provided to restrict flow to that needed at the machine layer containing the particle, the magnitude of the cenelement to be lubricated. Where a number of parallel trifugal force, and the mass difference between tho points are to be supplied with lubricant from a single particle and the oil volume displaced.
BEARINGS AND LUBRICATION
3.6 Coolers.
789
A purifier capacity of a t least 10 percent per hour of head requirements and in elevation between the most the in4ervice oil charge is recommended by reference remote bearing and other bearings in the system.
L I
[21]. Operation a t reduced capacity will take out finer pmticles than operation a t design values. Since the rate of oil deterioration is much higher for turbogenerators than for the main turbine, the purifier is arranged to take suction and discharge to each of the main tanks and auxiliary sumps in such a way that the possibility of cross oontamination is virtually eliminated. Separate supply 8nd return connections are located on the tanks to maximize the flow distance within the tank between purified and dirty oil to prevent recycling of pursed oil. The heaters must have the capacity of maintaining the lubricating-oil inlet temperature to the purifier between 160 and 180 F when operating a t a normal rate. The purifier oil heater requires the same precautions to limit mrface temperatures as were noted for the main lubrioating oil heater. 3.9 Condition Monitoring. Condition monitoring of a lubrication system provides information regarding the functional status of the system components, the lubrioant, and the bearings serviced by the system. Information should be supplied a t a rate and in sufficient depth that reliable performance is ensured. Since the condition of the lubrication system is essential to $he avail* bility of the main propulsion plant, an indication of aystem performance should be provided to the ships oommand and control system. Where staf3ng is not a problem, the report of the watch stander may provide sufficient assurance to meet command needs. Where the manning of the engineering space is reduced, an increased depth and sophistication of the condition monitoring input to the command and control system are required. In designing the condition monitoring system, there is a basic selection of passive and active systems. Passive systems present data for visual observation either a t the point sensed or a t a remote point; recorded data (manual or machine) belong in the class of passive systems. Active systems compare the data to some limits and provide an alarm or a change in operating conditions or both in the event that the data exceed the specified limits. Both passive and active systems may be designed to either continuously monitor the data or intermittently scan the data a t a prescribed rate; the sampling rate is chosen to be compatible with the datahandling response rates. Lubricating-oil pressures and temperatures are the characteristics most widely monitored to obtain an indication of eystem performance. By monitoring the pressure at the most remote bearing, assurance is obtained that the system head is being maintained; however, it is recognized that the proper pressure a t the most remote bearing does not necessarily ensure that oil is flowing to all bearings. If the pressure a t the most remote bearing drops below a preset limit, an alarm sounds and, in addition, the low-pressure signal may start the standby lubricating-oil pump. Low-pressure alm settings ahodd be made taking into &ccount the dieerence in
A pressure measurement a t each pump confirms pump performance. Low pressure a t a pump may indicate an excessive oil temperature, piping leakage, suction difficulties, or internal pump wear. An abnormally high pump pressure may indicate a !ow oil temperature or an obstruction downstream of the pump. A simultaneous measurement of both oil pressure and temperature aids greatly in defigng the true problem. A measurement of the differential pressure across strainers is used to monitor the increase of resistance as deposits of dirt are built up, and such measurements indicate the need to shift strainers. An excessive pressure drop across the waterside of coolers usually indicates excessive fouling and the need to clean the tubes a t the eext convenience. Positive assurance that oil is reaching a bearing is ~rovidedby the use of a sight-flow fitting, often called a bubbler. The oil supply to the sight-flow fitting is taken from the bearing discharge; through a g l window ~ in the sight-flow fitting, the watch stander can observe the stream of oil being discharged from the bearing. Where the bubbler is fitted on top of a turbine or gear bearing it is often necessary to bypass enough oil from the bearing inlet to provide the head necessary to raise the oil to the top of the bearing cap. Sight-flow fittings operating on bypass oil indicate the excess of inlet oil over that being discharged through a bearing. The flow is thus a maximum when the bearing fipw is at its minimum and decreases as the bearing flow increases. The requirement for sighting of the bubbler by the operator, the absence of a quantitative relationship between the oil flowing in the bearing and the flow in the sight flow, and the absence of a remote alarm capability are other disadvantages of this system. Sight-flow fittings are frequently fitted with thermometers to provide an indication of bearing discharge temperature. Temperature is one, if not the best, condition indicator. Temperatures are usually measured in the lubricating-oil sump, a t the outlet of the oil cooler, and a t the discharge of each bearing. Sudden changes in temperature usually indicate a situation that requires investigation to ensure that the cause will not result in system damage. Since the success of a monitoring system depends to a large degree on the amount of time provided to take corrective action, the rate of response of the sensor is of critical importance. The rate of response of a sensor can be measured on the basis of the time to respond to a step increase. The characteristic time of a sensor is defined as the time required for %be sensor t o rise to 63 percent of the amplitude of a step increase. In a period of three times the characteristic time, the sensor will indicate more than 95 percent of the true amplitude of the step increase. Typical characteristic times are from milliseconds to seconds for thermocouples, from three to ten seconds for resistance thermometer elements and thermistors, and t n to thirty seconds for glass thermometem. Glass [herrnorneters amd many metal ther-
790
MARINE ENGINEERING
CHAPTER XXI mometers are installed in thermometer wells that greatly slow down their response. Thermocouples and resistance elements may be installed in the babbitt layer of the bearing to shorten response by locating the sensor close to the source of heat. Alarm points are set about 25 deg F above the normal operating temperature measured by the sensor during maximum-power operation. There is no clearly optimum type of temperature sensor. Thermocouples, resistance thermometer elements, thermistors, and glass and metal thermometers have all been used as lubricating-oil temperature sensors.
Surface Profile on the Load Capacity of Thrust Bearings with Centrally Pivoted Pads," Trans. ASME, vol. 77, 1955. 15 B. Sternlicht, J. C. Reid, and E. B. Arwas, "Review of Propeller Shaft Thrust Bearings," ASNE Journal, May. 1959. 16 A. J. H. Goodwin, "The Design of a Resonance Changer to Overcome Excessive Axial Vibration of Propeller Shafting," Trans. IME, vol. 72, 1960. 17 Rules for Building and C l a s h Steel Vessele, American Bureau of Shipping. 18 "Turbines, Steam, Propulsion, Naval Shipboard, " Military Specification MIL-T-17600. References 19 "Recommended Practices for the Design of Marine 1 Dudley D. Fuller, Theory and Practice of Lubri- Propulsion Turbine Lubricating Systems, " Joint ASTMcationfor Engineers, John Wiley and Sons Inc., New York, NEMA-ASME Committee on Turbine Lubrication, 1956. ASME Standard 111, New York, ASME, 1960. 2 0. Pinkus and B. Sternlicht, Theory of Hydro20 "ASTM-ASME Recommended Practices for the dynamic Lubrication, McGraw-Hill Book Co., New York, Flushing and Cleaning of Marine Propulsion Turbine 1961. Lubricating Systems, " Joint ASTM-ASME-NEMA 3 Standard Handbook of Lubrication Engineememng,Committee on Turbine Lubricating Systems, ASME J. J. O'Connor and John Boyd eds., McGraw-Hill Book Standard 113, New York, ASME, 1964. Co., New York, 1968. 21 "Recommended Practices for the Purification of 4 Proceedings International Symposium on Lubri- Marine Propulsion Turbine Lubricating Oil," ASME cation and Wear, D. Muster and B. Sternlicht eds., Standard 114, New York, ASME, 1964. University of Houston, 1965. 22 "Recommended Practice for the Design, Opera5 H. C. Rippel, Cast Bronze Hydrostatic Bearing tion, and Maintenance of Marine Auxiliary Machinery Design Manual, second edition, Cleveland, Cast Bronze Lubricating Systems," ASME Standard 115, New York, Bearing Institute, 1965. ASME, 1968. 6 D. F. Wilcock and E. R. Booser, Bearing Design 23 "Recommended Practices for the Design of Steam and Application, McGraw-Hill Book CO., ~ e ~wo i k , Turbine Generator Oil Systems," ASME Standard 116, 1957. New York, ASME, 1968. 7 H. F. P. Purdy, StreamlineFlow, Constable & Co., 24 "Recommended Practices for the Cleaning of London, 1949. Steam Turbine Generator Oil Systems," ASME Standard 8 Mayo D. Hersey, Theory and Research in Lubri- 117, New York, ASME, 1968. cation, John Wiley and Som, Inc., New York, 1966. 25 "Recommended Practices for the Purification of 9 H. C. Rippel, Cast Bronze Bearing Design Manual, Steam Turbine Generator Oil," ASME Standard 118, Cast Bronze Bearing Institute, Cleveland, 1965. New York, ASME, 1968. 10 H. C. Rippel, Cast Bronze Thrust Beaping Design 26 "Recommended Practices for the Flushing and Manual, Cast Bronze Bearing Institute, Cleveland, 1967. Cleaning of Marine Auxiliary Machinery Lubricating 11 A. G. M. Mitchell, Lubrication: Its Principles and Systems," ASME Standard 119, New York, ASME, Practice, Blakie, London, 1950. 1968. 12 B. L. Newkirk and L. P. Grobel, "Oil Film Whirl 27 "Marine Machinery Lubrication, " The Texas -A Non-Whirling Bearing," Trans. ASME, vol. 56, Company, 1959. 1934. 28 J. B. Poole and D. Doyle, Solid-Liquid Separation, 13 0. Pinkus, "Solution of ~a~ered-p and Sector Ministry of Technology, Her Majesty's Stationary Thrust Bearing," ASME Paper 57-A-152, December Office, London, 1966. 1957. 29 Fluid Metere, ASME Research Committee on 14 A. Raimondi and J. Boyd, "The Influence of Fluid Meters, ASME, New York, 1959.
i W. 0. Nichols
1 Automation
Section 1 Auton~alior~Systenls 1.1 Introduction. Controls are defined as those variously designed devices provided to direct, govern, or influence the operation of machines. The control device is therefore the interconnection between man and the machine. The interface between man and the control/machine complex is the point where man exercises the commanding direction or influence upon the maohine's activities. The progress of technical achievements has permitted this interface to become more and more remote from the basic machine. Control systems have increased in complexity and extent to the point where they are recognized under a separate broad cat* pry termed "automation." In the broad sense the "automation packageu on a vesseL consists of all controls that are installed regardless of whether they are for propulsion, cargo handling, navigational control, et al. In this case the entire ship is the control/machine complex and the degree of automation depends upon the extent to which human judgment and effort are used in the control process. For the aake of simplicity, the ship may be subdivided into its various functional systems; however, it is important to retain recognition that each system is a part of the total complex and its relation to the whole must be considered if the degree of automation applied is to be effective. To be effective the applied automation must accomplish a worthwhile objective. It is common to think that reduction of operating personnel is the prime objective of automation. Actually this is only one of the many possible purposes for automation. Operability is another. Some functions could not be acwmplished without it; a good example is the control of an electric generating set. Automation may also provide increased efficiency and safety. All of these factors ultimately affect overall operating costs; so, as in the case for so many endeavors, economics becomes the primary motive force directing the desire and controlling the extent of automation. The determination of the functions to be accomplished through automation and the coordination of subsystem control into a total system concept is the work of the marine engineer. He must provide the lcnowledge peculiar to the marine environment, establish the basic design parameters, and exercise the selective action (both technical and economic) that will serve to produce the practical installation desired. In these days when
/'
!
technical capability provides little or no restriction, the selective function takes on paramount importance if "overdesign," excessive cost, and undue complication are to be avoided. Reliability and maintainability are essential determinations in the evaluating process. The overall success of an automated system or plant depends on more than merely the control system itself. The substitution of mechanization for the manual function can be accomplished through several approaches. In one approach, the function may be eliminated through component or system design; this means simplification and with simplification increased reliability usually follows. Alternatively, the frequency of the manual function may be reduced to the point where the usage of mahpower is not objectionable, thus again eliminating the need for automatic controls. The provision of automated features should proceed after consideration has been given to the preceding approaches. In summary, control system design cannot be divorced from overall system or plant design. For the accomplishment of a specific task, the knowledge of the marine engineer should overlap that of the control designer and supplier. It is essential that a common ground of understanding exist between the two ~ a r t i e sconcerning desired functions, operational demands, machine limitations and performance characteristics and control characteristics. This is desirable whether the control is a simple automatic valve or a complex subsystem. 1.2 Control System Types. A control system is a system which measures the condition of some entity and, with this information, governs the state of a variable; i.e., speed, temperature pressure, position, etc. Control eystems may be classified by four major characteristics: 1 System configuration (open or closed loop) 2 Type of operation (manual or automatic) 3 Type of copmand,input (constant or variable) Type of primary control medium @neumatic, electronic, etc.) The open-loop system is the simplest form a control system may assume. It is distinguished from the closedloop system by the lack of an input which measures the state of the controlled variable; such an input is called, a feedback. 79 1
i
1
792
MARINE ENGINEERING
1
CONDENSATE PUMP
v
CONDENSER
I
FLUID OUTPUT AT DESIRED TEMPERATURE
FLUIW TEMPERATURE SENSOR
FEED BACK (1 AND 5 ) CAPILLARY TUBE
HEATED FLUID CONTROL-LED VARIABLE
HEATING MEDIUM
ERROR SIGNAL(3) DIAPHRAGM POSITION
HEAT EXCHANGER
Fig. 1
VALVE ACTUATING MECHANlSM(4) (STEM)
\
Diagram of open-loop system
\HEATING MEDIUM
L CONTROL
:."i."T FLUID
I
I
,
VALVE
Figure 1 is a diagram of an open-loop concept. It deFig. 2 Diagram of closed-loop control system picts a feedwater control system that was popular in the days when Scotch boilers were used. The flow of water to the boiler was controlled bv adjustine the volume output of the feed pump in reiatioh to ;he change in The diagram of a simple temperature regulating dewater level in the feed tank. A rise in the tank water vice shown in Fig. 2 illustrates the closed-loop concept. level was an indication of a greater condensate return In this example, function 1is accomplished by the sensor and therefore also indicated a higher steam and feed(thermostatic element) in the heat exchanger discharge water demand. The float-actuated feed pump control varying the force on a diaphragm through the capillary utilized this input information to increase pump speed tube. The command input, function 2, is introduced as a and thus compensate for the increased steam demand. spring position transmitting a force to the opposite side Although the prime controlled variable was the drum water level, this information was not fed back to the of the diaphragm. If these two forces are equal (zero error) the system is in equilibrium and no motion is imcontrol system. The system worked satisfactorily beparted to the valve. However, if the sensor detects a cause the ratio of water volume in the boiler to feed flow change in temperature, the pressure in the diaphragm was high. chamber will also change. This difference is the error The open-loop system can be made to work effectively signal, function 3. This signal is transmitted to the provided it is properly applied. The controlled equip- valve through the stem, adjusting the flow of the heating ment must be such that its inherent characteristics promedium as required, function 4. The effected change is vide the balance needed to maintain operation within sensed at the exchanger discharge, function 5, and thus -acceptable limits. In the modern ship the open-loop the loop has been closed. system has practically disappeared. Generally, modern The described system operates automatically; howmachinery requires that the controlled variable be mainever, a system does not have to be automatic in its oper* tained at a tolerance closer than that possible with an tion to be classified as a closed-loop system. A manually open-loop system. operated valve and thermometer could serve the same In contrast to the open-loop system, the closed-loop fundamental purpose. Here man provides the logic system works directly with the controlled variable. mechanism to make the comparison between actual and The basic functions performed 'by a closed-loop system desired conditions as well as the force to activate the are: valve. Thus it can be seen that the term "closed-loop system" designates a broad category that can be further 1 Detect actual condition 2 Provide reference for desired condition (command divided into manual and automatic systems. A manual system is .distinguished by the fact that the input) 3 Compare actual and desired condition (error signal) control action (output) is obtained only when man 4 Activate control mechanism to reduce error (out- recognizes a deviation from the desired condition and activates the control mechanism. The system being put) dependent on man is therefore subject to his limitations. 5 Detect effect of control action (feedback)
The degree of control obtained depends upon the individual's ability to recognize and respond to off-normal oonditions as detected by the instrumentation. Man's ability to sense and respond to a change is generally poor, therefore system stability should be high and response to a change in output should be slow in order for manual operation to be effective. Man is not constituted to be continuously on the alert; this factor further limits the effectiveness of manual operation. Manual control can be applied only where the demand for control action occurs at a low frequency when considered over an extended time frame. There is no physical limitation regarding the relative location of the command input and aontrolled variable since various mechanisms can be used to provide motion and instrumentation from any distance. These remote manually operated mechanisms are an integral part of the "automation package"; the ~ystemelements do not have to be "automatic" to be a part of the automation package. Examples of manual control systems in marine use are ehip's steering (nonautomatic), bridge and engine room throttle controls (positioning type), and fuel-oil tank filling and transfer systems. In many cases manual systems fulfill an important need in that they are inatalled to provide redundancy for automatic systems. Automatic systems, as the term implies, are those which function without manual action. Of course, man must set the desired command input and initially activate the system. The system must do the same things that the man would do if it were a manual system. However, since control devices can respond more rapidly, with greater consistency, and without fatigue, they have a great advantage over the manual system. Automatic Bystems are required where one or more of the following conditions exist : 1 Rapid and/or precise response is essential. 2 Continuous surveillance is required. 3 Manpower is qot available t,o serve the function. Examples of automatic systems in common use in marine applications are boiler combustiop and feed regulation, steering by gyro pilot, electric generator frequency and voltage controls, and the many varied fluid flow regulating devices for controlling temperature, pressure, levels, etc. Systems that attempt to maintain the controlled variable qt a substantially constant value (constant command input) are defined as regulating systems. The temperature regulator shown in Fig. 2 is such a system. In this type of system, the controlled variable will deviate from the command input (set point) us the demand changes across the operating range. This comes about by the fact that the control mechanism is activated by an error signal. Since t,he set point remains constant the error signal will vary in direct proportion to a change in the controlled variable. I t follows that although the set point remains constant, the controlled variable will deviate from this set point in direct proportion to the demand (i.e., an error must exist in order for there to
be a response). This deviation from the set point across the operating range is called the proportional band. Some controls have a means for adjusting this proportional band. Control stability is affected by the band; if it becomes too narrow (large change in error signal for a small change in demand), hunting may occur. On the other hand, if ,it is too narrow (small change in error signal for a large chaLge in demand), adequate control may not be provided. The actual deviation from the set p e t can be reduced through the use of relays where #the output signal can be amplified and varied proportionately to the input. Where the command input is a variable and a corresponding proportionate change in the controlled variable is desired, positioner systems are used. As in the regulator system, a feedback supplies the comparison to obtain an error signal. However, in this case, the mechanism is caused to be motivated until the error signal is reduced to zero. The ship's steering gear is a classic example of the positioner type system. The variable input is introduced through the steering wheel or gyro pilot, and the rudder position is fed back to the operating mechanism. The difference between the two signals is the error and the steering gear turns the rudder until this error signal is cancelled. The fourth classification (type ~f primary control medium) reflects the type of hardware involved. This classification is perhaps the one most commonly employed; the terms mechanical, pneumatic, electric, electronic, hydraulic, etc. are"prevalent1y used. Each of these types of systems has its advantages and disadvantages and there is a degree of overlap between the various system types. This usually results in the necessity for compromise when the system evaluation and seleation are made. The transmission of motion is required in almost every control system and, provided the distance is limited, the mechanical mechanism has an advantage. It is simple, direct, and precise provided backlash is minimized. Transmission of motion over an extended distance introduces the problem of hysteresis-type distortion from bending or torsional deflections. The centrifugal (fly ball) governor with direct operation of a steam valve as shown in Fig. 45 of Chapter 5 is perhaps one of the best examples of a purely mechanical control system. Most control systems have some mechanical action even though the primary control medium is of a different nature. Pneumatic systems are used extensively for fluid flow control. They permit a higher degree of freedom in the location of input and outpw-t than the direct mechanical system. One important advantage is the ability to store the energy in tanks to permit continued control action through temporary power losses. The ability to exhaust the used air directly into the ambient space allows simplification relative to other fluids. Pneumatic systems do, however, have disadvantages. There is a practical limitation in the power output available and transmission losses must be considered
MARINE ElMGINEERING
where extended distances are involved. The compression of large volumes of air a t high pressures involves high energy input and costly equipment. For this reason (and also to permit transmission through relatively simple low-pressure tubing) 125 psig is the maximum control air design pressure recommended. The use of air also presents other problems; it must be properly treated since dirt, oil, and moisture are detrimental to control action. The use of pneumatic equipment in the weather and unheated areas is generally avoided due to the potential problem of freezing moisture. The electrical and electronic fields are closely related; as a matter of fact, the division between the two is quite often not easily established. These two fields combine to provide the greatest range of application both in control and information systems. The degree of sophistication possible defies the imagination; however, it also often leads to a system that is in excess of that actually required to accomplish the desired task. For example, an off-limit condition may be alarmed simply by a relayoperated bell responding to the closing of a contact. On the other hand the condition may be sensed by a voltage variation and converted electronically to sound the bell, display the value digitally, print it out by teleprinter, introduce it to a computer to initiate corrective action or store it for future reference. There would obviouslv be a considerably disparity between the cost of the foimer and latter. Advantages provided by electrical/electronic systems include the capability of long-distance transmission and adequate "muscle" for most control operations. The components are small and a great deal can be accomplished within a limited space. A high degree of precision is possible if demanded by the application. Electrical/electronic systems also have disadvantages however. It is generally essential to protect electronic components from voltage surges (spikes) that are likely to occur in most ship electrical supply systems. Additionally, some care must be taken to ensure that electrical and electronic parts are in a proper environment, particularly from a heat and humidity standpoint. This is especially important in th'at the loss of electrical power results in an immediate loss of control. The system design must recognize this disadvantage and react to maintain a safe condition. The application of batteries to store potential energy for such an occasion is limited in practicality. Furthermore, experience has shown that- systems containing storage batteries are prone to be neglected for extended periods as a result of their infrequent use, and consequently when required to perform, they are incapable of doing so. The repair and maintenance of electrical and electronic systems usually requires special skills not always found in operating engineers; however, the "plug-in-replacement" component concept has proven to be an effective means of alleviating this problem. Hydraulic systems may be used to both control and power a system. This is perhaps a primary advantage of hydraulic systems. Hydraulic fluid can be used where
AUTOM
the explosion hazards of electrical equipment or the moisture problems of penumatic systems preclude their use. In general, hydraulic equipment requires a degree of system cleanliness that is in excess of that required for the majority of shipboard systems. The pressures involved are usually high and therefore the components are of heavy construction and require a high level of inspection and test. 1.3 Control System Design. The design of a system and its components to perform a specified control function is a specialized field. The problems involved are peculiar to each type of system and to the particular purpose it serves. The field is broad and much has been written concerning the design fundamentals and procedures involved. For these reasons, the following discussion is confined to the considerationswhich the marine engineer must entertain when applying control systems. The marine engineer must recognize that the systems are dynamic in nature and subject to analysis through application of the laws of motion. A'lathematical models can be developed to approximate control system operation and predict response, stability, etc.; the mathe matics involved parallels that, associated with other motion studies such as vibrations. Differential equations, set up to represent the motion, may be solved generally through the use of transforms. A detailed treatment of the various analytical techniques involved is contained in texts such as reference [I].' For a comprehensive treatment of the terminology used with automatic control systems, see reference [2]. The marine engineer must specify the system environment, characteristics of the controlled machine, and tho desired response. Tabulated below are representative particulars that should be specified. System Environment 1 Operator (crew) capability for operation, maintenance and repair activities. 2 Operator (crew) availability (number, location, duties). 3 Ambient conditions (temperature, humidity, cleanliness, corrosion and explosion hazards, etc.). 4 Relative location of control system components. System (Machine) Characteristics 1 Interrelationship of controlled events. (a) dependent and independent operations (safety 2 interlocks, etc.) (b) requirement for logic sequencing (c) requirement for computed outputs 2 Characteristics of controlled variable. (a) number and characteristics of inputs (b) system response (time factors, stability, etc.) (c) permissible deviation from ideal conditions (pregsure, temperature, force limitations, etc.) 3 Effect of failurc.
Numbers in brackets designate Refererices at end of chapter.
!
!
Desired Response 1 Operator participation. (a) manual or adtomatic (b) performance monitoring 2 Control reliability. . (a) redundancy (b) test facilities (c) effect of failure 3 Number and characteristics of outputs. 1.4 Information Systems. Automation not only oonsists of control systems but also includes information nystems which serve the parallel function of monitoring rystem or plant performance. The monitoring function may serve a number of purposes; specifically it may 1 Indicate values of controlled variables to guide remote manual control operations. 2 Indicate values of controlled variables under automatic control. 3 Warn of off-limit conditions. 4 Warn of incorrect operating set up. 5 Provide information for trouble-shooting. 6 Provide record of performance. The instruments required to guide remote control operations should be of the continuous reading type in order that response may be immediately recognized. Instruments which reflect long-term trends (recording units) are not generally suitable for manual control. Where automatic control is involved, a continuous display may not be required; however, in many cases a redundant manual control is provided which may require limited continuous display. In this case a visual and audible alarm may be provided to warn of off-limit conditions and, associated with this, an instrument which will provide the value of the controlled variable upon demand. The latter is normally arranged to indicate a number of different variables through selective switching arrangements. The warning of off-limit conditions or an incorrect operating setup is normally announced by both visual and audible means. The visual signal is usually a flashing light which converts to a steady light when the alarm is acknowledged by the engineer. This acknowledgment also serves to cancel the audible signal. There are two ways in which the results of corrective action may be handled. With one method, the indication of an offlimit conditioli automatically cancels when the condition returns to normal; alternatively, the indication may be designed to persist until the fault is both cleared and recognized by the operator. Note that the latter may require two actions on the part of the engineer: one to acknowledge that the fault exists and silence the audible signal; and another to cancel the indication after corrective action has been taken. The question of which method to apply is somewhat controversial and its selection depends largely on the personal opinion of the operator and the manning a t the control station. The indication of an alarm condition serves to direct operating personnel to the problem area but does not
serve further in the trouble-shooting action. Corrective action must be accomplished through analysis of the instrument indication and visual inspection by the operator in the normal manner. However, where the control station is unmanned, assistance can be given by recording the off-limit condition or conditions and the s quence of their occurrence. An information system should be arranged, where practicable, to be self-monitoring. This requires that a failure in the syatem produce readily recognizable results. In the case *here alarms are involved, this is generally accomplished by sounding an alarm. Recorded performance conditions may be presented either by digital or analog display. ,Where the recording instrument serves a number of varied inputs, it is referred to as a data logger; where only a few inputs are involved it may be called a recorder. The digital display must, by its nature, be intermittent and record either on an automatic time cycle or upon specific demand. The recorded information is the instantaneous value at the time of recording; and operating trends must be obtained by an evaluation of a series of readings. The analog instrument provides a continuous graphical display and therefore variations in the controlled variable are more easily detected. However, under relatively steady-state conditions they may have the disadvantage of excessive paper usage, in which event the recorded data may have limited value. As noted in the foregoing, neither of these two types' of instruments is generally suited to provide information for direct contro,l.,-The manner employed to present the information to the operating personnel is an important consideration in automation. The reduction of personnel associated with automation generally means that the plant or system is operated by no more than one or two individuals. It is therefore necessary to centralize the information and remote controls a t a single location. The location of the control center will depend upon the manning arrangement and type and layout of the system (propulsion, liquid cargo handling, etc.). If the center is to be normally unattended during operation, it is desirable that it be in a location that is readily accessible since, except for routine checks, attendance will only occur under off-limit conditions. I t is also desirable that where possible the center be adjacent to the controlled machinery to permit direct surveillance. Direct surveillance is, of course, also desirable if the center is to be manned during operation. It is generally recognized that direct observation is desirable where certain hazards, such as fire, exist. The fiying aisle in a steam power plant or the deck of a tanker foT cargo control are eximplea of cases in which visibility is desirable. The design of a control center should follow the principles of human engineering [3-51. These principles define the criteria essential to establish the environmental conditions under which man will function most efficiently and consistently. Those conditions which serve to keep him alert, such as freedom from excessive vibration. proper illumination, low noise level, and good ventilation,
7
MARINE ENGINEERING
Fig. 3
AUTOMATION
Fig. 4
Vessel 0--control center console
are prime requisites. Without these conditions fatigue soon takes over, resulting in deterioration of the decisionmaking capability which is the one major function man can provide in a manner superior to the machine. Man normally utilizes all the senses of sight, feel, smell, and sound to evaluate the performance of machinery. When he is removed from direct use of these senses, substitutions may be necessary. For example, where control stations are remote from the operating machinery, devices sensitive to rate of temperature rise or excessive vibration may be required to assist in the evaluation of the operating condition. Information. presentation should provide a logical display and permit rapid assessment of the operating condition. Sectionalizing and subsystem grouping are important. Continuously displayed performance information can best be surveyed when mounted on a vertical surface; this permits a general assessment of the situation to be made from a position some distance from the board. Remote control actuators may be located on
Vessel 0-mimic
display
a sloping surface since, here, a specific rather than overall view is required. Control-actuating devices performing similar functions should operate in the same manner; they should also be readily distinguishable from purely indicating devices. Rapid assessment of operating conditions can be assisted if the instruments are arranged to form a specific pattern under normal conditions (e.g., lights form a straight line, pointers are all in the same general direction). Instrumentation may also be arranged in the form of a diagram (mimic) representing the system. A mimic diagram is shown on a console in Fig. 3 and is shown in detail in Fig. 4. This arrangement requires more console space but is particularly advant* geous in complex systems involving the transfer of fluids from a variety of sources to a variety of discharges. The presentation of the information should be aN simple and logical as possible. The time required to understand the control function should be minimized; this is particularly important in the marine field whero rapid crew turnover is sometimes the case.
Section 2 Applications 2.1 General. The techniques of automation can be applied to almost any phase of ship operation. The extent to which they are applied is a matter of judgment primarily made by the owner/operator of the vessel. The economic gain pbtained through manning reduction and increased efficiency must be balanced against the factors of first costs, plant reliability, maintainability, and operating costs. The solution is quite complex since the factors of efficiency, reliability, and maintainability are often difficult to assess and quantify.
This situation naturally leads to a wide variation in not, only the extent of automation applied but also in tho type of equipment used in automatic control systems. All of the various systems and types of control and information components used are not discussed in t h i ~ chapter; instead, the discussion is limited to certain examples of the shipboard systems which have been automated and the general principles involved. The propulsion plant is probably the system that has received the greatest attention; also, the automation of the pro-
f
pulsion plant has had a major effect on ship operation. However, cargo handling, mooring, food handling, fuel transfer, and many other shipboard systems have been automated to some degree and have also made a contribution to the reduction of the operating burden. All ship installations have, to some extent, control and information systems to assist their operation. Therefore a good question is: "When does automation begin?" Obviously it is a purely arbitrary point which may depend to some extent upon the installation in question. If one were forced to establish such a point, one criterion might be when the operation of a major component is made sufficiently automatic to no longer require operating personnel. Another might be when operating functions are centralized. These two criteria seem to fit most installations; no doubt there are others that can be applied. 2.2 Regulatory Body Requirements. Many regulatory bodies associated with the marine field have established recommendations for the midance of designers in develo~ingautomated installations 16-11]. ~irthermore,the au6mation of a propulsion machinery installation is sufficiently important for classification societies to establish minimum reauirements for class record and designation. Although ihe numerous bodies have varied interests which to some extent are reflected in their recommendations, the feature of safety appears to be a common denominator. The safetv features encomand safety pass the safety of the ship, safety of of the property within the influence of the ship's presence. Safety is approached in several different ways. Integrity of the control is perhaps one of the most important. Reliability is stressed through good component design and redundancy of vital elements to insure continuity of the control function. The desirability for self-monitoring features in control and information systems and the ability for rapid correction of faults are recommended features. I n view of the limited manpower on an automated vessel. safetv considerations cannot be limited to the controlled machinery alone. Fire detection is an important feature and the methods used, as well as the extent of the system, vary depending upon the arrangement and manning level. The detection methods available include sGtems which have sensors sensitive to smoke, rate of temperature rise, and ionized particles. Each type has its limitations, and location of the sensing element is particularly important. In an installation intended to operate under unattended watch, it may be necessary to provide a detecting system which utilizes all three methods. Another auxiliary system required for safety is flooding control. A high water level in ,the bilge may be simply alarmed; or, the bilge pump may be arranged for automatic operation. Of course the system can be further sophisticated to w a n of excessive frequency or extended periods of pump operation; however, such refinements are not usually considered essential. To insure that all systems are in working order and properly maintained, test procedures are established and
-
periodic inspections are made. These further serve to prevent unsafe operating conditions. 2.3 Steam Propulsion Plant. It has been generally accepted that where the control and information systems of a steam propulsion plant are sufficient to reduce the watch standing personnel to two or less, the plant is to some degree automated. Thus three levels of engine room automation (two men, one man, and unattended) are open to consideration. Only the general aspects of these alternathe considerations can be discussed since system detafls depend upon the specific plant/llesign, of which there are many. A reduction in manning is one of the objectives of automation, and the degree of automation, is often expressed in terms of the number of watch standing personnel; however, the actual number of crew members does not decrease in proportion to the number of watch standers. As the number of watch standers decreases, it may be emected that there will be an increase in the numbkr of personnel assigned to daily tasks. The ultimate, of course, is the unattended watch where the crew is on duty for an eight-hour period during the day. The machinery space is unattended for the intervening sixteen hours; however, a licensed officer usually is assigned the responsibility of each "watch" and responds to alarms which sound during his assigned period. The daily workers' tasks become primarily that of mainte nance rather than operation. It therefore becomes evident that automation not only affects the design of the machinery but also the type of personnel and the underlying philosophy of the ship operational procedure. As a general consideration, the propulsion plant control system should be designed so that the engineer retains responsibility for the satisfactory operation of the machinery and the watch officer on the bridge remains free from machinery performance considerations. The complications involved in starting a cold plant have dictated that the starting procedure be accomplished under direct manual control with gradual transfer to the automatic features. First, consider the steam generating plant since its automation is generally common to all three manning levels. Control systems of the nonautomated plant have included combustion controls, feedwater regulation, and to some extent power-actuated soot blowers. Informatioh systems have been limited to water level indication and a set of direct-reading gages to indicate steam, oil, and air/gas conditions. The major features that have been added for the automated boiler are widerange burner capability, steam temperature control, and safety shutdowns and albrins. Wide-rahge burners are used with two major variations in control concept. One system depends upon the vari* tion in fuel flow only and all burners are in continuous use. The other system involves the automatic lightingoff and securing of burners as well as utilizing the wide range capability of the burners. The latter arrangement provides for a wider range by reducing the minimum fuel flow capability to that of one burner. It does, of course,
798
AUTOM
MARINE ENGINEERING STATUS INDICATION
READY TO START
I
2
procedure stops and sounds an alarm. Indication may also be given that the failure to start was caused by low oil pressure. The gland seal control valve is likewise given a specific time in which to open. If the position sensor faas to detect an open valve in the allotted time, the sequence stops and the alarm is sounded as in the case of low oil pressure. If the sensor detects that the gland seal control valve is open, the sequence continues and the pump suction valve is opened. In this case both position and pressure are checked before the system is cleared for the next and final step of opening the steam supply valve. The described system is a relatively simple one; however, complexity is easily achieved if the desire for automatic operation is extended. In this case the system was kept reasonably simple by requiring manual setup for certain valves and power supplies, and the starting sequence was manually initiated. Where an unattended watch is required, the additional feature of automatic start-up of the standby unit in the event of operating unit failure would be provided. Such a feature would have to possess the ability to determine which of the installed units will operate and which will serve as standby. For an unattended boiler, additional protective features are also required to monitor the critical conditions. To substitute for the fireman's eye, each burner is equipped with a flame detector. The detectors are sensitive to the ultraviolet emission from the flame and autornatically shut off the fuel in event of a flameout. The condition is, of course, alarmed. Off-normal water levels are also alarmed. If the level gets below a safe limit, the flames are automatically extinguished. A high level may cause damage to the machinery by water carry-over into the steam lines; therefore, protection may be provided' by shutting the feed valve. In addition, closing the main throttle or extinguishing the fires may prove beneficial by causing a "shrinkJ' in the drum water level. 'Alarms may also be provided for such off-normal conditions as low steam pressure, high steam temperature, and low level in fueloil tanks. The main propulsion turbine and gear unit on an automated vessel will have the following features not provided on a nonautomated vessel:
- VALVE ACTUATINO POWER I
-
I
MANUAL VALVE SET-UP LOCAL CONTROL ON'REMOTE~
(
PUMP NOT RUNNlNQ 1
MASTER
AUXlLl ARY. LUBE OIL PUMP LUBE OIL SENSOR
Fig. 5 Logic diagram, boiler feed pump remote start-up
POSITION SENSOR
PUMP RUNNINO
require additional control equipment to operate air registers, control fluid flows, and provide ignition. Either system is generally suitable for port conditions because the minimum fuel range capability of the burners is lower than that required under port standby conditions. However, the continuous-use type may generate excess steam during a transient reduction in power. To prevent the safety valve from popping, a feature may be provided to artificially increase steam demand by dumping steam to the condenser. To maintain satisfactory combustion throughout the full range of fuel flow, refinements have been required in the systems controlling both air flow and fuel flow. Inlet vanes on forced-draft blowers have been supplemented with discharge vane control and with automatic speed changing to minimize excess air a t low flows. Fuelpump discharge pressure is programmed to maintain a constant differential pressure across the flow-control valve which permits good control characteristics even a t low flows. The boiler feed pump is a vital auxiliary and in most cases is located remote from the control station. Plants
FAILURE TO START
equipped for a one-man watch may be provided with means for remotely starting the pumps. Adost marine plants have steam turbine-driven feed pumps and their starting requires a control system that provides the sequential steps necessary. Such systems are usually referred to as logic systems since they contain the intelligence required to initiate each step in proper order and to check if each step is accomplished satisfactorily. Figure 5 is a logic diagram showing the steps that might be required; obviously the system will vary depending upon the particular feed pump and driver installed. Figure 4 shows such a system installed on a console. In this example there are four inputs to the system to indicate that the power and piping are set up in such a manner that the system is in a ready state. With these 1 tpermissivesJJin proper array the start switch may be manually actuated to initiate the starting cycle, which begins with the running of an auxiliary oil pump. The establishment of normal oil pressure permits the sequence to continue and energize a gland seal control valve motor operator. Note that in the event that pressure is not obtained within a specified time (time delay unit) the
1 Remote operation of the ahead and astern throttle valves. 2 Automatic operation of astern guardian and bleeder valves. 3 Extensive monitoring systems. 4 Remote emergency shutdown. !
1
i
The throttle remote-operating system is the major addition. Usually there are two remote-control stations, one a t the control center and one in the wheelhouse. Dependingupon its design, various means may be utilized to actuate the valve. Electro-hydraulic servo mechanisms have been used to the largest extent because they not
only have sufficient force to move the valve but they can accurately position the valve. This is important since at low power levels very little valve movement is required to obtain a significant change in propeller revolutions. The throttle valves may be either the globe or barlift type; usually the ahead valve is of the bar-lift type since it provides a ' single mechanism to give the full range of power with minimal throttling losses. There are two basic concepts for the control of propeller revolutions : one permits sqbstantially infinite variation; " the other commands speci5 increments, generally following what might be equivalent to tele graph commands. A feedback system is employed to accurately position the valve. The feedback input may be simply valve position, in which case the command signal represents strictly a valve position. Valve position is, of course, related to power, assuming standard conditions. The command unit may be characterized to approximate propeller revolutions. Variations in sea conditions, displacement, hull fouling, etc. affect the relationship between propeller revolutiqns and power. A positioning system may be refined to correct for this by introducing a feedback signal proportional to propeller revolutions. This signal serves to bias the position signal, resulting in a command for propeller revolutions rather than a specific power. Speed feedback is generally made effective in the maneuvering range only; speed feedback is usually not employed at the higher revolutions associated with free route at sea because the motion of the "vessel due to sea conditions mav result in rDm variations (such as under conditions where the broaches' the water). The corresponding variation in propeller rpm would cause unnecessary movement of the throttle mechanism in an attempt to hold constant rpm and therefore increase wear. Variations of speed feedback have been used to increase the speed of response by opening or closing the valve to
Table 1
Automated Features Provided for Various Manning Levels --MANNING LEVELUnat2 Men 1 Man tended
AUTOM.~T~:D FEATURE Steam generating plant. . . . . . . . . x Bridge propulsion control . . . . . . . Propulsion auxiliaries: console start.. . . . . . . . . . . . . . . . automatic stadby, start. . . . . . Fire detection system+&larm. . . . Fire extinguishing system-console control. . . . . . . Bilge system: level alarm. . . . . . . . . . . . . . . . . .
console automatic c6ntrol.. control. .. . . . . . . . . . . . Data logging-off limit. . . . . . . . . Alarm system: location-control center. . . . . . . x location-Engrs. quarters. . . . .
'
x
x
x
x
x
x x
x
x
x
x
x
X
x x
x x
AUTOMATION
MARINE ENGINEERING Table 2
VESSEL"A" (2 Men) A.
Table 2 (cont'd)
Control Systems Provided for Vessels A, 8, and C VESSELliC" (Unattended) C.
Stmm Ornuding Pknl
1. Combustion Control.. ........... Siigle Element (Boiler Pressure)
5. Fuel-Oil Pumps.. ............... Remote Start/Stop
Remote Start/Stop
6. Forced-Draft Blowers. . . . . . . . . . . Remote Start/Stop
Remote Start/Stop
7. Feed Pumps..
..................
10. Superheated Steam Temperature Control. ........... 11. Boiler Sgteguards Steam Dum ............... ~ e e dshot ............... Fuel-Oil Shut Off. . . . . . . . . . . . .
8f :
Local Start/Stop
Three Element (Boiler Pressure, Steam Flow, Rate of Throttle Movement) Two Elemeqt Automatic Sequencing Automatic Analyzer Automatic Treatment & Blow Down Remote St?rt/Stop, & Automatic Transfer
Remote Start/Stop & Automatic Transfer Motorized (High-Level Shut) Local-Manual
\*4.
Eiuhic P o w r Oenemting Units
................... S eed Control. . . . . . . . . . . . . . . . . . Arbine Safeguards. . . . . . . . . . . . .
1. Start/Stop..
Local Stmt/Stop
2 3:
At Switchboard Low Oil Tri overspeechrip
-
Synchronization.. .............. Manual 5. Generator Safeguards.. . . . . . . . . . . Low Voltage Tri 6. Emergency D i d Generator. Aubomatic overcurrent Start, 8!p
....
Local Stop 7. Standby Lube-0.il Pump (Mam Unit Dlrect Dr~ven)..... None 8. Auxiliary Circulating Pump. ..... None 9. Auxiliary Condensate Pump.. .... None
10. Auxiliary Air Ejector.. . . . . . . . . . . None 11. Remote Controlled Valves.. ...... None
Local/Manual
Automatic
Automatic
High Pressure ~ i g Level h Forced Draft Fan Failure, Flame Failure, Low Water
High Pressure High Level Forced Draft Fan Failure, Flame Failure Low Water High l$ater Remote Mechanical
Not Installed High Level Forced Draft Fan Falure, Flame Failure, Low Water
Remote Electric from Bridge/Engine Room Console.
Remote Electric from Bridge/Engine Room Console.
Safety Valve Easing Gear. . . . . Remote Mechanical
VESSEL"C" (Unattended)
VESBEL"A" (2 Men) Main Local Start/Stop, Standby Remote Start/Stop At Switchboard Low Oil Tri ~ v e r s p e ehf i p
Local Start
,
Automatic r" voltage,+ri Automatic ~vercurrent Start, &p
Zip
LOW
Local Stop
Remote Stop
Remote Start/Stop Automatic Start Remote Start/Stop Remote Start/Stop
Remote Start/Stop Automatic Start Remote Start/stop Remote Start/Stop Automatic Tiansfer Remote Start Automatic Lube Oil Cooler Water In%, Remote Control of: Exhaust V ~ V W ~uxiliar ~ i r c u i ~ t i n ~ ~uction~~iarge, and Auxiliary Condensate Sucbon/Dis-
Local Start/Stop Exhaust from Main Unit, Steam to Standby Unit
Motorized
B. Main hpuirion Turbinw
1. Throttle Control (a) Type.
....................
Remote Electric from Bridgemngine Room Console.
Valve position above
Opening Rate Limited
2. Feed Make-U . . . . . . . . . . . . . . . . . 3. Deaerator S $. . . . . . . . . . . . . . . . . 4. Machinery #pace Vent Fans. ..... 5. Engine Room Saltwater Service Pumps. . . . . . . . . . . . . . . . 6. Fire Pumps. ... :. . . . . . . . . . . . . . .
AutomaticAutomatic Remote Start/Stop Local Start/Stop Local Start/Stop
7. General Services Pumps (Bilge, Ballast, etc.). .......... Local Start/Stop
8. Fuel-Oil Transfer Pumps.. . . . . . . . Local Start/Stop 4.
Bearing Oil Control. . . . . . . . . . . . . Maintains Constant Pressure to Most
Maintains Constant Pressure to Most
6. Motorized Astern Guardian Valve.. . . . . . . . . . . . . . Remote Close, Automatic Open 7. Emergency Stop. ............... Engine Room Console
Remote Close, Automatic Open Engme Room Console
8. Auto Rotation.. ................ Ahead-2 min afkq Shaft Stopped
Ahead-2 min after Shaft Stopped Local EngagelDisengage
Pressure to Header
Automatic Open & Close Bridge/Engine Room &mole Alternate Ahead and Astern 3 min after Shaft Stopped Local Engage/Dlsengage
9. Remote Controlled Valves (Motorized). . . . . . . . . . . . . . . . . .
Boiler Feed Sto Astern ~uard%n, Turbine Bleeder
!G%!c%oard Low O i l Trip, Overspeed Tri Rotor I n t e r n 2 Clearance Semi-Automatic Low Voltage Tri Overcurrent Automatic Start,
High ~alinit: Automatic Automatic Remote Start/Stop
, ,--
Remote Start/Stop Remote Start Stop, (One Unit nlv)
d
Local Start/Stop, Remote klltop
Local Start/Stop Remote Stop Boiler Feed Sto Astern ~ u a r c k n , Turbine Bleeder, Main Sea Suctions (High $ Low), M a n Clrculatlna ~ystein, Condensate Pump Discharge, Fuel Oil Pump Suction & Emergency Discharge,Fire andPump Suction & Discharge
Local Control, Automatic Dumpin When Maneuvering !t On High Salinity Automabc Automatic Remote Start/Stop Remote start/stop Replote Stiut/Stop (One Unit from Bridge) Remote Start/Stop, i Biige Unit Automabc Start/Stop & Stop on Oily Overboard Discharge -Local StartJStop, Auto matic Stop High Level Boiler Feed Sto Astern ~uarcfkn, Turbine Bleeder, Main See Suctions (Hip, Low &Scoop) Mam Clrculatlng S stem, Set Erhayt, r Ejector
g~
Settlin T p k High &cbon, Auxiliary Condenser Distiller ~ e e Heater Recirculating d iary Condenser), (Main & Auxiland Drain L.P. Turbine Emergency Mmeuvering
12. Main Condensate Pump.
........
Local Start/Stop
Remote Start/Stop
802
MARINE ENGINEERING Table 3 Information Systems Provided for Vessels A, B, and C VESSEL VESSEL VESSEL "C" "A" 'IB1l None 60.ppin t Off-limit Digital D i ~ t a l Analog None None Yes
Data Logger. . . . . . . . . . . . . . . Bell Logger. . . . . . . . . . . . . . . . Fire detection system. . . . . . . Off-watch personnel call system. . . . . . . . . . . . . . . . . . Pressure indications (all systems). . . . . . . . . . . . . Tem erature indications (aRsystems), . . . . . . . . . . . . Liquid level indications. . . . . Auxiliary running indications . . . . . . . . . . . . . . . . . . Valve position indications. .. Off-limit alarms: (a) pressures.. .......... (b) temperatures.. . . . . . . (c) hquid levels. . . . . . . . . (4 incorrect set-up. . . . . . (e) miscellaneous.. ...... Smoke indicator.. . . . . . . . . . . Flame detector voltage.. . . . . Throttle valve position.. .... Bridge command lever position. . . . . . . . . . . . . . . . . . . Engine order telegraph.. . . . . Shaft revolution indicator and counter.. . . . . . . . . . . . . Horsepower meter. . . . . . . . . . Generator voltage, current, load . . . . . . . . . . . . . . . . . . . . Generator frequency. . . . . . . . . . Generator power factor.. ......
Yes
Yes
Yes
Yes Yes Yes
Yes Yes Yes
9 Yes Yes Yes
Yes Yes
Yes Yes
Yes Yes
Yes Yes
Yes Yes
Yes None
Yes Yes Yes
Yes Yes Yes
Yes Yes Yes
13
positions beyond that required for steady-state conditions. The speed of response is also affected by the rate of valve movement (which corresponds to a rate of change of power). It is important that this rate be compatible with the ability of the steam generating plant to change its rate of output. Shrink and swell in the boiler drum Must stay within acceptable limits and combustion controls should change the rate of firing with acceptable pressure variations and smoking. A time of 20 to 40 sec from fully open to fully closed appears to work well for the average plant. The incentive to decrease this time is low because during the first 20 to 40 sec of a fuH-power quick reversal, the hydrodynamic resistance of the hull provides the major braking form to slow the ship. Furthermore, the propeller cavitates heavily if reverse power is applied too abruptly; consequently the propeller has limited effectiveness until the ship has reduced speed IR appreciably. The throttle control system may be arranged to accomplish allied functions, such as : 1 Automatic closing of extraction and opening of astern guardian valves upon reduction in power. 2 Automatic intermittent rotation of the turbine when under a stopped-shaft condition. 3 Throttle closure or modulation upon loss of boiler fires, low steam pressure, overspeed, etc.
Auxiliary automatic controls may be provided for
starf up of standby condensate and lubricating-oil pumps, gland seal pressure control, condensate recirculation for air ejector condenser, lubricating-oil temperature controll etc. as may be desired depending the plant - upon design and the desired manning level. The operation of steam turbines has always required instrumentation for monitoring and control purposes. It is therefore not surprising that the additional information systems required on the automated vessel are not extensive. Typically, they may include alarms to warn of a high bearing temperature, low condenser vacuum, or loss of throttle control power. The electric generating plant is a prime auxiliary associated with the propulsion plant. The automatic control of voltage and frequency is essential for proper performance of the electric plant and, therefore, such control is commonplace. The automated vessel, however, may be equipped for remote start-up and automatic synchronization of the standby unit from a remote station. The control and monitoring provided for other auxiliary systems depend upon the manning level, type of machinery, and machinery arrangement. I n general the watch stander's normal duties must be made consistent with his available time; in many cases this means the elimination of a function or its assignment to a dav worker. The extent to which automatic z r remote operation is applied is limited only by economics. Table 1 gives the general features to which automation might be applied in a steam propulsion plant depending upon the manning level. These may be considered as basic requirements; actual installations will vary considerably in scbpe and detail. The focal point of all control and information systems in the engine room is the control center. This center is usually located in or adjacent to the machinery space such that there is ditect vision of the boiler firing aisle (this is a preference but not a necessity). Minimum require ments for the instrumentation are established by the regulatory bodies [6-111 and depend upon the manning level desired; actual installations vary to suit not only the specific plant but also operator's and designer's preferences. Three ships, designated as vessels "A," "B," and "C," have been selected for illustration and are tabulated in Tables 2 and 3. It will be noted that there is considerable difference in-scope and arrwgement even though each center serves to control a single-sci-ew geared-turbine plant with two boilers. These diffetences represent variations in regulatory body requirements and the owners' particular interests and operating practices. Vessel A is intended for a two-man watch. The control center, Fig. 6, is located on a flat open to the machinery space and adjacent to the firing aisle. The main generator switchboard forms a part of the station. The console, Fig. 7, faces athwartship and is high since sight over it is not required. The height permits the use of conventional instruments without requiring excessive length. This vessel is equipped with bridge throttle
i
-
-.--
-
COMBUSTION CONTROL CHAMB
-
-
ON LOWER L E V E L
I
BOILER LIQUID LEVEL INDICATORS
FEED WATER T E S T FACILITY 7
DRINKING SALINITY PANEL
1
PCNl
ENGINE ALARMS CONTROL RELAY
-STOP VALVE
Fig. 6
Vessel A-trol
control which is in excess of regulatory body minimum requirements. The operators, in this cape, considered the feature desirable to permit the watch standers to assume duties other than throttle operation. The r e cording of data is limited to a strip chart recorder for command location, throttle command, and shaft revolu-
center arrangement
tions plotted against time. Note that the vertical portion of the console contains the information display which can be readily observed from the deck area in front of the console. Control systems are located on the lower portion of the console where they are convenient for manual manipulation.
Fig. 7
Vessel A-control center console
Vessel B is intended for a one-man watch. The control center is located in the open machinery space just forward of the firing aisle, Fig. 8. Propulsion and auxiliary remote control instrumentation is located on a console, Fig. 3, which faces aft. The engineer views the boiler burner area from over each end of the console. In this case the main generator control and distribution switchboard is located on the same level but somewhat remote from the console. Although normal control of the electric plant is at the switchboard, the console is equipped for remote start and automatic synchronization of the standby unit. A rather extensive analog type data logger is provided to record plant performance. The large mimic covers the information and control systems required for remote start-up of the standby turbine generator set. This display covers condenser circulating water, steam, and condensate systems as well as automatic synchronization of the generator. This equipment is shown in more detail in Fig. 4. Other features of Vessel B include remote start of the boiler feed pumps and more extensive monitoring. Vessel C is intended for eventual unattended operation when underway at sea. The control center, Fig. 9, is located in a separate air-conditioned compartment with easy access from the quarters and also direct access to the machinery space. This has been made possible through the use of top-fired boilers, the firing aisles of which are visible from the control station through shatterproof glass windows. The generator control and distribution switchboard is also located in the control center, thus eliminating the need for duplicate instrumentation. The generators are installed directly below the switchboard, minimizing the length of unprotected cable. The console, Fig. 10, has extensive control and information systems. Automatic start of vital propulsion auxiliary standbys is provided. The size of the console has been minimized by arranging the occasionally read data
to appear on an electronically actuated digital display device called a demand readout. Each data point is identified by number, the number is set up on a selective dial and the value is read on the indicator. Continuous display is provided for vital information. Indicators located on the console are electric instruments; a sensor at the measuring point sets up a voltage proportional to the actual value which is transmitted to the console instrument. Instruments which would be required to operate in the event of a power failure are directly actuated. These are located in a separate enclosure to prevent damage from leakage of the actuating fluid. One of the special features of this installation is the improper setup alarms; these are arranged to indicate that circuitry, valves, etc. are in their proper relation prior to system activation. Control of the propulsion plant from the bridge is an important and essential feature of the automated marine power plant. The ability of the deck officer to directly control speed and direction decreases response time and frees the engineer to concentrate on assuring availability of power. The installation should be kept as simple as possible and the deck officer should be reasonably free of monitoring machinery performance. The control system should automatically handle all response limitations. Figure 11shows a bridge console. The throttle is a lever arranged so that forward motion from stop position is ahead, thereby minimizing the possibility of an incorrect command input. Lights indicate the location of control. Alarms indicate the transfer of control and the loss of control power. Direct telephone communication is provided to assist the transfer of control. The remaining instrumentation is navigational and interior communication. Figure 12 shows the propulsion control portion of the console. Note the long arc through which the throttle command lever passes. The long arc helps to minimize sensitivity in positioning the lever for small changes in propeller revolutions. It also serves to improve accuracy in rapid handling because a variation from desired command will be less. This can be a vital factor in areas where navigation requires careful maneuvering and the watch officer must concentrate on ship movements and not lever position. The duty of maintaining a bell log may be handled by an automatic alpha-numeric printer (digital data logger). This unit is electronically controlled and receives inputs from a master clock, throttle cornsand, actual propeller rpm, and location of control (bridge or engine room). The printout is initiated by a change in command and is repeated two or three times at 30-sec intervals to record c engine response. 2.4 Diesel Propulsion Plant. The basic concepts described for the steam propulsion plant are equally applicable to a diesel installation. The specific instrumentation will of course differ considerably. As in the steam plant, there are many varied combinations of engines, drives and associated auxiliaries; each has its own requirements for control and information systems.
805
AUTOMATION
MARINE ENGINEERING
n
f
A
n
n
A
A
A
A
A
A
A
GROUP CONTROL " C n NON- VITAL a R E ~ R I G .
/
y
-
'
/
CENTRALIZED CONTROL PANEL
E L E C T R I C A L CENTRALIZED CONTROL EQUIPMENT I.
i I
!
!
i
I
1
I MAIN \SWITCHBOARD
N TURBO-GENERATOR
115V E M E R G . PANE
I
H E A T E R PANEL
i
I
Fig. 8
/ L.O. P ~ M PCONTROLLERS
Vessel B-contrpl center arrangement
A
806
MARINE ENGINEERING
rangements were simply remotely controlled means to manipulate the usual fuel setting and reversing levers on the engine. Automated systems, however, are more sophisticated and operate through sequencing devices and interlocks for maximum engine protection while permitting bridge personnel to maneuver the vessel without concern for engine operation. For example, one directly reversible engine control system accomplishes the following sequential steps when the control mechanism is given an astern command from an ahead position: q
Fig.
eclRculT T BREAKER o u
M
9 Veuel C--control center arrangement
I I
The diesel engine may be considered a self-contained power generator, provided fuel, air, and a cooling medium are supplied. It therefore may be somewhat simpler than steam from the standpoint of controls. In view of this it is not surprising that diesel-propelled ships have operated with two watch-standing personnel (one less
8 07
AUTOMATION
than steam) prior to the general acceptance of the term "automation." Thus in the case of diesels the automatod power plant considers only two levels of watch manning (i.e., one man and unattended). The throttle ~ontrolis the prime element in the diesol control system. Pilothouse control has existed for somu
t
Fig. 12 . Bridge console-propulsion
control
1 Fire in the scavenge belt 2 Excessive vibration 3 Turbocharger failure 4 Low lube-oil pressure or high temperature 5 High cooling water temperature
Emergency controls are provided to override the shut1 Move the fuel control to stop, shutting off fuel and down features if necessary, The control system also removing a blocking device from the reversing contains provisions for avoiding operation at critical mechanism. speed points, below 25-percent rated speed, and limiting 2 Move the reversing mechanism to astern, blocking the fuel control from the fuel position and position- fuel increase during propeller freewheeling. Where controllable-pitch propellers are provided, a ing the starting air distributor and camshaft brake programmed relationship between engine rpm and propilot. peller pitch may be provged to obtain maximum pro3 Move the fuel control to air start, applying starting air, verifying rotation and direction of rotation, pulsion efficiency. Multiple engine installations are usually fitted with and removing the fuel block. clutches whieh may be remotely controlled. Automatic 4 Move the fuel control to fuel position, setting the governor, and then controlling acceleration in controls must also be considered for the various auxiliary systems, including fuel oil, lubricating oil, and jacket accordance with a torque limitation program. water systems, steam generating plant (waste heat and Additional protection is provided for the engine by oil-fired boilers), and the electric generating plant, in order to obtain the desired manning level. automatic shutdown or slowdown in the event of: '
MARINE ENGINEERING
808
,
LUBE OIL COOL
TURBINES FOR
GAS TURBINES FOR
F. W. GENERATOR
Fig. 13
Diesel propulsion-control
Fig. 14
center arrangement
Diesel propulrio~controlcenter cowole
The acoustic characteristics of the engine generally dictate a separate sound-proofed enclosure for the engine room control center. The information systems will naturally vary to suit the machinery installation; however, the method of display will parallel that of the steam plant. Fig. 13 illustrates a typical direct-drive diesel plant arrangement, and Fig. 14 shows a typical control center console. For additional discussion regarding the automation of large slow-speed diesel engines, see Chapter 8. 2.5 Gas Turbine Propulsion Plant. The gas turbine, like the diesel engine, is a self-contained prime mover, and the engine requires certain automatic controls in order to function properly (see Chapter 6). These control systems are an integral part of the engine and are normally provided as part of the engine package. Adaptation of these controls to suit marine automation naturally depends upon the method selected to transmit the power to the propeller. However, as in the diesel installation, it involves the provision of remote (engine room and bridge) control systems to transmit the command input to the engine control system. Starting cycles are automatically sequenced. Remote start therefore involves primarily the decision as to
where the engineer and the information systems to monitor the procedure should be located. This selection may depend to some-extent on the degree of automation provided for the auxiliary systems such as fuel handling, lubrication, cooling, and compressed air systems. These systems must be placed in operation to support the prime mover and the complexity of automating these systems for remote start may not be economically feasible. 2.6 Cargo-Oil Pumping System. The handling of cargo oil on tankers presents an interesting opportunity to gain advantages of not only reduced manpower but also increased safety through better surveillance of the process through applied automation. The cargo system usually consists of piping systems leading from the tanks to a pump room in which several pumping units, usuallv driven by steam turbines, are located; &owever, whe; deep-well pumps are employed, the pumps are individually located on deck and there is no central pump room. The pumping units provide the power to discharge the cargo to shore facilities through deck lines and hose connections. These same lines are used in part or in whole for the loading operation. Great care must be exercised when loading and discharging cargo. The loading and unloading procedure must be controlled to prevent undue stresses occurring in the hull structure. Also, the rate of flow must be adjusted to coincide with shoreside facilities; and in the loading cycle, care must be taken to prevent overfilling, which results in spilling cargo on deck and into the harbor waters with attendant fire and pollution hazards. -. . 'I'he mammoth tanker has stimulated the application of central control systems 'to ease the burdens of these operations. The very size of these large vessels has made the direct manual operation dif£icult and to a degree impractical. The control systems for cargo-oil service consist of both remote controls and information systems. Remote controls are applied to both the piping system valves and the pumps. The ever-present explosive hazard limits the selection of control components. Electrical systems, if used, must be intrinsically safe; i.e., they must have a sufficiently low energy level such that a short circuit will not result in a discharge that is capable of initiating a fire or an explosion. This means that the power level must be maintained at a low value, and therefore the application of electrical equipment ialimited to information-type systems. Hydzaulic and pneumatic systems have been used most extensively. Many of the cargo valves are located in the tanks, and hydraulic systems have been most extensively applied for operating these valves. In some cases the valves are actuated by hydraulic pistons located directly on the valve; deck-mounted hydraulic motors which drive the valve through extended shafts have also been applied. Where valves are accessible, such as in a pump room, pneumatic systems have been provided. In either case, the control station not only provides the means for commanding valve movement, but also provides an indication of the valve position. Quite often such a system will bo
I
AUTOMATION
diuplayed in the form of a mimic depicting diagrammutically the location of the valves in the system. Pump controls praent an interesting operating problem. The initial starting of a unit is customarily handled by the ship's engineers; however, the actual control af the pumping cycle is controlled by the pumpman, who is a member of the deck crew. Controls are therefore provided for adjusting pump speed and stopping the unit at the control station. Information systems include an indication of pump speeds and discharge pressures. The central control would not be complete without an information system indicating the liquid level in the various tanks. This is particularly important at the high and low levels to prevent overfilling or pumps running dry. Communication is also an important part of such an installation and direct communication is provided with such locations as the deck loading station, bridge engine room, and in many cases, shoreside facilities. 2.7 Other Applications. There are few systems aboard ship that have not been affected, at least to some degree, by automation. The complexity of the modern vessel would not be possible without it. No one gives a second thought to the air-conditioning system operating completely automatically. ks in the case for propulsion equipment, the application of automation principles is a matter of coordinating the man, machine, and control system complex to result in an application which accomplishes one or more objectives such as reduced manning, improved efficiency, safety, and operability. The diversity of break-bulk cargo handled made the application of automation difficult until the concept of containerization was established. The establishment of "standard" container dimensions subsequently made a control system feasible. I t is interesting to note, however, that many container vessels have the material handling equipment ashore rather than on shipboard in order to gain greater utility from the investment. The tending of mooring lines to compensate for changes in ship position relative to the dock has been relieved by the automatic mooring winch. The winches have been fitted with devices that sense the line tension, and the control system uses this input to maintain the line tension within a selected range by directing the winch to
809
pay out or haul in the line. The benefits accruing from such an installation are more than merely manpower reduction. The vessel is maintained in a more stable position at the dock, thus assisting in the more efficient handling of cargo. Navigational control is another area that is receiving atteqtion. The "automatic pilot" designed to control the steering mechanism to hofd a fixed course has long since been perfected. However, careful observance and course :tdjustmen),to avoid collision remains the direct duty of man. Bere the problem of constant surveillance becomes the stimulus to develop an automatic waaning and course adjustment system. References
1 H. L. Harrison and J. G. Bollinger, Introduction to Automatic Controls, International Textbook Company, 1968. 2 "Terminology for Automatic Control," The American Society of Mechanical Engineers, ASA C85.1-1963. 3 Ernest J. McCormick, Human Factors Engineering, McGraw-Hill Book Co., 1964. 4 Wesley E. Woodson, Human Engineering Guide for Equipment Designers, University of California Press, Berkeley, 1960. 5 "Human Engineering Design Criteria for Wfilitary Systems, Equipment, and Facilities," Military Standard 53 MIL-STD-1472. 6 "Guide for Shipboard Centra zed Control and Automation," American Bureau of Shipping, December 1965. 7 "Automation in Ships," Lloyd's Register of Shipping, 1963, London. 8 "Automated Main and Auxiliary Machinery," Navigation and Vessel Inspection Circular No. 1-69, United States Coast Guard. 9 "Rules for Equipment of Unattended Engine Plant in Motor Ships," Danish Ship Research Institute, October 1967, Report No. DSF-15E. 10 "Regulations for Automatic or Remote Controlled Installations in Seagoing Ships," Germanischer Lloyd, 1965. 11 "General Technical Conditions Concerning the Automation of Ships," Bureau Veritas, Guidance Note.
'I t
1 I
I 1
i I
CONSTRUCTKI N MATERIALS
CHAPTER XXll
W. 1. Williams M. R. Gross
I Construction Materials
Seetion 1 Prefacing Remarks A single chapter on construction materials must be limited in scope and restricted to items of greatest general interest. Therefore, this chapter is confined to discus sion of metals for marine engineering applications. It is recognized that nonmetallic materials, such as plastics, elastomers, and ceramics, have their places of usefulness. However, metals have been and will continue to be the principal materials of construction for nearly all vital shipboard machinery and equipment. The large number of metallic materials and the complexities involved in their selection for various applications dictate that metals . receive primary consideration. Material selection involves considerations of cost, fabricabiiity, maintenance, life, and freedom from failure. Failure may range from simple malfunction to catastrophic destruction. Inasmuch as shipboard equipments are constructed principally of metals, the marine engineer is concerned about "why" and "how" metals fail in service. The complete fracture of metal compbnents from gross overstress (e.g., overtorqued bolts or accidental overloads) is easily understood and is not of primary concern. The vast majority of metal failures are more subtle and aTe attributable to wear, seizure, repeated stressing (fatigue), mechanical shock (impact), crkep, stress rupture, embrittlement, corrosion, or the combined action of stress and corrosion (stress corrosion, corrosion fatigue). Because of the chainlike reaction that can occur when one part fails in a complex piece of machinery, it is not always easy to identify the source of failure.
I n contrast to a critical problem faced by the naval architect in selecting materials for hull structures, the marine engineer is not so concerned with "brittle" fracture of materials which normally behave in a ductile manner; therefore, this chapter does not devote special attention to this mode of failure, except for brief mention in Section 5 under inaterials for low-temperature and cryogenic equipment. Failures caused by corrosion, fatigue, and high-temperature creep are of particular importance, however, and receive special discussion in Sections 2. 3. and 4. Section'5 'considers the selection and application of metallic materials in some of the more critical equip ments of concern to the marine engineer. The uses and types of materials for these equipments normally are covered by regulatiohs and specifications developed by regulatory bodies and technical societies. These regulations and specifications are updated periodically, and contain a large amount of detailed information which should be available to the practicing marine engineer. References [I-1211 are typical examples. A glossary of selected metallurgihal terms is included a t the end of the text for the convenience of the reader. Finally, a selected list of key references appears a t the end of the chapter. The references discuss certain subjects in greater detail, and many of them contain extensive bibliographies for those who may wish to explore further. I
Corrosion of Metals 2.1 Infroduction. Much vital equipment aboard ship must perform srttisfactorily in the presence of corrosive environmente. Therefore, the control of corrosion deterioration is an important consideration in marine engineering. Control does not neceesarily mean complete elimination of corrosion. More commonly it involves careful selection of materials for particular design and operating parameters to optimize cost, maintenance, reliability, and equipment life.
It is not expected that the marine engineer have a complete understanding of complex corrosion theories and corrosion-engineering practices. However, the marine engineer should be aware of the general types of corrosion and some of the material and environmental circumstances under which they occur. He can then
1
Numbere in brrrckets designate References st end of chapter.
,
recognize when problems exist and when to consult a corrosion specialist. References [13-221 are suggested for additional detailed discussion with regard to general corrosion theory. 2.2 Types of Corrosion. Corrosion in the form'of direct chemical attack, without the presence of an electrolyte, is one general type of corrosion. It can occur from reaction with certain chemicals, such as chlorine and sulfur. In such cases solution of the problem involves elimination of the corrodant or selection of a material with inherent resistance to the corrodant. Direct chemical attack also can occur from hot gases such as those encountered on firesides of boilers and in gas turbines. These cases may result from fuel constituents such as sulfur or vanadium, or from salt contamination in the fuel and/or combustion air; and reducing the corrosive constituents, lowering of operating temperatures, and selecting more resistant materials are the cures most frequently employed, as noted in Section 5. Another general type of corrosion occurs by electrochemical atlack in the presence of an electrolyte. Seawater is the most common electrolyte encountered aboard ship, and is responsible for the most complicated and troublesome corrosion problems. The reasons for this are the great chemical activity of seawater as compared to other electrolytes such as fresh water, the wide assortment of environmental and material variables which affect seawater corrosion of metals, and the wide variety of forms in which deterioration can occur. Other electrolytes leading to electrochemical attack aboard ship usually involve rather specific environmental circumstances and materials of construction, and are mentioned in Section 5 as appropriate. 2.3 Seawater Corrosion. The nature and extent of seawater corrosion depend on the alloy and its compositional and heat-treatment variations. Of equal importance are environmental variations such as degree of aeration, alternate wetting and drying, fouling, dissimilar-metal couples, velocity, turbulence, cavitationerosion, temperature, and crevices. The various types of corrosion which can occur in a seawater environment are as follows: a. General Corrosion (Uniform Attack). With rare exceptions, the important forms of electrochemical degradation occur by severe local attack rather than by excessive rates of general corrosion. When general corrosion is of concern, it usually can be controlled with paints, metallic coatings, sacrificial anodes, impressedcurrent systems, or by selection of more resistant materials. b. Galvanic corrosion can occur when two dissimilar metals are electrically coupled in the presence of an electrolyte such as seawater. Current will flow through the electrolyte from the anodic material to the cathodic material. The corrosion resulting from this "battery effect" is known as galvanic corrosion. Usually, corrosion of the anode is accelerated, and corrosion of the cathode is decelerated. The potentials of metals in the electrolyte, as might be
81 1
measured with a saturated calomel half cell, determine which metals will be anodic to others. Table 1is such a listing of metals in seawater, arranged in approximate order from the most anodic in behavior (magnesium alloys) to the most cathodic in behavior (graphite, graphitized cast iron). The listing is actually not exact, as relative positions may change due to variations in water velocity, temperature, oxygen content, and so forth [23]. Furthermore, the listing is based on opencircuit When current begins to flow between two metals inta closed circuit, polarization effects occur due to reactions at the metal-liquid interfaces. The polarization reduces the potential difference between the coupled metals and tends to reduce the galvanic current and corrosion. The suibbiiity of a dissimilar-metal couple in practice may depend on the relative areas of the anode and cathode. If the anode is small and the cathode large, the anode may suffer an unacceptable rate of deterioration. On the other hand, if the cathode is small and the anode large, the corrosion and average penetration rate of the anode may remain a t a tolerable level.- For example, Monel rivets in a large steel plate might be acceptable, whereas steel rivets in large Monel plate would not be acceptable. Occasionally, circumstances may dictate that dissimilar metals be coupled with an undesirable cathode/anode area ratio. In such cases it may be possible to reduce corrosion of the anode with paint. However, the paint should not be applied to the @node,because exposure of the anode a t any imperfection in the paint film would result in an even more unfavorable cathode/anode area ratio and concentration of the galvanic current a t the point of imperfection in the paint. Although the cathode is not the metal requiring protection, painting of the cathode would provide a more favorable cathode/anode area ratio and a reduction of galvanic current; and any corrosion of the anode woqld be distributed instead of concentrated a t one place [20]. The same line of reasoning leads to a warning against the application of noble-metal coatings on a less noble base metal: for example, chromium plating on carbon steel. If $he coating contains any imperfection or holiday, the large cathode/anode area ratio can cause severe and rapid corrosion of the base metal in the localized area of the imperfection [24]. Table 1represents an endeavor to rate the acceptability of various metal couples in seawater a t 40 to 80 F as a function of relative cathode/anode area ratio. The table is useful only as a qualitative guide. The ratings tend to be conservative so that any uncertainties will be in the safe direction. However, due to variations in oxygen content, temperature, velocity, and turbulence, the indications can apply only in a broad sense. This is particularly so for the couples marked with "X" and "C". Those marked with "U" should certainly be avoided. c. Pitting corrosion is a form of attack that may occur from local cells (potential diierencee) which develop on the eurface of a single material [14, 20, 211. The cells
t
MARINE ENGINEERING
Fig. 3
Impingement attack of a Mawoter pipe in the turbulent area downstream of a globe valve Fig. 4
Fig. 1
Seawater crevice corrosion d an austenitic stainless steel (attack within crevice)
arise from local environmental differences in oxygen, temperature, liquid velocity, and so forth. Oxygenconcentration cells are the most frequently encountered in marine environments. Pitting on boldly exposed surfaces is usually associated with quiet or stagnant environments which permit environmental heterogeneities to exist. Severe pitting of aluminum alloys in seawater is often associated with local cells resulting from the anodic behavior of the aluminum matrix and the cathodic behavior of heavy-metal alloying elements such as copper, nickel, and iron. For this reason, the aluminum alloys with lower heavy-metal contents, such as some of those in the 5000 series, generally have the best resistance to mwhter corrosion. However, even these alloys may suffer accelerated attack if the water contains heavymetal ions: for example, copper leached from antifouling paints. Noble metals such as stainless steels are often more susceptible to severe pitting than less noble metals such as carbon steel. This is because the nobility of the stainless steel is associated with a protective passive oxide film, and any local breakdown of the film exposes a loeal active area of less noble character. If the film cannot reform for any reason, the potential difference then ca&s accelerated attack a t the point of film breakdown. The chloride ion in seawater is especially aggressive in penetrating the passive film on these materials. d. Crevice corrosion is an especially severe form of pitting from concentration-cell effects. It can oecur on susceptible materials where the corroding liquid is trapped between two surfacewhence the name "crevice. " It is often studied in the laboratory by using a crevice created by bolting together two pieces of the same alloy,
Fig. 2
Seawater crevice corrosion of a copper-base alloy (attack outside crevice)
or by bolting a piece of plastic to the alloy surface. The localized corrosion may be in either of two locations [13, 251. Crevice corrosion of some metals, such aH austellitic stainless steels and many of the nickel-baso alloys, occurs within the crevice area. The chloride ion in seawater will penetrate the protective oxide film, and this creates an active surface which attempts to repassivate by combining with the oxygen dissolved in tho entrapped water. When the oxygen in the crevice in depleted, film repair is no longer possible and a galvanic cell results between the active surface in the crevice and' the passive surface on the outside. A second cell result^ from the difference in oxygen content inside and outsido the crevice. Both cells tend to accelerate corrosion of the alloy in the crevice area. An example of corrosioli inside a crevice is shown in Fig. 1. Crevice corrosion of other metals, such as some of tho copper-base alloys, may occur just outside the crevica area. I n this case, corrosion in the crevice caust:n saturation of the entrapped water with copper ions which stifles further corrosion in the crevice area. The difference in copper ion concentration inside and outside tho crevice results in a local cell tending to accelerate corrosion around the outside edges of the crevice. An example of corrosion outside a crevice is shown in Fig. 2. Marine organisms such as barnacles can attach themselves to a surface and create effective crevices. Many of the copper-base alloys combine antifouling characteristics with good resistance to pittingand crevice corrosio~~. This has been one of the important reasons for the widespread use of copper alloys in seawater systems. e. Velocity effects are important in systems handling seawater. As the water velocity increases beyond 2 or 3 fps, fouling tendencies diminish and pitting of the moro noble metals decreases or disappears if mecha~~ical
I
crevices capable of entrapping stagnant water are not designed into the system. Thus, the austenitic stainless steels and many nickel-base alloys that pit under lowveloeity conditions remain inert and passive a t higher velocities [251. Tihnium and certah nickel-base alloys are outstanding in that they remain essentially passive m d free of $tt2ng under both stagnant and movingwater conditions. As the water velocity incteases, the copper-base alloys begin t o lose their protective films with a resulting increase in general rate of corrosion [20]. The "critical velocityn for copper is approximately 3 fps, and increases to i 5 fps or more as one goes from copper to Admiralty metal, aluminum brass, aluminum bronze, 90/10 Cu-Ni (1.5% Fe), and finally 70/30 Cu-Ni (0.5% Fe). Thus, water velocity is an important criterion in material selection for tubes and pipes carrying seawater. Copper alloys, nevertheless, are often used a t velocities in excess of their "critical velocity"; examples &re pump casings, impelIers, and propellers where the wall thicknesses are such that moderate corrosion rates can be tolerated. Under severe mnditions, however, the point is finally reached where more noble materials must be selected. f. Impingement and Cavitation-Erosion. Insofar as material d e g r h t i o n is concerned, water veloeity alone is not necessarily the mo& important limiting factor. More frequently, seriok degradation occurs locally by impingement and cavitationerosion attack associated with turbulent areas in moving-water system$ [26]. For example, the effects of excessive water velocity first become apparent a t inlet ends of condenser tubes or in pipes immediately downstream of fittings where turbulence exists and smooth-flow conditions have not become established. I n these turbulent areas the protective films can be "scourednaway, leaving a n rtctive surface on which corrosion may be accelerated from galvanic currents set up by the surrounding passive surfaces. Poorly fitted gaskets, elbows, valves, or anything else that creates turbulence can lead to accelerated attack of piping when the water velocity is approaching the
Graphithatioh corraion i a a grey cast iron seawater pipe
smooth-flow tolerance of the alloy [q. Thus, designing for streamlined flow can be advantageous. However, this is seldom possible throughout a system, so the alloy n selection remains important. Globe valves are particularly bad from the standpoint of turbulence, a s may be noted from Fig. 3 which illustrates typical downstream deterioration resulting from an excessive water velocity with such a valve. The perforation damage in Fig. 3 occurred after 1.3 years at 15 fps water velocity in a pipe made of 70/30 Cu-Ni alIoy containing less than 0.1% iron. If the alloy had contained about 0.5y0 iron, the pipe would have suffe~ed very little damage after several years of operation. The higher iron content would provide a much better protective corrosion film; this illustrates the importance of care in the selection and specification of materials. g. Intergranular corrosion and/or selective-phase corrosion may take several forms: Graphitization corrosion is a type of corrosion common to gray cast irons, although not to austenitic nickel cast irons [20]. An example is shown in Fig. 4. The iron corrodes away, leaving a residue of the free carbon (graphite) that gives gray cast irons their name. The residue retains the original shape of the part, but has no significant mechanical strength. Dezim$cation is another type of selective corrosion common to brasses and bronzes containing more than 15% zinc [14]. The damage is thought to occur by solution of the alloy folIowed by redeposition in situ of the copper. The result is a part retaining its original shape but containing a porous, reddish-copper deposit of little strength. Muntz metal, naval brass, Admiralty metal, and aluminum brass Me subject& dezincification in seawater unless "inhibited" grades are specified. Dealuminiua6ion is a type of selective-phase attack occurring in some aluminum bronzes, particularly cast alloys 124, 281. An example is shown in Fig. 5. The attack is associated with selective corrosion of a fine network of aluminum-rich " g a m phase" in the microstructure of the alloy. Control of the composition, especially the copper and iron contents, and the addition
816
MARINE ENGINEERING
Fig. 5
Seawater dealuminization of a cast aluminum bronze
of 4% nickel leads to a suitable alloy for seawater service. Dealuminization can be particularly insidious, because the corroded gamma phase can occupy such a ' minor part of the alloy's volume that there may be no outward appearance of corrosion. The attack is evident only on fractured surfaces, yet the alloy may have suffered drastic loss of strength and ductility. Intergranular corrosion of austenitic stainless steels is yet another form of selective attack. The carbides in steels sensitive to this form of deterioration tend to precipitate a t grain boundaries when the steels in the annealed condition are reheated into the 800 to 1600 F range. Such temperatures are reached in the heataffected zones adjacent to welds. Preferential corrosion can then occur along the grain boundaries. This type of attack can be avoided by using low-carbon (0.03% maximum) varieties of the alloys, using stabilized grades (containing titanium or columbium), or putting 'the precipitated carbides back into solution by quench cooling (annealing) from above 1950 F [20]. Exfoliation (delamination) is a special form of selectivephase attack in which the corroding phase occurs in a highly preferred orientation because of rolling or other forming operations. An example is shown in Fig. 6. If it is necessary to use an alloy with this susceptibility, it is desirable to prevent exposure of the cross section to the corrodant. Sometimes this can be accomplished by "buttering" the edge with weld metal. On the other hand, pitting of the surface eventually may cause penetration to the nonresistant phase. When this occurs, the corrosion can proceed rapidly beneath the surface by a tunneling effect. If corrosion products are voluminous, internal pressure may cause blistering of the external surface. h. Stress corrosion is a form of local deterioration
be dangerous to design on this basis, because the' degree of residual stress is seldom known, and because local corrosion pits or other discontinuities can serve as sources of unanticipated stress concentration. Stress corrosion of a particular material usually occurs only in specific environments, and these environments may be essentially noncorrosive or only mildly corrosive in the absence of stress. Thus, the affected part may appear perfect except for the appearance of a crack. Examples are the cracking of hardened martensitic stainless steels in hot high-purity water [30] and the "season cracking" of brasses in environments containing a few parts per million of ammonia [31]. Pi-actically all low-alloy and high-alloy steels can stress corrode in the marine environment if they are heat treated or otherwise produced to a yield strength in excess of 150,000 psi [32]. The dangerous yield-strength range does vary with the alloy, however. With many of the steels, electrochemical processes seem to predominate when stress corrosion occurs a t the lower end of the susceptible yield-strength range. It is sometime^ possible, therefore, to extend the useful strength rango by cathodic protection to stop electrochemical attack, However, cathodic protection can cause generation of hydrogen on the surface of the steel, and this can result in failure from hydrogen crackingJ331. Indeed, a t high enough yield-strength ranges, the steels can suffer hydrogen cracking from the hydrogen generated solely by natural corrosion processes. It is wise to examine tho circumstances very carefully whenever a steel in excess of 150,000 psi yield strength is to be used in the mari~lo environment. Numerous high-strength aluminum alloys, particularly in the 2000 and 7000 series of alloys, are susceptible to stress corrosion in seawater. The direction and rate of crack propagation can be highly dependent on tlio direction of the applied stress in relation to directio~lal microstructural variations or "texturing" associated with
CONSTRUCI'ION MATERIALS
817
rolling operations 1341 (see the preceding discussion of 150 F, and then the problem is usually associated with conditions leading to the concentration of salts on sxfoliation). Some titanium alloys can stress corrode in seawater. evaporating surfaces [37]. (See Section 5 regarding the An examp1e.i~Ti-7Al-2Cb-1Ta alloy, in which the stress- use of austenitic steel in nuclear power plants.) i. Corrosion fatigue, which results from the combined corrosion susceptibility is associated with tendencies action of a corrosive environment and cyclic stresses, is toward formation of submicroscopic Ti3A1 precipitates in the structure [24, 351. Immunity to stress corrosion covered in Section 3. 2.4 Cathodic protection. eathodic protection somein this alloy can be accomplished by lowering the aluminum to 6% and adding 1% molybdenum to times is used to provide partial or .complete corltrol of electrochemical cyrosion problems [13]. One method suppress the formation of Ti3AI. Titanium alloys that do stress corrode in seawater involves the ipdtallation of sacrificial anodes (zinc, usually will not do so in the absence of a stress concen- magnesium, or iron), thereby setting up a galvanic tration such as a crack or corrosion pit. Since titanium couple which reduces corrosion of the protected material alloys are practically immune to all forms of marine by making it the cathode. Another method involves corrosion, including pitting, the existence of stress the use of an impressed-current system with an external corrosion tendencies was not recognized for many years; power source; the anode is usually an inert material, it was noted only when sharply notched specimens were such as platinum, to prevent consumption. Impressedexposed to seawater and found to support smaller loads current systems are not often used for the protection of machinery and auxiliary equipment handling seawater than similar specimens exposed to air [36]. The common grades of austenitic stainless steels do not aboard ship, because of difficulties in voltage corltrol and pose stress-corrosion problems in seawater except above current distribution on protected surfaces.
Section 3 Fatigue 3.1 Introduction. The term "fatigue" as used herein refers to the failure of a component under the action of repeated stress. Fatigue is probably responsible for as many as SO percent of all machine part failures. Of these, failures due to bending forces are the most prevalent type, with the torsion type next and the axial type seldom occurring. Although service conditions usually involve combinations of variable and steadystate stresses, and sometimes both bending and torsion, usually a dominant type of failure is found. 3.2 Fatigue Fracture. Figure 7 shows a typical fatigue fracture. I n many instances these failures are easy to recognize because of distinctive characteristics associated with fracture appearance. These are a "flat, brittle" appearance of a large portion of the fracture, and "beach marlis" or "oyster shell" marliings which focus on the origin or nucleus of failure. However, the absence of such marliings does not necessarily eliminate fatigue as a cause of failure inasmuch as the level and kind of loading, temperature, and e~lvironment can affect fracture appearance greatly and may malie diagnosis on the basis of appearance alone impossible. I n such cases the use of an electron microscope to examine the fracture surfaces a t high magnifications may be helpful. Fatigue failures are progressive, beginning as minute craclts that grow under the action of the fluctuating stress. Failure usually consists of three distinct stages, namely: cracli initiation, slow craclt propagation, and rapid terminal fracture. If a cracli is not present initially, a certain number of stress cycles \\-ill be con-
sumed in generating a crack from a "stress raiser" to a size sufficient to be visible to the unaided eye or to be detectable by nondestructive inspection. The crack will continue to grow in size with each stress cycle, usually a t an increasirlg rate, until the remair~ir~g section is no longer able to support the load. Complete fracture then occurs instantaneously. The terminal fracture may be ductile or brittle depending on the material and surrounding conditions. 3.3 Cyclic Stress and Strain. lcigure 8 shows the stress-strain relationships liliely to develop under cyclic loading conditions. The relationship in Fig. S(a) occurs when the applied force or moment is completely reversed
Fig. 7
Fatigue failure of a manganese bronze propeller blade
MARINE ENGINEERING
819
CONSTRUCTION MATERIALS CYCLE
-
+s ALTERNATING STRESS AMPLITUDE, S,
/
MEAN STRESS, 5, MAXIMUM STRESS. S,
I
'I
MINIMUM STRESS. S,;
I
10'
,
I
I
I
lo5
lob
10'
LOG N. NUMBER (X CYCLES TO WURE
Fig. 10 Typical high-cycle S-N dlagram
TIME
(01
Fig. 9
{b)
Fatigue cycle nomencloture
Fig. 8 Stmu-strain relationships under cyclic loading
but within the elastic region. St is the total stress range and et is the total strain range. Figure 8(b) shows the relationship that develops under reversed loading into the plastic region. The stress-strain relationship is no longer linear, but follows the hytfteresis loop BCDEB during each cycle aftm initial loading along OAB. 3.4 Fatigue Testing. Over 95 percent of our knowledge of the fatigue behavior of materials has developed from laboratory tests of relatively simple specimens. Some 200 laboratories in the United States are engaged in fatigue studies. The number outside the United States is estimated to be even larger. It can be assumed that nearly all of these laboratories are generating specimen data of one form or another. In spite of all the information developed, the fatigue behavior of materials is not completely understood. As will be seen, the subject is highly complex and to a great extent continues to remain empirical in nature. Until the middle 1950's, practically all the fatigue information was in the so-called hi&-c~ele region. That is, design and test data were concerned with the failure of materiab aftermany millions of stresscycles. such information continues to have direct applicability to the majority of fatigue problems in marine engineering. It was recogxibed, however, that some structures and equipments were failing by fatigue because of a liniited number of high-stress cycles. One of the principal causes of such failures was thermal gradients produced by periodic heating rand cooling of heavy-&ion pressure vessels and machinery components such as turbine casings. It is now recognized that other types of equipment are subject to this type of failure because of other factors, such as periodic high-pressure excursions. Failure by fatigue a t finite numbers of stress cycles, i.e., less than 5 X lo6, is called low-cycle fatigue. There is general agreement among investigators that the highcycle fatigue performance of metallic materials is related to tensile strength, whereas low-cycle fatigue performance is related to tensile ductility. Accordingly, the amount of cyclic strain that a material undergoes becomes
the dominant factor in low-cycle fatigue. Although subsequent sections may discuss high- and low-cycle fatigue separately, it is desirable in evaluating materials to consider the broad spectrum of fatigue behavior ranging from about loe to lo8 cycles. There are several methods of applying the load in fatigue testing. The specimen may be subjected to direct tension or compression, to bending, to torsion, or to a combination of these. The simplest and most frequently used method for high-cycle fatigue testing is the rotating beam test [38-401. The specimen may be loaded either as a simple 4point loaded beam or as a cantilever beam. The specimen is usually dead-weight loaded and rotated a t a constant speed, thereby creating fully reversed bending stresses with each revolution. In low-cycle fatigue testing, both axial and flexural loading have been used with success [41]. 3.5 Fatigue Data Presentation. I n presenting or orgadsing high-cycle fatigue data, the following nomenclature, depicted in Fig. 9, is used: Stress cycle. The smallest section of the stress-time function which is repeated periodically and identically. N~minalstress, 8. The dress Calculated by simple theory without taking into account variations in stress caused by geometrical discontinuities, such arr holes, grooves, and fillets. Maximum stress, Smm. The highest algebraic value of stress in the cycle. Minimum stress, Sm&. The lowest algebraic value of stress in the cycle. Stress range, ST. The algebraic difference between tho maximum and minimum stresses in one cyclo, Sr = Smm - Smc. Alternating stress amplitude, S.. One half the strestl range, S , = Sr/2. Mean stress, S,. The algebraic mean of the maximum and minimum stresses in one cycle, Sm = (S,.. Smin)/2*
+
asymptotic thereto.
,200 KS1 STEEL. SMOOTH SURFACE AIR ENVIRONMENT
J
\
-
100
I
MARINE ENGINEERING
though the measured parameter is strain. The disadvantage is that unrealistically high stress values are obtained in the very-low-cycle region. Figure 11 shows the broad-spectrum fatigue behavior of two steels with markedly different yield strengths. The solid lines compare the steels under ideal conditions, i.e., a smooth surface in a noncorrosive environment. As mentioned previously, the high-cycle fatigue life under these conditions is related to the strength properties of the steels. Thus, the fatigue limit of the 200,000-psi steel is considerably higher than that of the 40,000-psi steel. However, the curves cross each other in the lowcycle region. The higher tensile ductility of the 40,000psi steel is conducive to higher fatigue strengths in the very-low-cycle region where strain deformation is the principal controlling factor. The dashed curves in Fig. 11 depict the performance of the two steels under adverse conditions, i.e., a stress raiser in a corrosive environment. Behavior of the two steels is similar despite the large difference in static strength properties. Under such conditions, there would be no advantage in selecting the more expensive, higher-strength steel. 3.7 Factors Affecting Fatigue Life. The following paragraphs briefly describe the effects of various factors on the fatigue behavior of metals [43-491. a. Stress Raisers. Practically all fatigue failures start a t a stress raiser (stress concentration) on the surface of a machine part. The majority of stress raisers fall into one of the following broad groups: Those caused by changes in the geometry of a part, such as steps a t changes in diameter, abrupt corners, holes, keyways, threads, press or shrink fits, junction of bolt shanks and heads, and so forth. Surface discontinuities such as nicks, notches, machining marks, pitting, and corrosion. Defects inherent in the material such as nonmetallic inclusions, local discontinuities, minute cracks, and voids. The ability of a discontinuity to concentrate stress is dependent upon its shape and size. Cracks have the highest stress-concentrating effect, whereas generous fillets with a smooth, polished surface have the lowest. Analytical methods have been developed for calculating the stress-concentrating effects of discontinuities based on geometry, dimensions, and assumed elastic behavior. The effect arrived a t in this manner is called the theoretical stress concentration factor K t [38, 501. However, the actual effect of a given stress concentration may vary both within and among materials. By means of tests it is possible to establish the reduction in fatigue strength caused by a particular stress-concentration factor for a particular material a t a particular strength level. By comparing these data with unnotched (smooth) test data, one can arrive a t the so-called fatigue notch factor Kf. From K t and Kf the notch sensitivity index, q, of the material can be calculated as follows:
For most metals the notch sensitivity tends to increase with increasing strength. Thus, in the presence of sharp notches (Kt 1 3), it is not unusual to find little or no advantage for higher-strength materials. b. Maximum and Mean Stresses. The maximum tension stress or strain developed in the cycle has an important bearing on fatigue life. The results discussed thus far have been based on completely reversed (equal tension and compression) cyclic stress or strain independent of any steady-state conditions. This is the exception rather than the rule. Most practical applications result in stress-time patterns of the type shown in Fig. 9. It is apparent that if the maximum and minimum stresses are known, then all constant stress amplitude problems can be resolved into mean stress and alternating stress components. I n general, in the high-cycle region, the permissible alternating stress decreases with increasing mean stress. Through the use of mathematical or diagrammatical relationships, such as the Goodman diagram or the Haigh-Soderberg diagram, it is possible to convert the stress conditions shown in Fig. 9 to an equivalent, completely reversed stress condition. It is important to recognize that maximum stress controls the whole course of events. If the maximum stress exceeds the elastic limit, the mean stress is decreased and a new set of conditions comes into play. Factors most likely to alter the mean stress are stress raisers and residual stresses. The influence of mean stress decreases with decreasing fatigue life, and when the alternating stress equals or exceeds the yield strength, the mean stress becomes zero. Accordingly, whether or not mean stress is an important factor in low-cycle fatigue depends upon the yield strength of the material. In high-strength materials, it may be necessary to consider mean-stress effects a t fatigue lives ranging down to 100 cycles. c. Residual Stresses. Residual stresses may bo either favorable or unfavorable insofar as fatigue life i~ concerned [48]. Fatigue cracks initiate and/or propagab only in a tensile-stress field. Therefore, tensile residual stresses are usually detrimental to fatigue life whereu~ compressive residual stresses are beneficial. Nearly all fatigue failures initiate a t the surface. Accordingly, processes that introduce compressive stresses in tho surface layers of a machine part can be effective in improving fatigue performance. The most commonly used metallurgical processes for increasing fatigue resistance are case carburizing, nitriding, or carbonitriding. In these processes, carboll and/or nitrogen is diffused into the surface layers of tho part. When properly applied and heat treated, tho resultant metallurgical structure in the diffused laycr occupies a greater volume than that of the parent metal. The greater volume causes compressive residual stresaw. Compressive residual stresses can also be introduced by cold working the surface material by such techniques
.
82 1
CONSTRUCTION MATERIALS
as hammer peening, shot peening, and cold rolling. To produce the desired effects, it is necessary that the stressing technique plastically deform the surface layers of the metal. It is estimated that 80 percent of the improvement in fatigue resistance is due to favorable residual stress development and 20 percent to physical change of the surface metal. The benefits in fatigue of surface cold working tend to decrease with (a) increasing yield strength and (b) decreasing fatigue life. There are probably three reasons for the decrease with increasing yield strength: the higher forces required to deform the surface metal result in a shallower layer; the notch sensitivity of both the deformed and undeformed metal is higher; and the lower ductility of the high-strength material is conducive to cracking induced by some stressing techniques. Unfavorable residual stresses, i.e., surface tensile stresses, originate principally from heat treatment, welding, misfits, and cold forming. It is possible to relieve unfavorable residual stresses in metals by a socalled stress-relieving heat treatment and thus improve fatigue resistance. This treatment requires that the materials be heated to elevated temperatures. The applicability of the stress-relieving treatment is limited by facilities for handling large structures and by the fact that the temperature required for stress relief may have a detrimental effect on other properties such as yield strength and impact resistance.
d. Corrosion. The combined effect of a corrosive environment (such as salt water) and cyclic stressing is called corrosion fatigue and can be highly detrimental to the life of metals. This is particularly true in the case of carbon and low-alloy steels wherein the high-cycle, corrosion-fatigue strength in salt water is about the same regardless of composition or strength level (see Fig. 11). As might be expec%d, metals that are either partially or completely resistant to the corrosive environment will be less affected. For example, the fatigue behavior of cups6nickel and nickel-copper alloys normally used for salt%ater equipment is not greatly different whether exposed to salt water or to air. Many titanium alloys are exceptionally resistant to corrosion fatigue in salt water. Protection of susceptible materials against corrosion fatigue is similar to that taken against general corrosion, i.e., the use of protective coatings that are both resistant and impervious to the environment, and cathodic protection. Both of these methods are simple in principle but often difficu t to put in practice. e. Other actors. The foregoing sections have discussed four of the most important factors that affect fatigue life. Other factors that may be of importance are cumulative damage, prestressing, metallurgical structures, weldments, creep, temperature, surface finish, size, and stress state.
2
Section 4 Behavior at Elevated Temperatures 4.1 Introduction. An elevated temperature can be defined as any temperature a t which mechanical behavior is affected by time a t temperature. I n the broadest sense it can include such things as the change in tensile . properties by aging of certain aluminum alloys (including a few that age a t room temperature), the change in tensile properties and hardness by heat treatment of precipitation-hardening alloys, the release of residual stresses by stress-relief heat treatment of weldments, and the loss of ductility by temper embrittlement of steels. However, a more restricted definition of elevated temperature is applied herein, I n the context of this discussion, an elevated temperature is any temperature a t which strain and load-carrying ability are time dependent. Below the elevatedtemperature range, design for any life can be based on properties measured in "short-time" tests such as the conventional tensile test. Within the elevated-temperature range, however, mechanical properties must be measured by means which account for a change of properties with time. Elevated-temperature behavior occurs over different temperature ranges for different materials. For example, the mechanical properties of lead can be time dependent
i
a t room temperature and above, whereas for low-alloy steels they are time dependent only above about 700 F. Accepted practice dictates that, insofar as possible, a metallic material be used in a condition of metallurgical stability a t the operating temperature. For example, a quenched and tempered steel normally would not be used unless the tempering temperature were a t least 100 deg F higher than the intended operating temperature; otherwise, the strength obtained by the temper would be reduced on heating to the operating temperature. 4.2 Time-Dependent Properties. When a metal is subjected to a static load a t an elevated temperature, an immediate elastic strain occurs followed by a timedependent permanent plastic strain called creep. A typical strain-versus-time curve is shown in Fig. 12. The creep occurring a t a diminishing rate a t the initial part of the curve is called primary creep. This is followed by seconhrg creep, characterized by a relatively constant rate of strain. The creep rate may ultimately accelerate and lead to rupture in a stage called tertiary creep. It is rather obvious that stress-temperature combinations leading to tertiary creep within the life of the equipment should be avoided.
822
MARINE ENGINEERING
IN MATERIALS
RUPTW PRIMARY
SECONDARY CREEP
I
I I
I
I
1
' " """
' " ""'I
lo2
10'
TIME
10'
' ' ' "'
' " ""'I 10'
10'
RUPTURE TIME, HOURS
Fig. 12 Typical strain-time relationship of an alloy undergoing creep Bg. 1 4 Typical presentation of strescrupture data (see Section 4.2d for line extrapolation4
31 ""1
1
0.01
1
1
1
1
,
0.1 MINIMUM SECONDARY CREEP RATE.% PER 1000 HOURS
Fig. 13 Typical presentation of secondary aeep rate data
Allowable stresses for high-temperature design are often set by regulatory bodies such as ASME 151. For example, the allowable stresses for each material included in the ASME's code dealihg with power boilers were arrived a t by using the lowest of the following criteria a t each design temperature in the creep range [51]:
putting static tensile loads on specimens and measuring the timedependent strain with sensitive extensometers. The tests are made a t several stresses for each temperature of interest. Each test must be conducted long enough (usually several thousand hours) to establish the minimum strain rate in the secondary stage of creep. A conservative average stress for a secondary creep The data from such tests commonly are plotted as shown in Fig. 13, from which the stress to produce a particular rate of 0.01 percent per 1000 hours. creep rate can be estimated. 60 percent of the average stress for rupture in b. Stress-Rupture Tests. These tests are also called 100,000 hours. 80 percent of the minimum stress for rupture in creep-rupture tests, and are conducted by static tensile loading of specimens a t stresses high enough to cause 100,000 hours. failure. The tests are made at several stresses for each The use of average and minimum values assumes scatter temperature of interest. The rupture time is measured, in the rupture data from tests of many lots of materials. and strain-time curves also can be obtained if it be The designer must also apply additional safety factors if desired to extend creep-rate curves to higher rates of he must accommodate unusual conditions of instability, strain. The stress versus rupture-time data are plotted corrosion, possible overstresses, and the like. on semilog or log-log graphs. A typical example is European practice often bases allowable stresses on shown in Fii. 14. total creep (primary plus secondary) rather than the rate c. CreepRelaxation Tests. These tests are also of secondary creep. This has an advantage if close called relaxation tests. They are similar to creep tests, dimensional tolerances must be maintained. Data on except that stress is the variable rather than strain. primary creep are not commonly available for American After the specimen is loaded, the gage length of the alloys. specimen is held constant by reductions in stress so that a. Creep Tests. Creep data usually are obtained by elastic contractions will exactly balance any extensions
due to creep. The data from such a test commonly are plotted as shown in Fig. 15. The similarity between the relaxation test and bolting applications is apparent. Upon tightening to an initial stress, a high-temperature bolt retains its dimensions, but there is a fall-off of the stress as a function of time and temperature. The stress indicated by the asymptotic part of the relaxation curve is related to the ultimate "holding power" of the bolt. However, the asymptotic stress can vary as a function of initial stress on loading, and may also be different after a second loading, such as might occur by retightening of a bolt [52-541. Residual stresses can be introduced in a material by welding, cold forming, and other processes. The reduction of these stresses by stress-relief heat treatments can be said to occur largely by the creeprelaxation process. If the stress-relief temperature be restricted for any reason (such as not exceeding the tempering temperature of a quenched and tempered steel), the relaxation test is a useful means of estimating the level to which residual stresses can be reduced within a particular time a t a given temperature 154-551. d. Data Extrapolation. It is apparent that hightemperature creep and stress-rupture tests are expensive and time consuming, and the luxury seldom can be afforded of conducting tests over the full design life of equipment. Nevertheless, knowledge of long-time properties is necessary. For example, it is important to know that tertiary creep will not occur during long equipment life if a design stress be based on a secondary creep rate measured in a test lasting only a few thousand hours. Another example would be the 100,000-hr (11.4-year) rupture-stress criterion used by ASME in setting allowable stresses for boilers and unfired pressure vessels. A common method of obtaining long-life rupture-stress properties is to use a straight-line extrapolation of data such as shown by the dashed extensions of the lines in Fig. 14. There is debate as to whether the lines are actually linear or somewhat curvilinear, even for a metallurgically stable structure. Of greater concern is the possibility of temperature-induced and/or straininduced microstructural changes which are known to produce slope changes in some alloys. This circumstance is illustrated by the dotted lines in Fig. 14. The microstructural changes normally occqr in shorter times a t higher temperatures. This is also illustrated in Fig. 14 by the shorter time for the "break" in the curve a t temperature T2 than a t T1 (T2 being higher than TI). The concept that time and temperature bear an equivalence in the creep process, and that which occurs in a
40
823 I
I
TIME; HOURS
Fig. 15 Creoprelaxation cunes for a quenched and tempered carbonmolybdenum botting stpel
long time a t one temperature will occur in a shorter time a t a higher temperature, has been the basis for several parameter methods to estimate long-time properties from short-time tests. No single parameter works best for all materials, apd indeed the validity and accuracy of parameter methods is the subject of much debate. Reference [56] is an excellent critical review of timetemperature parameter methods and their usefulness. Although there is uncertainty about the actual longtime strength properties of many materials, it remkins an engineering necessity to make a best estimate for design purposes. The matter becomes critical if design safety factors are to be low. The problems associated with data analysis and e&mation of long-life hightemperature properties are described in references [57] and [58]. Different finishing heat treatments during manufacture are sometimes permitted by a material specification, and these may be reflected by differences in short-time rupture-strength properties. However, the higher strengths associated with some manufacturing processes may disappear after a very long time a t service temperature. Therefore, it is common practice to set allowable design stresses for long-life equipment on the basis of the "weakest" condition under which a material will be marketed [51, 571. 4.3 Other Considerations. Notches, biaxial and triaxial stresses, cyclic loading, environmental effects, and temperature "over-shoots" are additional factors isfluencing creqp and rupture behavior. Also, the temperature and strain history of a material undergoing creep may have an effect on other properties such as loss of fracture ductility. This is known as creep damage. References [42] and [59-611 are suggested as sources of . additional information regmding the& factors.
MARINE ENGINEERING
CONSTRUCTION MATERIALS
Section 5 Applicatio~~sof Materials 5.1 Introduction. The design, construction, and operation of modern shipboard machinery and auxiliary equipment are usually the result of years of experiment and experience. The wealth of information so gained is documented, and up-dated occasionally, in general specifications available for the guidance of the marine engineer. References [I-51 are typical examples. I n addition, certain of the military specifications contain general guidelines for specific items of naval machinery (e.g.1 MIL-B-18381 for Naval High Pressure Steam Boilers, and MIL-T-17600 for Naval Steam Propulsion Turbines). These general specifications refer in turn to a variety of specific material specificationscontaining details of the materials acceptable for particular components and operating conditions. Several organizations publish material specifications, with those of the American Society for Testing and Materials being the ones to which reference is most often made. Examples of typical materials used in critical applications, with some rationale as to why they are used, and the limitations under which they will provide satisfactory service are given in the following sections. 5.2 Boilers and Superheaters. The materials used in critical components of modern marine boilers have evolved over many years into a series of standardized alloys which optimize cost and mechanical properties. I n addition, they include where necessary other attributes such as castabilit~,v~eldabiity, forgeability, and/or sufficient ductility to permit forming operations such as tube bending and flaring. Most of the boiler components are made of carbon and low-alloy steels. However, their suitability in service is highly dependent on proper boiler design, construction, and operation. This is particularly important in marine boilers, and especially naval boilers, which are more compact and subjected to greater and more frequent changes in load demand than land-based plants. The design must incorporate adequate provision for flexibility to accommodate expansion and contraction fromtemperature cycling without development of stresses high enough to cause fatigue failure. Particular attention must be given to joint designs, such as a t tubeheader connections and tube that stresses are not concentrated excessively a t these locations. Proper placement and operation of burners are also helpful. The materials in boilers have limited tolerance for temperature, so design and opelation must guard against overheating. Burner placement and operation are important in this regard, as is proper circulation of hot gases and placement and maintenance of refractories. Good fireside maintenance is critical, so that deposits do not accumulate and interfere with circulation and heat
he water and steam side is also important from the
heat standpoint, because heat removal by water and steam circulation is necessary to keep tube temperatures within tolerable limits. The design must provide for circulation under all steaming conditions, and operators must not allow low water to occur. The watersides must be kept clean and free of deposits which can interfere with proper heat transfer. Corrosion also must be prevented. This is accomplished on the water and steam side by several means. Deaeration of makeup water prevents corrosion and pitting from oxygen. An alkaline-phosphate treatment of the boiler water provides an optimum pH for rninimiZing corrosion, and also controls damaging scale-forming salts by producing insoluble phosphates which can be removed by blowdotvn or easy mechanical cleaning. Proper water treatment and blowdown also prevent damage by foaming and excessive carry-over of solids into the superheater. Volatile amines may be used to control corrosion in condensers and return lines, and to prevent deposition of copper in boiler tubes. Caustic embrittlement, especially of boiler drums, is not the common problem today that it was in the past. Proper water treatment has helped. Of equal importance has been the use of fusion-welded construction in lieu of riveted construction. This has eliminated many of the crevices within which waterside chemicals could concentrate. Fireside corrosion from fuel-combustion products can be controlled adequately with good boiler design, established fireside cleaning procedures, and use of morc refined fuels. There are, however, certain parts such as superheater support members which must tolerate high temperatures and severe corrosive conditions. These parts are especially vulnerable when vanadiumbearing fuels are used and temperatures are reached a t which vanadium-bearing molten slags can form and accumulate [62]. The best material known to resist this attack is a cast 60Cr-40Ni alloy. ASTM A351 grade CH20 has been used successfully in commercial boilers; this alloy about 24Cr and 13.5Ni.
Another type of firesidecorrosion has been encountered occasionally, especially in cold boilers. This has beer, traced to acids formed when the fireside is not kept dry. Rain entering the stack has been a source of trouble ~ I I Some cases. The obvious solution is to keep the firesido dq. If conditions leading to failure by fatigue, overheating, and corrosion are kept under control, the primary considerations in material selection then reduce to product form, fabricability, cost, and the use of standard grades of alloys in accordance with regulations for temperature and allowable stress. The most common guide is that of the ASME Boiler and Pressure Veswl
i
1
825
Code, wvhich sets allo~vabledesign stresses as a function of temperature for each commonly used alloy. For the low-alloy steels, the tempekature and stress tolerances tend to increase progressively in the series of carbon, carbon-XMo, 1xCr-XMo, and 2xCr-IMo grades. There are, of course, other competing alloys, but these four are the most frequently employed. The useful temperature ranges for these alloys can overlap if suitable adjustment of the design stress is made. Holvever, carbon steel is seldom used above 750 F because of its low strength. If it is used a t higher temperatures, it is advisable that the steel be a grade deoxidized \vith silicon instead of aluminum. Alurninum-killed carbon steels have a greater tendency to become embrittled by decomposition of iron carbide (graphitization) after prolonged high-temperature exposure. Graphitization occurs most readily near welds. Molybdenum and particularly chromium are strong carbide-forming elements and inhibit graphitization tendencies in alloys containing these elements. Typical applications of carbon steels (including minor modifications in chemistry for welding, etc.) include steam and water drums, water-wall headers, generating tubes, downcomers, and economizer tubes and headers. Depending on applicable specifications, either or seamless tubes can be used. Inspection for
are not subject to this danger if the steam passing through them is already dry. There are places in boilers where special metals may be employed. Mention has already been made of cast 60Cr-40Ni alloy for superheater support members. Alloys such as Inconel may be used for oxidationresistant brick bolts; desuperheater parts are often made of a 16Cr-lNi alloy; and cBsings or other parts exposed to corrosive flue gases, particularly a t temperatures over 700 F, may be,mde of stainless steel. 5.3 Main and Auxiliary Sfeam Piping Systems. 1n general, the guideli~lesfor main and a w i l i a r ~steam piping systems on merchant vessels are set by applicable sections of the American National standards ~nstitute Code, ASME Boiler and Pressure Vessel Code [5, 101, and U.S. Coast Guard Marine Engineering Regulations [I]. Navy practice follo~vsthe same pattern, differing only in details. As in the case of boilers and supeclleaters, the rationale for choosing between carbon and alloy steels for steam lines is largely a matter of service temperature and pressure. However, in contrast to certain parts in boilers which are exposed to hot combustion gases, steam lines are not exposed to temperatures beyond those of the contained steam. The carbon and low-alloy ferritic steels can be used
alloy for tube metal temperatures up to 1100 $' and a columbium or titanium stabilized lgCr-CJNi austenitic stainless steel for metal temperatures between 1100 and
material muat have good hobforming properties. Both hot bending and cold bending are used in pipe
Beyond 1050 to 1100 F the high-temperature strength and oxidation resistance of the ferritic steels fall off rapidly. Hence, it is necessary to go to the more expensive austenitic stainless steels in high-temperature superheaters. It 'is common practice to u* ferritic steel tubes in the beginning of the superheater Pass1 and to restrict the austenitic steel tubes to the final portion of the pass. Economy is not the only reason for this. If stainless steel were used in the initial portion of the pass, the stainless would be exposed to wet steam entering the superheater. Any boiler-water solids carried over would concentrate on the tube surfaces and could lead to stress-corrosion cracking. Stainless tubes
fabrication. 1850 F with Hot the pipe bending packed is with sand done to at help retain to shape apd circularity. A somewhat higher temperature may be used for 2xCr-1Mo pipe. Cooling the outside of the bend may be employed to prevent excessive thinning in this area. A bending mchine with a mandrel is used for cold bending, and a minimum bend radius of 5 diameters is ,g8ually recommended. , After hot bending, carbon steels are usually annealed at about 1200 I?. The Mo arid Cr-Mo alloy steels are usually normalized from 1650 F and then drawn a t 1200 F. Both a r b o n and alloy steel piping are generally annealed after cold bending. Seamless pipe is used for the more critical applications, especially for the higher temperatures and pressures. Welded pipe is permitted for many other applications.
I/ I I1
I
82 6
I
'
bN MATERIALS
MARINE ENGINEERING
Copper or copper-alloy pipes with brazed joints can be used in noncritical, low-temperature, low-pressure service. For example, fabricated copper systems are permitted for steam and steam drain service on merchant ships up to 320 F and 75 psi [2]. Cast or forged valves are used in steam lines. For all practical purposes, cast steel valves can be considered as cast versions of the wrought alloys used for the piping. Minor compositional variations are mostly associated with the need to enhance the "castabilityn of the alloy to assure high casting quality. The seats and disks in steam valves require special attention because of tendencies to erode-a condition commonly known as "wire drawing." The tendency increases when valves are used for throttling service, and can be especially severe with wet, saturated steam. Austenitic stainless steels are quite resistant to wire drawing, but are not ushlly employed for valve trim because of poor resistance to galling and because of a difference in thermal expansion coefficients between the ferritic steel valve body and the austenitic steel trim. Nitrided steel has been used for valve trim on occasions, but nitriding has lost favor because of poor performance in comparison with other materials. The best trim material for resisting wire drawing is a weld-deposited, tungsten-cobalt-molybdenum hardfacing alloy known commercially as Stellite. Stellite to Stellite and Stellite to 13% Cr stainless steel are suitable combinations for the mating surfaces of disks and seats, although Stellite on both parts is preferred for the more critical systems. The Stellite facing of the seat can be applied by direct welding to the valve. However, welding procedures can be controlled better, and seat repair can be facilitated, if the Stellite is applied to a separate seat ring which can then be shrunk or screwed into place in the valve body. Steam line gaskets are made mostly of thin spirally wound metal ribbons spaced with asbestos. The metal must be "soft" so that it will deform during joint tightening and sealing. Soft iron and Monel have been used with success, but modern practice calls for stainless steel for high-temperature and high-ppssure services. Less costly gaskets, such as asbestos-metallic cloth and asbestos composition, can be used in low-temperature, low-pressure applications. Carbon or alloy steels are used for bolts on steam line joints. Carbon steel bolts are used a t lower temperatures. As the temperature increases, and depending on the temperature, one of the alloy steels must be used. The alloys most frequently utilized are Grades B6 (12% Cr), B7 (Cr-Mo), and B16 (Cr-Mo-V) of ASTM Specification A193. Alloy steel bolting normally is used in the quenched and tempered or normalized and tempered condition to enhance strength properties and to assure the highest possible elastic limit. Minimum tempering temperatures are specified to be well above the intended operating temperature, and for metallurgical reasons are usually within the range of 1100 to 1200 F.
Reference [2] is an excellent guide to materials for steam piping systems for merchant vessels. It contains a tabulated breakdown of material specifications for pipes, joints, bolts, gaskets, valves, and fittings for each of the various service systems aboard ship. Temperature and pressure limitations and other explanatory information are arrapged for convenient use. 5.4 Nuclear Reactor Plants. The nuclear reactor plants most often considered for merchant vessels aru the pressurized-water reactor (PWR) type. This mearltr that the reactor core is cooled by water kept undcr pressure to prevent boiling. Current practice is to barn the design of PWR pressure vessels on conditions of 650 F and 2500 psig. These levels provide a margin lor operational maneuvering, transients, and set-point ranges for relief and safety valves. The selection of the materials used in the reactor coro itself (fuel and fuel containment) is a highly specializcd field that is not normally under the direct control of tho marine engineer. I n somewhat the same category aro the mechanical devices used to manipulate the reactor control mechanisms which operate in the pressurizedwater environment. Corrosion and wear resistance arc) important considerations for these devices [30]. Reference [63] is a good source of general information on fuol elements and reactor controls in PWR plants. The pressurized water used to cool the reactor corc i~ contained in the primary coolant system. America~t practice to date is to make the primary system or corrosion-resistant materials (mostly austenitic stainlonn steels) to minimize the formation of corrosion productn. Control of corrosion products is important, as they t c ! ~ ~ t l to deposit preferentially in the core. Water circulation in the primary system is provided by a "canned rotor" pump so that only corrosion-resbtu~~b material is in contact with the water. Parts not math of austenitic stainless steel, such as springs and a few olllclr parts in the control devices, must be selected carefully uo as to be immune to stress-corrosion cracking, hydroac311 cracking, or other forms of deterioration. Austonibic? stainless steel cladding is used to isolate the low-rtll~~y steel from which the reactor pressure vessel is made. The primary coolant water is a high-purity deminoralized water with the oxygen maintained below 0.1 ppnl. Lithium hydroxide or ammonium hydroxide is addccl for pH control. Boric acid is added, as the boron car1 IIR used for fine control of reactivity "(known as chernio~d shimming). Reference [64] is recommended for furtl~crr details on primary water chemistry. With proper W I L ~ . ( + ~ maintenance, corrosion has not been a particlrlsr problem on the primary side, except for the threat of stress corrosion of specialty items made of q)ri~lg= tempered materials [30]. The stainless-clad reactor pressure vessels are wvcrrul inches thick and therefore pose special consideratiorln of cost, fabricability, and freedom from possible "brit41tln1' failure. A cautious approach has been used in ma1,cwLI selection [65]. Some early reactor vessels were matlcr of carbon-silicon steel (ASTM A212, Grade B). 'I'o
improve fracture toughneslg, later vessels were made of a nickel-modified version of ASTM A302, Grade B, a quenched and tempered Mn-Mo steel now designated ASTM A533. Other steels considered for possible use include 2%Cr-lMo (ASTM A542) and 3Ni-l%Cr-%Mo (ASTM A543). The selection of pressure-vessel steels has emphasized the development of uniform properties throughout the thick sections used, with particular attention to good lowtemperature notch toughness. An important consideration has been the change in mechanical properties brought about by bombardment with highenergy nuclear radiation. The most important change is the tendency toward progressive embrittlement, including a shift in the brittle-to-ductile fracture transition temperature. References [65-671 cover these aspects in detail. Since the primary coolant system is made of austenitic stainless steel, the same material must be used for the steam generator. This poses a special problem not encountered in conventional boilers made of low-alloy ferritic steels. The problem is associated with caustic and chloride stress-corrosion tendencies on steamgenerating surfaces made of austenitic stainless steel [68, 691. The primary water circulates inside the generator tubes, and on the secondary side there is a tendency for boiler water solids to concentrate, particularly in "hideout" areas such as tube-to-tube-sheet crevices. Even a few parts per million of chloride in the boiler water can be concentrated in the crevices to the extent that stainless tubes can fail rapidly by stress-corrosion cracking. There is evidence to indicate, however, that chloride cracking will not occur in the absence of oxygen. If there is no oxygen, high chloride contents can be accommodated temporarily from accidental in-leakage, such as from a condenser leak [30,68]. For these reasons, the secondary (boiler water-steam) side of a reactor plant must be maintained with extremely low chloride and oxygen contents, aiming a t less than 1 ppm and 0.5 ppm, respectively. I n view of the rapidity with which stress corrosion can occur above 170 to 200 P, it is necessary that the oxygen be removed before plant start-up. An oxygen scavenger such as sodium sulfite is commonly employed. Alkaline-phosphate treatment of the boiler water must be kept a t the proper pH-phosphate concentration, so that phosphate rather than free caustic residues will occur on evaporating surfaces. If free caustic residues occur, the stainless steel tubes can fail from caustic stress corrosion. Details of secondary-system water chemistry are covered in reference [MI. Some effort has been directed toward use of nickel alloys (such as Monel and particularly Inconel) to reduce chloride stress-corrosion hazards in steam generators. The secondary side of a PWR plant operates on saturated steam, so turbine blades must be resmtant to w e t - s h m erosion. Wet-steam conditions have also prompted use of cast 12% Cr steel turbine casings (in
827
lieu of the cast low-alloy steel in conventional plants) to improve corrosion and erosion resistance. 5.5 Fresh- and Saltwater Piping Systems. Carbon steel pipe (ASTM A53) commonly is used for miscellaneous non-saltwater applications such as boiler feed piping, low-pressure drains, auxiliary machinery piping, and engine cooling. The particular application may influence whether welded or qamless, or plain or galvanized pipe is chosen [2]. Copper pipe may be used in condensate systems. Polyvinyl chloride (PVC) plastic pipe (ASTM D1785) also is used for freshwater systems for auxiliary machinery and engine cooling. Cold and hot freshwater systems for domestic, sanitary, heating, and air conditioning services normally are made of PVC plastic pipe or seamless copper tubing (ASTM B88). The main and auxiliary saltwater piping systems present far greater difficulties than freshwater systems because of the severe corrosion problems which are encountered (see Section 2). Experience has shown that it is false economy to select cheap materials for initial construction if these materials will lead to costly shutdowns, repairs, and replacements. Water velocity and pipeline configuration are the most important variables (other than material) influencing the performance of a shipboard saltwater piping system. Most metallic materials of interest for saltwater systems have low corrosion rates in quiescent water because of the formation of corrosion films which protect the underlying metal. As the water velocity is increased, a velocity may be reached a t which erosion will begin. That is, the protective film will be swept away such that fresh metal will be exposed constantly to the water, and corrosion will be accelerated. I n general, the water velocity in piping systems is seldom high enough to bring about general corrosionerosion of the pipe material. However, velocities commonly are experienced that lead to damage in localized areas. This action occurs in areas of water turbulence. Because of its impinging effect, turbulent flow can sweep away protective corrosion films in local areas a t wate; velocities that are not destructive where flow is more or less streamlined. The rate of impingement attack in turbulent areas is a function of both the water velocity and the configuration of the piping system. The attack nearly always occurs just downstream of fittings and joints. Knowledge of this fact peranits a n easy inspection of a piping system, since the critical areas can be observed simply by breaking the joints. I t is seldom necessary to know the condition of a pipe beyond the area that can be seen readilv a t the inlet end. pipe bends do not mu&- significant impingement attack if the bend radius is generous enough to prevent wrinkling of the pipe or more than a moderate collapse of the ~ros~section, - ~ l b o w son , the other hand, can &use damaging turbulence if the nominal water velocity is fairly high. Long-radius elbows cause less damage than short-radius elbows. Other fittings, such as tees, reducers, and valves, also
11 I I I.
I
828
3N MATERIALS
MARINE ENGINEERING
thickness of the zinc and the water velocity, and may vary from as little as 3 months to 2 or 3 years. Other types of protection for steel pipe (lining with enamel, paint, rubber, lead, and so on) have been tried and generally abandoned because of cost and difficulty of installation, and maintenance of coating integrity. Stainless steel piping is not suitable for shipboard saltwater piping because of local pitting and crevicecorrosion tendencies. Aluminum is attractive because of its low cost and light weight, and certain alloys have good resistance to high water velocities; however, it has serious disadvantagen. One is its poor antifouling characteristics. Another in the tendency to pit from heavy-metal ions which may bo contained in the water from upstream equipment or other sources. But the most serious drawback is undoubtedly the susceptibility to rapid galvanic corrosior~ when coupled to heavier metals such as steels and coppcr. Use of 70/30 Cu-Ni welds for fabricating 90/10 base alloys. The use of bronze valves and fittings in a n alumirium Cu-Ni alloy. piping system, or the coupling of copper-alloy pipes Lo Use of Monel trim in a tin-bronze valve. aluminum pipes, would be intolerable. Any sucl~ The general considerations described in the foregoing combination would require completely insulated jointn, are applicable to any saltwater piping system almost including assurance against accidental coupling of tho regardless of the material from which it is made. The dissimilar metal components through the hull structurn. following paragraphs now discuss the performance Experience indicates that this complete isolation in characteristics of specific materials which have been used difficult, although not impossible, to achieve undor or proposed for use in saltwater systems. References practical shipboard conditions. Copper pipe can be used with some success in saltwater [20, 21, 23, 27, and 70-721 are suggested sources for lines. The principal drawback is its poor resistance to further elaboration. Experiments have demonstrated little diierence in the erosion and impingement attack a t more than moderatnr performance of plain steel, low-alloy steel, and wrought water velocities. I t also tends to corrode excessively by iron in seawater pipelines. Therefore, the term "steel" concentration-cell effects under stagnant water conditions. Between these extremes, copper can give good will be used to cover these materials as a group. Steel has fair resistance to impingement corrosion, but service, but in most piping systems it is difficult to poor resistance to cold salt water and very poor resistance guarantee that tolerable water-flow conditions can ho to hot aerated salt water. It is most useful in closed maintained. Aluminum brass and aluminum bronze alloys ( c . ~ . , systems where there is no continuous supply of oxygen. Steel has no antifouling properties, and marine growth 76/22/2 Cu-Zn-Al, 95/5 Cu-All or 90.5/2/7.5 Cu-Fe-Al) can accumulate under stagnant water conditions such as have better resistance than copper to erosion atit1 in dead-end lines. This can be a serious problem when impingement attack. However, some of these a1lo.y~ the ship is operating in shallow tropical or semitropical suffer from local forms of corrosion such as pitting, arltl also tend to be more difficult to solder or braze. waters. The cupronickel alloys are the best from the stantlThe corrosion products of steel may create serious rust point of all-around performance in shipboard saltwahw problems with small pipes. It is probable that steel systems. Their resistance to general corrosion, crevic!tr pipes under 1% in. or 2 in. in size would be unsuitable corrosion, and pitting is good, and these forms of atl;aalc because of rusting and fouling. seldom need consideration. The cupronickel alloys aro For various reasons, it would be desirable to use bronze easy to form, weld, and braze. They have good antivalves even if a piping system were made of steel; fouling characteristics. Additionally, of all the copyarhowever, such a combination would introduce a galvanic base alloys, the cupronickels have the best resistance to corrosion problem owing to accelerated corrosion of the high water velocity and impingement attack. steel in the vicinity of the valves. Installation of waster Two cupronickel alloys are commonly used: 70/:10 pieces between the valves and pipes would be desirable. Cu-Ni and 90/10 Cu-Ni (ASTM B467). If nomirlr~l Protective coatings are desirable for steel pipelines in pipe water velocities are held within conservative limibri, seawater service. Galvanizing has been tried in both these alloys will give long, essentially troublefree lira. service and experimental systems, and the zinc coating The 90/10 alloy should give 15 to 25 years of servico if is effective, but it also corrodes away after a time. The velocities do not exceed 8 to 10 fps. The 70/30 alloy extra pipe life gained by galvanizing depends on .the will provide in excess of 20 years of service a t velocit~ion cause turbulence. Globe valves used for throttling service probably lead to more pipe damage than any other type of fitting (see Fig. 3). Properly designed and installed unions are perhaps the least damaging. If a piping system is designed adequately from the standpoint of material selection and water velocity, there should be no major deterioration from standard fittings. An occasional failure may occur, but frequent failures should be an indication that the system is inadequately designed or is being used improperly. Whatever piping material is used, good design practice dictates that fittings, welds, and connections be made of compatible materials from the galvanic-corrosion standpoint. I n areas where dissimilar metals must be used, it is important thah insofar as possible, a favorable cathode/anode area ratio be provided and that the key components be made of the more noble material [20]. Examples would be:
up to 12 fps. The long life and generally satisfactory service provided by these alloys make them costeffective despite their high initial cost. The outstanding performance of modern cupronickel alloys is partly dependent on the addition of iron as a n alloying element (see Section 2.3). The 70/30 alloy requires a nominal 0.5% iron content, and t h e severe damage illustrated in Fig. 3 would not have occurred if this iron content had been provided. The 90/10 alloy requires a nominal iron content of 1.5% to provide maximum corrosion resistance. The 90/10 Cu-Ni alloy is the one most frequently used in the construction of merchant ships. It is also used in Navy surface ships. Because of its somewhat superior properties and reliability, the 70/30 alloy is used in sea-connected systems of Navy submarines. Titanium may find use for saltwater piping in special circumstances [73]. For example, the high strength-toweight ratio of titanium might be attractive for systems aboard weight-critical vessels such as hydrofoils, surfaceeffect ships, and deep-ocean submersibles. Pipe walls can be thin because of high strength, and because there is no need for a corrosion allowance in wall thickness calculations. Furthermore, pipe sizes can be smaller because high water velocities (over 20 fps) can be tolerated without damage. However, there are a t least three disadvantages of titanium. First, it is difficult to fabricate and repair in the field, and operating personnel cannot use simple techniques such as brazing. Second, galvanic corrosion must be talten into account wherever there is a connection to less noble metals. Third, titanium has no antifouling characteristics. Polyvinyl chloride plastic can be used for some salb water piping applications [2]. It is limited to lowpressure and low-temperature applications, and safety must be considered if the consequences of loss of the piping from fire damage would constitute a hazard. Valves, fittings, and pumps in saltwater systems should be compatible with the piping material. With cupronickel piping, it is common practice to use cupronickel or bronze flanges, and cast-bronze silver-braze unions. Valves for saltwater systems can be made of a variety of materials. Cast tin bronzes set the standard for highperformance valve bodies on both merchant and naval vessels. However, ductile iron and galvanized cast steel are commonly used, especially in systems of large pipe sizes. Alloys such as austenitic nickel cast iron, stainless 20, Monel, and 70/30 Cu-Ni may be used on special occasions. The seats in bronze valves can be cast as a part of the valve body, but this would be suitable only for very mild service. Otherwise, the seats would erode in turbulent water. The more common practice for high performance is to use seats and disks of Monel. This alloy has greater inherent erosion resistance, and the surrounding bronze body enhances performance by providing cathodic protection to the trim. Titanium, certain stainless alloys, and plastics have been used for trim in experi-
829
mental valves and have shown promise for applications involving severe erosion problems. Water-circulating pumps may be made of a variety of materials. Pumps for hot water heating systems frequently are made with a cast iron casing, brass or bronze impeller, and a stainless or nickel alloy shaft. Sometimes both the casing and impeller are of cast bronze. The best performance for saltwater circulation is obtained when the impeller alloy is not only resistant to erosion and impingement attack, but also is more noble than the casing $aterial so that cathodic protection of the impeller is obtained. One common combination is a cast iron casing with a bronze impeller and shaft. Better service is obtained with a cast bronze casing and Monel impeller and shaft. This combination is frequently used on merchant and naval ships. Monel impellers provide better erosion resistance than bronze impellers. All-titanium pumps have been tried experimentally and show promise of extended life with little or no sign of deterioration. Titanium pumps may be used in the future as a means of resisting severe conditions of corrosion, erosion, and cavitation damage. However, provisions would have to be made to control galvanic corrosion damage to less noble metals in connecting systems. 5.6 Condensers, Heat Exchangers, and Distillation Units. The tubes are the most critical components in
steam condensers. They must have thin walls for effective heat transfer, yet must provide isolation of the steam from the salt water used for cooling. Therefore, a prime consideration is resistance to tube deterioration on the saltwater side. Waterside deterioration can be considered from two aspects: how resistant a tube material is to the water velocity within the tube, and how resistant it is to erosion from turbulence and impingement a t the inlet end. Fortunately, the tube materials of greatest interest for shipboard service tend to rank themselves in the same order on both counts. Admiralty metal is the least resistant, and its use should be restricted to very low water velocities, preferably not over 3 to 4 fps. If Admiralty metal is used, a n inhibited grade should be selected to prevent dezincification. An inhibited aluminum brass (77Cu, 21Zn, 2A1, As) would be next in rank for resisting velocity and turbulence. This alloy is also resistant to brackish and polluted wpters, and is widely used in marine service. The highest performance standards among the copper alloys are obtained with the cupronickels. The 90/10 Cu-Ni (1.5% Fe) alloy is the one most frequently employed and has been used by the Navy in recent years for all main condensey on surfam vessels. Somewhat better performance and reliability can be obtained with the 70/30 Cu-Ni (0.5% Fe) alloy, but its higher cost favors its use only for special applications. The 90/10 and 70/30 alloys give good performance up to water velocities of 10 and 12 fps, respectively [74], although conservatism usually dictates somewhat lower limits in actual design.
830
MARINE ENGINEERING
In older practice, Muntz metal and naval brass were often used for tube sheets; however, the trend now is to use the same material for the tubes and tube sheets, and this is preferred. In any case, the tube sheet alloy must be resistant to saltwater corrosion, and should not be more noble than the tube alloy so as to prevent galvanic corrosion of the tubes. Excessive inlet-end erosion of condenser tubes sometimes has been a problem, even with cupronickel tubes in the earlier years of their use before iron additions to the alloys were made. Inlet-end erosion can be controlled in several ways. Waterboxes can be designed so as to minimize turbulence, impingement, and velocity gradients across the tube banks. I n some cases, plastic ferrules are inserted in the inlet ends to carry the water through the first few inches of the tubes where the water is most turbulent. In other cases, galvanic protection of the tube ends is provided by installing sacrificial zinc or iron anodes in the waterbox, or by making the waterbox itself of steel or cast iron. Monel waterboxes frequently are used with 70/30 Cu-Ni tubes; when this is done, the Monel is solder-wiped to reduce adverse galvanic effects. If design water velocities are maintained conservatively low, the modern iron-modified cupronickel alloy tubes require little or no supplemental protection for a satisfactorily long-life service. Tubes of 90/10 Cu-Ni work well with 90/10 alloy sheets and waterboxes. Tubes of 70/30 Cu-Ni work well with 70/30 sheets and either 70130 or 90/10 waterboxes. Other combinations are possible if their use does not lead to galvanic corrosion of the tubes. Navy practice in material selection for surface steam condensers is outlined in Military Specification MIL-C15430. This covers both the primary component8 as well as the miscellaneous hardware and attachments. Several materials other than copper-base alloys occasionally have been considered for condenser tube applications. These have included certain austenitic stainless steels and nickel-base alloys [20,74]. However, . titanium probably offers the most interesting potential for future shipboard conde~lserapplications [73]. Titanium condensers and heat exchangers can be used to advantage where weight is critical, such as on hydrofoils, surface-effect ships, and deep submersibles. The weight advantage arises from more than just a substitution of titanium for heavier metals. Thinner tubes can be used, as no corrosion allowance is required. High heat-transfer rates can be obtained with titanium despite its lower coefficient of thermal conductivity in comparison with conventional copper-base alloys. This is so for =era1 reasons. First, thin tubes can be used. Second, heat transfer is not stifled by a build-up of insulating corrosion films. Third, higher cooling water velocities (above 243 fps) can be tolerated, the limit being primarily a matter of the power costs for the circulating pumps. In addition, wet steam impingement is sometimes a problem on the outer surfaces of condenser tubes. The usual practice is to install baffles
to prevent direct tube impingement. Titanium tubes would be highly resistant to this form of damage. The miscellaneous auxiliary heat exchangers using salt water as the cooling medium require the same rationale for material selection as steam condensers. However, the services are usually not as critical, and design water velocities normally are quite low. Distillation units for making fresh water aboard ship present a severe material selection problem. Water temperatures tend to be high, and concentrated brine and scale are encountered, all of which aggravate corrosion attack. The removal of scale from evaporating surfaces also places heavy demands on constructio~l materials. The scale is often broken away by means such as cold shocking, and, as in the case of the vertical corrugated basket type, also by mechanical flexure. These scale removal procedures cause a fatiguing action. Occasionally the units may have to be cleaned with acid uyashes. It is not surprising, therefore, that distillation units demand the best materials. Distillation units are made in a variety of sizes and configurations. Possibilities include vapor compression, thin film, submerged tube, Aash, and vertical basket types, as well as combinations of these. Low-pressure steam or hot fresh water are used as heat sources. Military Specification MIL-D-18641 contains detaih of the material selection for all components of all type^ of distillation units acceptable for naval shipboard applications. The compositions of the basic component^ are 70/30 Cu-Ni, 90/10 Cu-Ni, or Composition 5 aluminum bronze (approximately 9 1Cu-7A1-2Fe). However, Monel is specified for the corrugated baskets and shells in the basket-type units. The selection of othcr materials for distillation units on merchant vessels may, of course, be different. Titanium tubes have been used successfully in landbased desalination plants [75] and probably will find uuu in future shipboard distillation units. Titanium secmn to be unaffected by corrosion, except for a tendency toward crevice corrosion under gaskets, etc., if brirla temperatures exceed 250 F. Alloy development work indicates that 2% Mo, 2% Ni or 0.2% Pd additiorrn may alleviate. this problem. 5.7 Steam Turbines. Turbine rotors can be made from a variety of materials depending on stress ant1 temperature conditions. Carbon steel (usually about; 0.30 to 0.40% carbon) is commonly used a t temperaturon up to about 600 to 650 F. Low.-alloy steels of tho Ni-Mo-V and Cr-Ni-Mo-V types are used a t somewhat higher temperatures. For extreme temperatures (alch as between 750 and 1050 F), it is common practice to utu, s 1gCr-1xMo-XV alloy steel. If corrosion and erosio~l resistance is a major design consideration, as in saturatotl steam plants, the turbine rotor may be made of 12% (Ir steel for service up to 750 F. Because of high stress conditions, turbine rotors mu#C be of high-quality steel, free of injurious defects such sn seams, scale, porosity, and excessive nonmetallic inc111sions. To obtain a desirable metallurgical structuro, u,
CONSTRUCTION MATERIALS
254 to 1 and preferably 3 to 1 reduction in area should be used in converting the cast ingot into a forged rotor. It is necessary to avoid hydrogen "&kingn of the rotor forging by assuring that the forging is not cooled to room temperature until there has been time for hydrogen t o diffuse out. This can be accomplished by interrupting the cooling a t about 400 to 600 F, and holding a t this temperature for a prolonged period. Alternatively, a common practice to avoid hydroge~flaking is to use vacuum degassing techniques dqring the steelmaking process. The deaired rnicrost~cturesand properties in rotor forgings are obtained by quenching and tempering or normaling and tempering. A tempering temperature of 1100 F is recommknded (1200 F for 12% Cr steel). After the rotors have been rough machined, they &re stress relieved a t 50 to 100 F below the tempering temperature. This provides for maximum dimensional stability after final machining. Alloys such as Monel and brass have been used for turbine blades, but steam turbine blading is now made almost exclusively from one of the quenched and tempered chromium stainless steels. AISI Types 403 and 410 are the most common. These are ertsily machined and forged, have good corrosion and moderately good webateam erosion resistance, have good mechanical properties a t opew,ting temperatures, and show good resistance tofatigue. Of equal importance is the unusually high specific damping capacity of these steels 176, 771. This can be used t o advantage in reducing fatigue stresses arising from blade vibration. It is a particularly important characteristic while passing through speed ranges that might produce resonant vibrations in the blades. Types 403 and 410 steels may not have sufficient creep strength a t the highest temperatures and stresses encountered in some marine turbines. When this is so, a more highly alloyed steel can be used, such as AISI Type 422 (13Cfli-W-Mo-V). The added strength in this steel is accompanied by a reduction in specific damping capacity. If wet-steam erosion is too severe in any stage of the turbine, it is sometimes the practice to apply Stellite inserts on %heblades to enhance erosion resistance. Titanium has been tried experimentally for stertm turbine blades and may see considerable use in the future, particularly in the iow-temperature stages of turbines. Titanium has good wet-steam erosion resistance. Furthermore, since its density is only about 57-percent that of steel, titanium blades develop lower stresses from centrifugal forces. Turbipe casings and steam chests are generally cast, using a variety of materials depending on the temperature and pressure conditions. Carbon, carbon-Wo, 1XCr-%Mo, and 2%Cr-lMo are the most common grades. A 12%Cr steel is occasionally used in saturated steam plants for better corrosion and erosion resistance (particularly in nuclear plants; see Se~tion5.4). 5.8 G a s Turbines. The widespread use of gas
83 1
turbines in aircraft engines has provided the major incentive to develop special mate~ialsfor gas turbines, especially the cobalt-base and nickel-base superalloys used in hot-section components. It is understitndable, therefore, that the materials developed for aircraft engines have also been used widely in marine versions of gas turbines. This applies to both main propulsion and auxiliary drive units. Nevertheless, the adaptation of gas turbines to shipboard applications has led to certain problems which are not of primary concern to the aircraft indust@. Two materials-related problems in particular have been encountered in marine turbines: vanadium corrosion and sulfidation corrosion, both of which can result in rapid deterioration of critical hotsection components such as guide vanes and turbine bMm. Each of these forms of corrosion is discussed briefly. Early gas turbine5 suffered accelerated oxidation attack of hot-section components when turbines were operated on residual fuels. The attack was associated with vanadium and sodium in the fuels which combined during combustion to form ash or slag deposits. Sulfated-vanadium/alkali-metal slags were typical. There appeared to be a threshold temperature, dependent on the slag composition, above which the slag promoted catastrophic oxidation. The threshold temperature varied from abouf 1200 t o 1560 F for many of the slags studied. No completely acceptable solution to this problem has been found. Alloys vary in their resistance to attack, but none combine the necessarf mechanical properties and long life required for marine gas turbine service. Fuel additives such as calcium or magnesium compounds are effective in inhibiting vanadium corrosion, particularly when combined with desalting to remove any salt present. The resulting deposits are hygroscopic and nonadherent; consequently, they are easily removed. Proper fuel treatment along these lines has enabled many gas turbines to operate wtisfactorily on residual fuel. Much literature har~appeared on the problem of vanadium corrosion. Referenoea {62] and 178-801 are prticulxrrly m m m e n d e d for further details. I n spite of the possibilities of fuel-treatment systems to control corrosion, most marine gas turbines have been operated on essentially vanadium-free distillate fuels. Trends iri world petroleum production have indicated that "multipurpose" distillate fuels, suitable for boilers, diesel engines, and gm turbines, may be~omeavailable at prices attractive for marine use. These particular distillates qmy contain traeeszf vanadium (under 2 ppm). It is not yet certain how low the vanadium must-be kept to avoid destructive attack of turbine components during reasonably long periods between overhauls. Sulfidation corrosion (commonly called hot corrosion) is a catastrophic type of oxidation which can be particularly harmful to hot-section components of marine turbines. The attack disrupts the normally protective oxides on superalloys, and sulfides may form deep into
832
MARINE ENGINEERING
Fig. 16 Sultldation attack on a blade from a marine gar turbine
grain boundaries causing serious degradation of the structural integrity of the alloy. Figure 16 is an example of a marine gas turbine blade damaged by sulfidation. , The immediate agent responsible for sulfidation is Na2S04, which can form during combustion from reactions between NaCl in sea air and sulfur in the fuel. However, the use of sulfur-free fuel will not eliminate the problem, because Na2S04 is also present in sea air. There is evidence that NaC1-NanSO4 mixtures are more corrosive than NazS04 alone. The sulfate must be present on the alloy in a condensed phase for attack to take place. At pressures near atmospheric, the accelerated attack may occur over the temperature range of 1400 to 1900 F. The results of research on sulfidation mechanisms are reviewed in references [81, 821. Several approaches can be used to combat sulfidation attack [83]. One is the use of fuel additives to promote formation of slags with melting points above turbine operating temperatures. Another approach is the use of metal-diffusion coatings such as the NiAl type. A
third is the use of demisters or salt separators to remove as much of the sea salt from the intake air as possible. Still another is the use of cooled components to lower the metal temperature. Finally, the inlet gas temperature can be limited to keep component temperatures within tolerable limits. Fortunately, new alloys have been developed which combine both high strength and sulfidation resistance. These probably will permit marine turbines to operate a t high inlet gas temperatures (-1750 F) for several thousand hours between overhauls. Indeed, these new alloys, helped by blade cooling, coatings, and/or demisters, hold much promise in promoting the widespread use of gas turbines aboard ship [81]. Other material problems in marine versions of aircraft, turbines have not been so difficult to overcome. Somc changes have been needed to make components resistant to the marine environment. These have included tha coating of some components, the substitution of stainles~ alloys for some miscellaneous parts, and the substitutiorl of aluminum alloys for magnesium alloys. Stainlcs~ steel or titanium should be used for compressor blading, both for stators and rotors. References [84-881 aro typical papers dealing with the marinization of gaH turbines. 5.9 Main Propulsion Shafting. Propulsion s h a h constitute one of the critical components in singltrscrew ships. Although there have been periods in which propeller losses caused by tailshaft failures have beon epidemic (1946-1950), such disabling casualties are now infrequent. However, lesser casualties in the form of surface deterioration and/or penetrating cracks do occur frequently [89]. These are of serious concern to bot,l~ ship owners and underwriters. ~ r o ~ u l s i oshaft n casualties are caused principally hy corrosion, fatigue, and fretting corrosion [40]. Steps aro usually taken to protect the shaft from seawater corrosion by applying protective coatings or covering^. However, complete integrity against the ingress of H(!ILwater cannot be assured. The junctions between tho covering and metallic components, such as bearing sleeves and propeller hubs, are particularly vulnerahl(!. Unfortunately, these are the areas where the nectl for protection is greatest. When seawater contacts the shaft surface, it invariul)ly causes general corrosion and pitting. Corrosio~i I)y itself is not serious, inasmuch as it rarely affects the ovtrrall strength of the shaft. I n most cases pits r ~ r ~ t l corroded areas can be ground smooth, filled with a put,Oy compound, and the protective covering reapplied. If the pits are unusually deep, they are ground out and t t r t r r ~ built up by welding. However, corrosion pits can sat I ~ H local sites for the initiation of fatigue cracks. Corro~ionfatigue cracks commonly occur in shafts a t the aft t r t ~ t l of the bearing sleeve and at the forward end of illlo propeller hub. This is a region of maximum fluctual,irly bending stresses produced by propeller weight antl/or off-center thrust. The amplitude of the bending s t r o ~ ~ t ~ u may be greatly increased when the ship is travclli~~y
,
CONSTRUCTION MATERIALS
light. Complete failure of the shaft in this region means the loss of the propeller and perhaps total disablement of the ship. Fortunately, fatigue cracks normally progress slowly enough to be detected during overhaul. Fatigue cracks initiating in tailshaft lteyways con: stitute another source of casualties. As discussed in Section 3.7, geometric discontinuities raise the local stress level and thus decrease fatigue life. It has been shown that detrimental effects can be reduced significantly by contouring the keyway ends [40, 891. Fretting corrosion is caused by minute movements between mating parts. Because of the microscopic welding and tearing actions which take place between the mating surfaces, small microcracks which act as sites for fatigue crack initiation readily form. Fretting in tailshafts generally occurs under the ends of the bearing sleeve and under the forward end of the propeller hub. Accordingly, fatigue cracks formed in these areas cannot be detected without cutting away the ends of the sleeve or removing the propeller. Experience has shown that the fatigue life of tailshafts subjected to fretting can be improved greatly by cold rolling the surface of the shaft [40, 901. Cold rolling may not eliminate the formation of microcracks; it does, however, produce residual compressive stresses in the surface layers of the shaft, and this retards the propagation of fatigue cracks if they, do form. The U.S. Navy generally cold rolls tailshafts that are 6 in. and over in diameter. Rolling is usually restricted to about one diameter on each side of the point where the propeller taper begins. The practice is also followed in the design of some merchant ships but has not been generally adopted by private ship designers, presumably because of the expense involved. Propulsion shafts for merchant ships usually are made from normalized and tempered plain carbon steel forgings conforming to the requirements of ABS Grade 2 steel [3]. Military ships usually employ higher strength, quenched and tempered alloy steel shafts conforming to Specification MIL-S-23284. High-strength steels such as Class 1 of MILS-23284 are not generally used for outboard shafting where the possibility of contact with seawater exists. In instances where they are used, the higher strength is not considered in the determination of the shaft size. The reason is that the corrosion-fatigue properties of the higher-strength steel are no better than those of a lower-strength steel (see Fig. 11). However, high-strength steels are used for inboard shafting to reduce the weight of the -shafting and to increase its flexibility. Weldability is an important characteristic of shafting materials as shafting is vulnerable to damage that can only be repaired by welding. For example, the removal of a corrosion-fatigue crack by grinding, with a subsequent build-up of material by welding, can greatly extend the life of many shafts. 5.10 Main Propulsion Gears. Although gears of all types are used aboard ship, the main propulsion gears are of greatest interest and concern to the marine
833
engineer. Gears generally fail from tooth breakage or deterioration of the tooth contact surfaces. Tooth breakage is caused by fatigue cracks which normally initiate in the root of the tooth. Accordingly, such factors as material strength, root radius, surface finish, surface treatment, and bending stresses may greatly affect the life and performance of the gears. Deterioration of the ' contact surfaces is normally caused by "pitting" a t or near the pitch line. Pitting is a fatigue-type failyre generated by compressive rolling loads when the tketh are meshing. This failure may originate from 'either surface or subsurface fatigue, depending on the circumstances. Once a crack appears a t the surface, the hydrodynamic action of the lubricating oil assists in its propagation. Eventually a piece of metal pops out, leaving a pit in the surface of the tooth. Pitting problems can arise in a properly designed gear from misalignment, helix-angle errors, tooth-spacing errors, and improper lubrication. It is not uncommon to have light pitting on the contact surfaces during the first few months of service. This pitting usually ceases after a brief period, and no further trouble is experienced. Pitting of this type is sometimes called "corrective" or "incipient" pitting. On the other hand, "heavy" or "destructive" pitting is extremely serious and normally requires immediate attention to prevent total destruction of the gear. From a fatigue standpoint, it is desirable to use steels having high tensile strength for the gear elements. However, the strength that can be developed in a steel frequently is limited by section size and available heattreating facilities. The rims of large bull gears usually are made from high-quality forgings that are either normalized and tempered or quenched and tempered to improve their strength and toughness. Both carbon and alloy steels-are used. Typical steels in current use for slow-speed gears are AISI-1030, 1035, 1050, 4137, and 4337. Higher hardness and strength can be developed in pinions because of their smaller size. Normalized and tempered AISI-1042 and quenched and tempered AISI4140 and 4340 are typical steels for pinion applications. 5.1 1 Propellers. The following properties and characteristics must be given consideration in selecting a propeller material : Castability Strength and Ductility Corrosion-Erosion Resistance Cavitation Resistance Weldability Cost Most propellers are casti&s made by pouring molten metal into intricate molds. The propellers are cast close to the required dimensions because their size and shape preclude extensive machining. To meet these demanding requirements, it is important that the metal have excellent castability, i.e., the ability to flow freely into mold cavities and to solidify into sound metal. The strength and ductility of the metal are important
834
MARINE ENGINEERING
in establishing the design and ,serviceability of a propeller. High strength is necessary to resist the applied forces and t o provide fatigue resistance. Blade loss, although infrequent, usually can be attributed to.fatigue caused by vibration or unsymmetrical loading (see Fig. 7). On the other hand, blade damage frequently occurs from striking submerged or floating objects. It is important in these circumstances that the propeller material have sufficient ductility, not only t o bend rather than break, but to undergo subsequent straightening and repair. Water and entrained debris move across the faces of ~ r o ~ e l lblades er a t high velocities. This action tends to scour away protective films and expose bare metal to corrosion. Thus, high corrosion-erosion resistance is required in propeller applications. Also, the metal must be resistant to cavitation damage should conditions conducive to this phenomenon arise. The steel hull of a ship, and protective zincs attached thereto, provide a degree of galvanic protection which helps to reduce propeller deterioration. It is desirable, if not mandatory, that the metal be weldable. Good weldability is required not only to repair damage incurred during service, but also to repair casting imperfections on the surface of the blades during manufacture. The weld metal must be compatible with the base metal from the standpoint of corrosion to prevent undesirable galvanic couples. Copper-base alloys are the most widely used metals for marine propellers. Certain of these alloys exhibit unique combinations of the desirable features mentioned in the foregoing, plus a built-in resistance to fouling produced by the dissolution of copper. Manganese bronze, similar to that specified in MIL-B-16443, was used almost exclusively for marine propellers until about 1957. Since then there has been a n increasing trend toward the use of higher strength nickel-aluminum bronze and manganese-nickel-aluminum bronze similar to alloys 1 and 2 of MILB-21230. From time to time other types of alloys, such as 12y0 chromium stainless steel and titanium alloys, have been used for marine propellers itl special applications. B y and large, however, the copper-base alloys wiIl continue to find the most widespread usage as marine propeller materials. 5.12 Low-Temperature and Cryogenic Equipmen!. The use of merchant ships to transport refrigerated cargo and liquefied gases poses special material requirements. The materials for storage tanks, pressure vessels, piping, etc., must be capable of providing safe handling a t the
low temperatures encountered. I n addition, the equipment must be resistant to chemical attack by the cargo being handed. Two areas of material-application technology are recognized. These are "low-temperature" a n d "cryogenics," which involve materials for applications to -150 F and -459 I?, respectively. The low-temperature materials are those especially suited for handling relatively "warmn liquefied gases such as propane, anhydrous ammonia, carbon dioxide, and ethane. The cryogenic inaterials are capable of handling liquefied methane, oxygen, nitrogen, argon, hydrogen, and helium. Because of their relatively low cost and ease of fabrication, steels have been widely used for low-temperature and cryogenic applications [911. Not all steels, however, are suitable for these applications. Many of the lowalloy ferritic steels are susceptible to '%rittlen fracture a t low temperatures, and their so-called "notch toughness" must be thoroughly evaluated before use. Notch toughness is the ability of a metal containing a notch or crack to yield plastically under constraint and high local stress. This property is a prime requisite of metals for low-temperature applications and is vital to the selection of materials for the handling, transportation, and storage of liquefied gases. Methods of determining this property are given in reference [ll.]. The notch toughness of most ferritic steels decreases with decreasing temperature. Ferritic steels usually undergo a sharp transition from ductile to brittle fracture [92]. This does not preclude their use if great care ia exercised in selecting an alloy for a specific application. The ferritic 9yo Ni steel, ASTM A353, was developed specifically for cryogenic applications down to liquid nitrogen temperatures (-320 F). The austenitic Cr-Ni stainless steels of the AISI 300 series show no marked decrease in toughness with decreasing temperature. Furthermore, they are highly corrosion resistant. Accordingly, they find wide um over the entire range of cryogenic applications. Ferritic and martensitic stainless steels generally are not recommended for cryogenic use. The toughness of most nonferrous alloys is unaffected by cryogenic temperatures. Aluminum alloys have bee11 used extensively, because of their fabricability and low coat [93]. I n particular, the non-heat-treatable aluminum-magnesium alloys of the 5000 s&ies are umd widely for cryogenic tankage.
i
CONSTRUCTION MATERIALS
Sectba 6
Glossary of Metallurgical T e n s UsEd in Matefirls Engineering2 reference electrode of the composition: Pt/Hg-HgzClz/KCl solution. For one normal KC1 solution its potential versus hydrogen at 25 f= and one atmosphere is +0.280 volts. Carbide. A compound of carbon'with one or more metallic elements. Carbonitriding. Introducing carbon and nitrogen into a solid ferrous alloy by' holding above Acl (see transformation temperature) in an atmosphere that contains suitable gases such as hydrocarbons, carbon monoxide, and ammonia. The carbonitrided alloy is usually quench hardened. Carburizing. Introducing carbon into a solid ferrous alloy by holding above Acl (see transformation temperature) in contact with a suitable carbonaceous material, which may be a solid, liquid, or gas. The carburised alloy is usually quench hardened. Carbon Steel. Steel containing carbon up to about 2% and only residual quantities of other elements except those added for deoxidation, with silicon usually limited to 0.60% and manganese to about 1.65%. Also termed "plain carbon steel," "ordinary steel,'' and "straight carbon steel." Case Hardening. Hardening a ferrous alloy so that the outer portion, or case, is made substantially harder than the inner portion or core. Typical processes used for case hardening are carburizing, cyaniding, carbonitriding, nitriding, induction hardening, and flame hardening. Cast Iron. An iron containing carbon in escess of the solubility in the austenite that exists in the alloy at the eutectic temperature. For the various forms of gray cast iron, white cast iron, malleable cast iron, and nodular cast iron, the word "cast" is often left out, resulting in "gray iron," "white iron," "malleable iron," and "nodular iron," respectively. Cathode. The electrode where electrons enter (current leaves) an operating system such as a battery or an eleetrolytic cell. In a battery or electrolytic cell, it is the electrode where reduction occurs. Contrast with anode. Cathodic Protection. Partial or complete protection of a metal from corrosion by making it a cathode, using either a galvanic or impressed current. Cation. A positively charged ion; it flows to the cathode in electrolysis. Cavitation. The formation and instantaneous collapse of innumerable tiny voids or cavities within a liquid subjected to rapid and intense pressure changes. Cavitation produced by ultrasonic radiation is sometimes used to give violent localized agitation. That caused by severe turbulent flow often leads to cavitation damage. Cavitation Damage. Wearing away of metal through the formation and collapse of cavities in a liquid. Cavitation Erosion. See preferred term, cavitation damage. Charpy Test. A pendulum-tyfie single-blow impact test in which the specimen, usually notched, is supported at both ends as a simple beam and broken by a falling pendulum. The energy absorbed, as determined by the subsequent rise of the pendulum, is a measure of impact strength or notch toughness. Clad Metal. A composite metal containing two or three layen that have been bonded together. The bonding Selected from thelist ofdefinitionspublishedinMet& Handbook, may have been accomplished by co-rolling, welding, 8th Ed., Vol. 1, 1961, by the American Society for Metals.
Acid Embrittlement. A form of hydrogen embrittlemat whicK may be induced in some metals by acid treatment. Age Hardening. Hardening by aging, usually after rapid cooling or cold working. See aging. Aging. In a metal or alloy, a change in properties that generally occurs slowly at room temperature and more rapidly at higher temperatures. Air-Hardening Steel. A steel containing sufficient carbon and other alloying elements to harden fully during cooling in air or other gaseous medium from a temperature above its transformation range. Alloy. A substance having metallic properties and being composed of two or more chemical elements of which a t least one is an elemental metal. Alloying Element. An element added to a metal to effect changes in properties, and which remains within the metal. Alloy Steel. Steel containing significant quantities of alloying elements (other than carbon and the commonly accepted amounts of manganese, silicon, sulfur and phosphorus) added to effect changes in the mechanical or physical properties. Anion. A negatively charged ion; it flows to the anode in electrolysis. Annealing. Heating to and holding at a suitable temperature and then cooling at a suitable rate, for such purposes as reducing hardness, improving mchinability, facilitating cold working, producing a desired microstruc.ture, or obtaining desired mechanical, physical, or other properties. Anode. The electrode where electrons leave (current enters) an operating system such as a battery or an electrolytic cell. In a battery or electrolytic cell, it is the electrode where oxidation occurs. Contrast with cathode. Anode Corrosion. The dissolution of a metal acting as an anode. Austenite. A solid solution of one or more elements in facecentered cubic iron. Unless otherwise designated (such as nickel austenite), the solute is generally assumed to be carbon. Austenitic Steel. An alloy steel whose structure is normally austenitic at room temperature. Austenitizing. Forming austenite by heating a ferrous alloy into the transformation range (partial austenitising) or above the transformation range (complete austenitising). Bainite. A decomposition product of austenite consisting of an aggregate of ferrite and carbide. In general, it forms at temperatures lower than those where very fine pearlite forms and higher than that where martensite begins to form on cooling. Its appearance is feathery if formed in the upper part of the temperature range and acicular, resembling tempered martensite, if formed in the lower part. Brazing. Joining metals by flowing a thin layer, capillary thickness, of nonferrous filler metal into the space between them. Brittle Fracture. Fracture with little or no plastic deformation. Calomel Electrode (Calomel Half Cell). A secondary
.
836
1
3N MATERIALS
MARINE EN
casting, heavy chemical deposition, or heavy electroplating: Cleavage Fracture. A fracture, usually of a polycrystalline metal, in which most of the grains have failed by cleavage, resulting in bright reflecting facets. I t is one type of crystalline fracture. Contrast with shear fracture. Controlled Cooling. Cooling from an elevated temperature in a redetermined manner to avoid hardening, cracking, or iniernal damage or to produce a desired microstructure. This cooling usually follows a hot-forming operation. Cooling Stresses. Residual stresses resulting from nonuniform distribution of temperature during cooling. Corrosion. The deterioration of a metal by chemical or electrochemical reaction with its environment. Corrosion Embrittlement. The severe loss of ductility of a metal resulting from corrosive attack, usually intergranular and often not visually apparent. Corrosion Fatigue. Effect of the application of repeated or fluctuating stresses in a corrosive environment characterized by shorter life thanwould beencountered as a result of either the repeated or fluctuating stresses alone or the corrosive environment alone. Creep. Time-dependent strain occurring under stress. The creep strain occurring a t a diminishing rate is called primary creep; that occurring a t a minimum and almost constant rate, secondary creep; that occurring at an accelerating rate, tertiary creep. Creep Limit. (I) The maximum stress that will cause less than a specified quantity of creep in a given time. (2) The maximum nominal stress under which the creep strain rate decreases continuously with time under constant load and a t constant temperature. Sometimes used synonymously with creep strength. Creep Recovery. Time-dependent strain after release of load in a creep test. Creep-Rupture Test. Same as stress-rupture test. Creep Strength. (I) The constant nominal stress that will cause a specified quantity of creep in a given time at constant temperature. (2) The constant nominal stress that will cause a specified creep rate a t constant temperature. Crevice Corrosion. A type of concentration-cell corrosion; corrosion of a metal that is caused by the concentration of dissolved salts, metal ions, oxygen or other gases, and such, in crevices or pockets remote from the principal fluid stream, with a resultant building up of differential cells that ultimately cause deep pitting. Crystalline Fracture. A fracture of a polycrystallime metal characterized by a grainy appearance. Contrast with jibrous fracture. Damping Capacity. The ability of a metal to absorb vibration (cyclical stresses) by internal friction, converting the mechanical energy into heat. Decarburization. The loss of carbon from the surface of a ferrous alloy as a result of heating in a medium that reacts with the carbon at the surface. Deoxidizer. A substance that can be added to molten metal t o remove either free or combined oxygen. Dezincification. Corrosion of some copper-zinc alloys involving loss of zinc and the formation of a spongy porous copper. Ductile Iron. See nodular cast iron. Ductility. The ability of a material to deform plastically without fracturing. Measured by elongation or reduction of area in a tensile test, by height of cupping in an Erichsen test, or by other means.
837
* Electrolyte. (1) An ionic conductor. (2) A liquid, most often a solution, that will conduct an electric current. Electromotive Series. A list of elements arranged according to their standard electrode potentials. In corrosion studies the analogous but more practical galvanic series of metals is generally used. The relative position of a given metal is not necessarily the same in the two series. Embrittlement. Reduction in the normal ductility of a metal due to a physical or chemical change. Endurance Limit. Same as fatigue limit. Erosion. Destruction of metals or other materials by the abrasive action of moving fluids, usually accelerated by the presence of solid particles or matter in suspension. When corrosion occurs simultaneously, the term erosion-corrosion is often used. Exfoliation. A type of corrosion that progresses approximately parallel to the outer surface of the metal, causing layers of the metal to be elevated by the formation of corrosion product. Fatigue. The phenomenon leading to fracture under repeated or fluctuating stresses having a maximum value less than the tensile strength of the material. Fatigue fractures aro progressive, beginning as minute cracks that grow under the action of the fluctuating stress. Fatigue Life. The number of cycles of stress that can ba sustained prior to failure for a stated test condition. Fatigue Limit. The maximum stress below which a material can presumably endure an infinite number of stress cycloti. If the stress is not completely reversed, the value of t l ~ o mean stress, the minimum stress, or the stress ratio shoultl be stated. Fatigue Strength. The maximum stress that can be sustaincd for a specified number of cycles without failure, the stress being completely reversed within each cycle unless otherwise stated. Fatigue-Strength Reduction Factor (Kj). The ratio of llitr fatigue strength of a member or specimen with no strortn concentration to the fatigue strength with stress concerltration. Kj has no meaning unless the geometry, size, ant1 material of the member or specimen and stress range ~ ~ r o stated. Ferrite. A solid solution of one or more elements in bodycentered cubic iron. Unless otherwise designated (for instance, as chromium ferrite), the solute is generally assumed to be carbon. On some equilibrium diagr:~lnrt there are two ferrite regions separated by an austenittr area. The lower area is alpha ferrite; the upper, dcllr~ ferrite. If there is no designation, alpha ferrite is assumod. Fibrous Fracture. A fracture where the surface is charactorized by a dull gray or silky appearance. Contrast wild^ crystalline fracture. Flakes. Short discontinuous internal fissures in ferro~ln metals attributed to stresses produced by localized trannformation and decreased solubility of hydrogen duri~lg cooling after hot working. In a fractured surface, flakon appear as bright silvery areas; on an etched surface thoy appear as short discontinuous cracks. Also called "shntlor cracks" and "snowflakes." Flame Hardening. Quench hardening in which the heat in applied directly by a flame. Fretting (Fretting Corrosion). Action that results in surfnt:tr damage, especially in a corrosive environment, when thoroo is relative motion between solid surfaces in contact uniIols pressure.
Galling. A condition developed on the rubbing surface of one or both mating parts where excessive friction between high spots results in localized welding with subsequent spalling and a further roughening of the surface. Galvanic Cell. A cell in which chemical change is the source of electrical energy. I t usually consists of two dissimilar conductors in contact with each other and with an electrolyte, or of two similar conductors in contact with each other and with dissimilar electrolytes. Galvanic Corrosion. Corrosion associated with the current of a galvanic cell consisting of two dissimilar conductors in an electrolyte or two similar conductors in dissimilar electrolytes. Where the two dissimilar metals are in contact, the resulting reaction is referred to as "couple action." Galvanic Series. A series of metals and alloys arranged according to their relative electrode potentials in a specified environment. Compare with electromotive series. Grain. An individual crystal in a polycrystalline metal or a110y. Grain Growth (Coarsening). An increase in the size of grains in polycrystalliie metal, usually effected during heating a t elevated temperatures. The increase may be gradual or abrupt, resulting in either uniform or nonuniform grains after growth has ceased. Grain Size. For metals, a measure of the areas or volumes of grains in a polycrystalline material, usually expressed as an average when the individual sizes are fairly uniform. Grain sizes are reported in terms of number of grains per unit area or volume, average diameter, or as a grain-size number derived from area measurements. Graphitization. Formation of graphite in iron or steel. Where graphite is formed during solidification, the phenomenon is called "primary graphitization"; where formed later by heat treatment, "secondary g~aphitization." Gray Cast Iron. A cast iron that gives a gray fracture due to the presence of flake graphite. Often called gray iron. Half Cell. An electrode immersed in a suitable electrolyte, designed for measurements of electrode potential. Hard Drawn. Temper of copper or copper alloy tubing drawn in excess of 25% reduction in area. Hardenability. In a ferrous alloy, the property that determines the depth and distribution of hardness induced by quenching. Hardness. Resistance of metal to plastic deformation usually by indentation. However, the term may also refer to stiffness or temper, or to resistance to scratching, abrasion, or cutting, Indentation hardness may be measured by various hardness tests, such as Brinell, Rockwell, and Vickers. Heat-Affected Zone. That portion of the base metal which was not melted during brazing, cutting, or welding, but whose microstructure and physical properties were altered by the heat. Hydrogen Embrittlement. A condition of low ductility in metals resulting from the absorption of hydrogen. Impact Energy (Impact Value). The amount of energy required to fracture a material, usually measured by means of an Izod or Charpy test. The type of specimen and testing conditions affect the values and therefore should be specified. Induction Hardening. Quench hardening in which the heat is generated by electrical induction. Intercrystalline. Between the crystals, or grains, of a metal.
Intergranular Corrosion. Corrosion occurring preferentially at grain boundaries. Intracrystalline. Within or across the crystals or grains of a metal; same as transcrystalline and transgranular. -Killed Steel. Steel deoxidized with a strong deoxidizing agent such as silicon or aluminum in order to reduce the oxygen content to such a level that no reaction occurs between carbon and oxygen during solidification. Local Cell. A galvanic cell resulting from inhomogeneities between areas on a metal surface in an electrolyte. The inhomogeneities mg+ be of physical or chemical nature in either the metal m its environment. Macrograph. A graphic reproduction of the surface of a prepared specimen at a magnification not exceeding ten diameters. When photographed, the reproduction is known as a photomacrograph. Macroscopic. Visible at magnifications from one to ten diameters. Macrostructure. The structure of metals as revealed by examination of the etched surface of a polished specimen a t a magnification not exceeding ten diameters. Malleable Cast Iron. A cast iron made by a prolonged anneal of white cast iron in which decarburization or graphitization, or both, take place to eliminate some or all of the cementite. The graphite is in the form of temper carbon. If decarburization is the predominant reaction the product will have a light fracture, hence, "whiteheart malleable"; otherwise, the fracture will be dark, hence, "blackheart malleable." "Pearlitic malleable" is a blackheart variety having a pearlitic matrix along with perhaps some free ferrite. Martensite. A metastable phase of steel, formed by a transformation of austenite below the M , (or Ar") transformation temperature. I t is an interstitial supersaturated solid solution of carbon in iron having a body-centered tetragonal lattice. Its microstructure is characterized by an acicular, or needle-like, pattern. Martempering. Quenching an austenitized ferrous alloy in a medium a t a temperature in the upper part of the martensite range, or slightly above that range, and holding it in the medium until the temperature throughout the alloy is substantially uniform. The alloy is then allowed to cool in air through the martensite range. Mechanical Hysteresis. Energy absorbed in a complete cycle of loading and unloading within the, elastic limit and represented by the closed loop of the stress-strain curves for loading and unloading. Sometimes referred to as elastic, but more properly, mechanical. Micrograph. A graphic reproduction of the surface of a prepared specimen, usually etched, a t a magnification greater than ten diameters. If produced by photographic means it is called a photomicrograph (not a microphotograph). Nitridiig. Introducing nitrogen, into a solid ferrous alloy by holding a t a suitable tempergture (below ACI transformation temperature for ferritic steels) in contact with a nitrogenous material, usually ammonia or molten cyanide of appropriate composition. Quenching is not required to produce a hard case. Nodular Cast Iron. A cast iron tha't has been treated while molten with a master alloy containing an element such as magnesium or cerium to give primary graphite in the spherulitic form.
MARINE ENGINEERING 3Tonnatizing. Heating a ferrous alloy to a suitable temperature above the transformation range and then cooling in
Sacrificial Protection. Redding the extent of corrosion of a metal in an electrolyte by coupling it to another metal that
fracture at points of stress concentration. For example, in a notch tensile test, the material is said to be "notch brittle" if the notch strength is less than the tensile strength. Otherwise, it is said to be "notch ductile." Notch Rupture Strength. The ratio of applied load to original area of the minimum cross section in a stress-rupture test of a notched specimen. Notch Sensitivity. A measure of the reduction in strength of a metal caused by the presence of stress concentration. Values can be obtained from static, impact, or fatigue tests. Overaging. Aging under conditions of time and temperature greater than those required to obtain maximum change in a certain property, so that the property is altered in the direction of the initial value. See aging. Passivation. The changing of the chemically active surface of a metal to a much less reactive state. Passivity. A condition in which a piece of metal, because of
applied to stress-corrosion cracking of brass. Shot Peening. Cold working the surface of a metal by metalshot impingement. Silver Brazing. Brazing with silver-base alloys as the filler metal. Size Effect. Effect of the dimensions of a piece of metal upon its mechanical and other properties and upon manufacturing variables such as forging reduction and heat treatment. In general, the mechanical properties ara lower for a larger size. Slack Quenching. The process of hardening steel by quenching from the austenitizing temperature at a rate slower than the critical cooling rate for the particular steel, resulting in incomplete hardening and the formation of one or mom transformation products in addition to or instead sf martensite. Solution Heat Treatment. Heating an alloy to a suitablo
CONSTRUCTION MATERIALS The temperature a t which austenite transforms to delta ferrite during heating. Ae,,, Ael, Aer, Ae4. The temperatures of phase changes a t equilibnum. AT,,. In hypereutectoid steel, the temperature at which precipitation of cementite starts during coolmg. Arl. The temperature a t which transformation of austenite to ferrite or to ferrite plus cementite is completed / during cooling. Ar3. The temperature a t which austenite begins to transform to ferrite dudng cooling. Ar4. The temperature a t which delta ferrite transforms to austenite during cooling. 1M,(or AT"). The temperature at which transformation of austenite to martensite starts during cooling. Mf. The temperature a t which martensite formation finishes during cooling.
stressJ' denotes the stress where force and area are measured at the same time. "Conventional stress," aer applied to tension and compression tests, is force divided by the original area. "Nomind stress" is the stress computed by simple elasticity formulas, ignoring stress raisers and disregarding plastic flow; in a notch bend test, for example, it is bending moment divided by minimum section modulus. Stress Concentration Factor, Kt. The ratio of the greatest stress, in the region of a notch or other stress raiser as determined by advanced theory, photoelasticity or direct measurement of elastic strain, to the corresponding nominal stress. Stress-Gorrosion Cracking. Failure by cracking under combined action of corrosion and stress either external (applied) or internal (residual). Cracking may be intergranular and/or transgranular, depending on metal and corrosive medium.
or electrochemical reaction.
Radiography. A nondestructive method of internal examinaa t least 0.25%; ingot iron contains considerably less. tion in which metal or other objects are exposed to a beam Strain. A measure of the change in the size or shape of IL of x-ray or gamma radiation. Differences in thickness, body, referred to its original size or shape. "Linear straia" density or absorption, caused by internal discontinuities, is the change per unit length of a linear dimension. "'J'ruc~ are apparent in the shadow image either on a fluorescent strain" (or "natural strainJJ) is the natural logarithm of screen or on photographic film placed behind the object. tlie ratio of the length at the momerit of observation to thn Recrystallization. (1) The change from one crystal structure original gage length. "Conventional strain" is the lino~rr. to another, as occurs on heating or cooling through a strain referred to the original gage length. "Shearing critical temperature. (2) The formation of a new, strainstrain" (or "shear strain") is the change in angle (exprn~ucttl free grain structure from that existing in cold-worked in radians) between two limes originally a t right an~loa. metal, usually accomplished by heating, When the term strain is used alone, it usually refers to t 1 1 ~ Refractorg Alloy. (1) A heat-resistant alloy. (2) An alloy linear strain in the direction of the applied stress. having an extremely high meltiqg point. (3) An alloy Strain Hardening. An increase in hardness and strongldr difficult to work at elevated temperatures. caused by plastic deformation a t temperatures lower ~ I I I L I I Relaxation. Relief of stress by creep. Some types of tests the recrystallization range. are designed to provide diminution of stress by relaxation Stress. Force per unit area, often thought of as force ao0i11~ a t constant strain, as frequently occurs in service. through a small area within a plane. It can be divhlcvl Residual Stress. Stress present in a body that is free of into components, normal and parallel to the plane, cellatl external forces or thermal gradients. "normal stress" and "shear stress," respectively. "'J'~IIo
during cold working. Temper Brittleness. Brittleness that results when certain steels are held within, or are cooled slowly through, a certain range of tempemture below the transformation range. The brittleness is revealed by notched-bar impact tests at or below room temperature. Tempering. Reheating a quench-hardened or normalized ferrous alloy to a temperature below the transformation range and then cooling a t any rate desired. Thermal Fatigue. Fracture resulting from the presence of temperature gradients which vary with time in such a manner as to produce cyclic stresses in a structure. Toughness. Ability of a metal to absorb energy and deform plastically before fracturing. It is usually measured by the energy absorbed in a notch impact test, but the area under the strewtrain curve in tensile testing is also a measure of toughness. Transformation Temperature. The temperature a t which a change in phase occurs. The term is sometimes used to denote the limiting temperature of a transformation range. The following symbols are used for iron and steels: Ac,,. In hypereutectoid steel, the temperature a t which the solution of cementite in austenite is completed during
to austenite is completed during heating.
include brazing. White Metal. A general term covering a group of whitecolored metals of relatively low melting points (lead, antimony, bismuth, tin, cadmium, and zinc) and of the alloys based on these metals. Work Hardening. Same as strain hardening.
References
~Regulations," ~ U.S. Coast i ~ ~ d . ~ ~ ~ , , hfor ~ yips, d ~ l ~~ ~valves, b ~t i t~t and b ~~ ~ Symbols, " Maritime Administration. 3 Rules for the Classification and Construction of Steel Vessels, American B u r q u of Shipping. 4 "Naval Ships Technic61 Manual," Naval Ships Systems U.S. Navy' lor Construction Of 'resswe vesselS1" 5 ASME Boiler and Pressure Vessel Code.
1 #(Marine ~
~
~
MARINE ENGINEERING
(
9 Metals Handbook, American Society for Metals. 10 "Material Specifications," ASME Boiler and Pressure Vessel Code. 11 ASTM Standards, American Society for Testing and Materials. 12 "Steel Products Manual," American Iron and Steel Institute. 13 H. H. Uhlig, The Corrosion Handbook, 2nd Edition, John Wiley and Sons, Inc., New York, 1948. 14 H. H. Uhlig, Corrosion and Corrosion Control, An Introduction to Corrosion Science and Engineering, John Wiley and Sons, Inc., New York, 1963. 15 F. L. LaQue and H. R. Copson, Corrosion Resistance of Metals and -Alloys, Reinhold Publishing Corp., New York, 1963. 16 U. R. Evans, Il4etallic Corrosion, Passivity and Protection, Longmans, Green and Co., New York, 1948. 17 U. R. Evans, Corrosion and Oxidation of Metals, St. Martins Press, New York, 1960. 18 E. Rabald, Corrosion Guide, American Elsevier Publishing Co., New Yorlr, 1951. 19 Metal Corrosion in the Atmosphere, ASTM Special Technical Publication 435, 1968.
21 F. L. LaQue and A. H. Tuthill, "Economic Considerations in the Selection of Materials for Marine Applications, " Trans. SNAME, 1961. 22 F. L. LaQue, "Corrosion Testing," Proc. ASTM, 23 F. L. LaQue, "The Behavior of Nickel-Copper Alloys in Sea Water," Journal ASNE, February, 1941. 24 W. L. Williams, "Metals for Hydrospace," Journal of Materials, ASTM, December, 1967. 25 G. J. Danek, Jr., "The Effect of Sea-Water Velocity on Corrosion Behavior of Metals," Naval Engineers Journal, October, 1966. 26 Erosion by Cavitation or Impingement, ASTM Special Technical Publication 408, 1967. 27 W. L. Williams, "Shipboard Salt Water PipingInfluence of Material and Design on Performance," Bureau of Ships Journal, August, 1959. 28 R. B. Niederberger, "Composition and Heat Treatment Effect on Dealuminization of Aluminum Bronzes, " Modern Castings, March, 1964. 29 Stress-Corrosion Testing, ASTM Special Technical Publication 425, 1967. 30 W. L. Williams and J. F. Eckel, Corrosion and Wear Handbook Jor Water-Cooled Reactors, edited by D. J. DePaul, New York, McGraw-Hill Book Co., 1957. 31 Symposium on Stress-Corrosion Cracking of Metals, ASTM & AIME, 1945. 32 B. F. Brown and L. S. Birnbaum, "Corrosion Control for Structural Metals in the Marine Environment," Naval Research Laboratory Report 6167, November, 1964. 33 A. R. Elsea and E. E. Fletcher, "HydrogenInduced Brittle Failures of High Strength Steels,"
Defense Metals Information Center Report 196,January, 1964. 34 D. 0. Sprowls and R. H. Brown, "What Every Engineer Should Know About Stress Corrosion of Aluminum," Metal Progress, April and May, 1962. 35 I. R. Lane, J. L. Cavallaro, and A. G. S. Morton, Stress-Corrosion Cracking of Titanium, ASTM Special Technical Publication 397, 1966. 36 B. F. Brown, "A New Stress-Corrosion Cracking Test for High Strength Alloys," Materials Research and Standards, March, 1966. 37 Report on Stress Corrosion Cracking of Austenitic Chromium-Niclcel Stainless Steels, ASTM Special Technical Publication 264, 1960. 38 Manual on Fatigue Testing, ASTM Special Technical ~ublication91, 1949. 39 A Guide for Fatigue Testing and the Statistical Analysis of- Fatigue Data, ASTM Special Technical Publication 9 1 4 , 1963. 40 Large Fatigue Testing Machines and Their Results, ASTM Special Technical Publication 216, 1957. 41 Manual on Low-Cycle Fatigue Testing, ASTM Special Technical Publication 465, 1970. 43 G. Sines and J. L. Waisman, Metal Fatigue, McGraw-Hill Book Co., New York, 1959. 44 J. J. Burke, et al., Fatigue-An Interdisciplinary Approach, Syracuse University Press, Syracuse, N.Y., 1964. 45 A. F. Madayag, Metal Fatigue: Theory and Design, John Wiley and Sons, Inc., New York, 1969. 46 T. R. Gurney, Fat.igue of Welded Structures, Cambridge University Press, Cambridge, England, 1968. 47 Fatigue Design Handbook, Society of Automotive Engineers, 1968. 48 J . 0. Almen and P. H. Black, Residual Stresses and Fatigue in Metals, McGraw-Hill Book Co., New York, 1963. 49 R. C. Schwab and E. J. Czyr~ca,Eflecls of Environment and Complex Load History on Fatigue Life, ASTM Special Technical Publication 462, 1970. 50 R. E. Peterson, Stress Concentration Design Factors, John Wiley and Sons, Inc., New York, 1953. 51 P. M. Brister and W. E. Leyda, Proc.-Vol. I: Joint Interfiational Conference on Creep, ASME-ASTMIME, London, 1963. 52 E. L. Robinson, "The ~esis6inceto Relaxation of Materials a t High Temperature," Trans. ASME, 1939. 53 E. L. Robinson, "High-Temperature Bolting Materials," Proc. ASTM, 1945. 54 Relaxation Properties of Steels and Super-Strength Alloys at Elevated Temperatures, ASTM Special Technical Publication 187, 1956, and Data Series. Publication 14, 1961. 55 A. H. Rosenstein and W. H. Asche, Temper Embrittlement in Steel, ASTM Special Technical Publication 407, 1965. 56 S. S. Manson, Proc.-Vol. 11:Joint International
CONSTRUCTION MATERIALS
84 1
Conference on Creep, ASME-ASTM-IME, London, 1963. 57 G. V. Smith, "Evaluation of Elevated Temperature Strength Data," Journal of Materials, ASTM, December, 1969. 58 An Evaluation of the Yield, Tensile, Creep and Rupture Strengths of Wrought 304, 316, 391 and 347 Stainless Steels at Elevated Temperatures, ASTM Data Series Publication DS5S2, 1969. 58 Symposium on Eflect of Cyclic Heating and Stressing on Metals at Elevated Temperatures, ASTM Special Technical Publication 165, 1954.
for Seawater Service," Journal of Engineering for Power, Trans. ASME, October, 1966, 75 N. G. Feige and R. L. Kane, "The Application of Titanium for Desalination Plants," Metals Engineering Quarterly, August, 1967. 76 W. H. Hatfield, G. Stanfield, and L. Rotherham, "The Damping Capacity of Engineering Materials," Engineering, June, 1942. ' ..' 77 D. W. James, "High Damping Metals for Engineering Applications,:' Materials Science and Engineering,
Temperatures and Effect of NowSteady Load and Temperature Conditions on the Creep of Metals, ASTM Special Technical Publication 260, 1959. 61 Literature Survey on Creep Damage in Metals, ASTM Special Technical Publication 391, 1965. 62 W. J. Greenert, "High Temperature Slag Corrosion of Metallic Materials," Corrosion, February and
Physical Factors on the Formation of Deposits from Residual Fuels," Trans. ASME, 1955. 79 B. 0. Buckland and D. G. Sanders, "Modified Residual Fuel for Gas Turbines," Trans. ASME, 1955. 80 H. W. Schab, "A Bibliography of a Decade of R e s e a r ~ ~ oOil-Gas n Corrosion by Heavy Fuels, " Journal ASNE, November, 1958.
Potential of Light Water Reactors, AEC Publication WASH 1082, 1968. 64 P. Cohen, Coolant Chemical Technolugy of Aqueous Heterogeneous Reactor Systems, Gordon and Breach Science Publishers, Inc., New York, 1969. 65 G. D. Whitman, G. C. Robinson, and A. W. Savolainen, Technology of Steel Pressure Vesselsfor WaterCooled Nuclear Reactors, Nuclear Safety Information Center, Oak Ridge, 1967. 66 The Eflecls of Radiation on Structural Metals, ASTM Special Technical Publication 426, 1967. 67 L. E. Steele, "Neutron Irradiation Embrittlement of Reactor Pressure-Vessel Steels," Atomic Energy Review, Vienna, 1969. 68 W. L. Williams, "Chloride and Caustic Stress
82 Hot Corrosion Problems Associated with Gas Turbines,ASTM Special Technical Publication 421,1967. 83 G. J. Danek, Jr., "State-of-the-Art Survey on Hot Corrosion in Marine Gas Turbine Engines," Naval Engineers Journal, December, 1965. 84 R. W. Stuart Mitchell, "The Influence of Current Research and Development on the Application of the Open-Cycle Gas Turbine as a Propulsion Unit for Merchant Ships, " ASME Paper 66-GT/M33, 1966. 85 E. E. Stoeckly, "Development of the General Electric LM-1500 Gas Turbine as a Marine Power Plant, " Journal of Engineering for Power, Trans. ASME, April, 1966. 86 N. K. Bowers, "Gas Turbines in the Royal Navy, " Journal of Engineering for Power, Trans. ASME,
Steam, " Corrosion, August, 1957.
73 W. Williams, The Science, Technology and Application of Titanium, edited by R. I. Jaffee and N. E. Promisel, Pergamon Press, Ltd., London, 1969. 74 R. B. Niederberger, "Condenser Tube Materials
92 D. F. MacNaught, Principles of Naval Arckitecture, edited by J. P. Cornstack, SNAME, New York, 1967. 93 Aluminum, edited by K. R. Van Horn, American Society for Metals, Metals Park, 1967.
I
1
1 ,
I
I r
PETROLEUM FUELS
CHAPTER X X l l l I
VAPOR FEED
GAS PLUS
LIQUID FEED
LIGHT FUEL OIL
Section 1 Fuel Manufacture and Characteristics 1.1 Scope. This chapter deals with fuels which are used in main and auxiliary internal combustion engines and boilers aboard ship. Since wood- and coal-burning ships have virtually disappeared from the w s , these fuels are not discussed. On the other hand, nuclear fuels are a subject in themselves and a discussion of this source of energy, and the means employed to harness it, are presented in Chapter 4. This chapter, therefore, is devoted to petroleum fuels and covers the source and manufacture of the various types in use, the means employed to define their ~ r o ~ e r t i ethe s , relation of these properties to power-plant requirements, and lastly a digest of the worldwide bunkering system and fuel availability picture. 1.2 Distillation Processes. The source of petroleum fuels is crude oil, which consists of hydrocarbons ranging from very light volatile gases to heavy residues such as the familiar asphalt used in road paving. The refining process by which a host of products is produced can be considered essentially as a series of steps aimed a t separating hydrocarbons, rearranging certain types to produce more desirable hydrocarbons, treating to remove undesirable constituents, arad introducing additives to accentuate existing properties and/or impart totally new
The first basic step in petroleum refining is the separation of the crude oil into desired components or fractions. This is called atmospheric distillation and can be likened to boiling water in a tea kettle and condensing the steam on a cold surface which returns the steam tc liquid again. Water, of course, has a single boiling point. Crude oil, on the other hand, consists of thousands of compounds of different boiling points and, therefore, a more elaborate distillation and condensation process is involved. During the distillation process, the crude oil is heated in a pipe still to a temperature sufficient to produce the desired degree of vaporization, after which it paases into a fractioning column as illustrated by Fig. 1. There is a gradation of temperature in the fractioning column from top to bottom; therefore, a t any location in the tower, lighter materials remain in a vapor state and rise to the top, whereas the higher-boiling materials condense and fall to the bottom. On each fractionating tray there is continuous boiling and condensation which forms an equilibrium mixture of a given boiling range. The trays and drawoffs are located a t
FRACTIONATING
Fig. 2 Air-lift h m f w catalytic ua&ing process
HEATER "
CR,'DE
STRIPPING STREAM
Fig. I Atmospheric distillation process
points in the column which yield cuts of the desired boiling range. At atmospheric pressures, a pipe-still distillation is limited to about 700 F, since higher temperatures result in cracking of some of the heavy hydrocarbons and the formation of materials which are not desired a t this stage of refining. If higher-boiling frac~ionsare ulanted, a "vacuum tower" is used, which is essentially the same type of equipment except that it is operate,d under reduced pressures. This, of course, results in the hydrocarbons boiling a t lower temperatures and eliminates the high-temperature cracking condition. The heavy hydrocarbons which are recovered in the vacuum tower are unchanged from their composition in the original crude. Other refining processes follow either of these primary distillation steps and are aimed a t cracking and rearranging hydrocarbons to produce products with various properties. One such process is called "visbreaking" where the charge stock, usually a residuum from the
.
atmospheric distillation process, is subjected to high temperatures (856 F) and pressures in excess of 1000 psi. The hrge molecules of the heavy hydrocarbons break up under these conditions, and the resulting lighter materials are separated in a fractionating column. The remaining residuum is lighter in nature and less viscous, thus the name of the process "visbreaking. " I n order to gain a higher proportionate yield of light products such as gasoline and kerosine, various processes utilize the heavy distillate from the primary atmospheric distillation as charging stock which is then subjected to cracking and reforming. The objective is to rearrange various hydrocarbons to those having more desirable properties for the intended product. The processes which first were developed were named after their inventors, i.e., Dubbs and Cross Thermal Cracking Units. The furnaces in these units operate a t temperatures of 950 F and pressures of 800 psi. The volatile fractions are separated in a fractionating column and the residuum from these processes is a heavy tar material. Such units have largely been replaced by catalytic cracking units in which lower temperatures and pressures can be employed because of the action of catalyst materials. (A catalyst is a material which accelerates a chemical reaction without itself entering into the
design, largely centering on methods for continuously introducing active catalyst and removing spent catalyst to be regenerated. In a process called TCC (Thermofor Catalytic Cracking) the catalyst in bead form is carried by an air lift to the top of the structure where it flows down through a reactor vessel and then to a regenerator kiln to begin the cycle all over again; the TCC process is illustrated by Fig. 2. PreE&ted oil enters the reactor catalyst near the top to flow downward with the catalyst. The cracked products leaving the reactor are separated in a fractionating column. Another process called the Fluid Catalyst Process utiliaes a catalyst in a powdered form as compared to beads in the TCC process. The powdered catalyst is circulated by oil, steam, or air pressure and arranged to feed into the reactor vessel with tlie feed stock to be cracked. Great advances have been made in catalyst chemistry such that materials are available which provide high activity, .long durability, and of most importance, selectivity to produce desirable end products. Through the proper use of catalysts, it has been possible to significs~tlyincrease the yield of products in great demand, such as gasoline and light fuel oil, and, a t the same time, reduce the portion of less-desirable materials such as coke, fuel gas, and heavy fuel oil. Various types of catalytic cracking units are employed; Reforming is another process by which hydrocarbons their names generally denote the method for introducing may be rearranged to ob_tain more desirable properties. and regenerating the catalyst. The first process was Low-boiling naphtha may be fed to a furnace where it is developed by Eugene Houdry and was called a "fixed heated to 1100 F under pressures of 1500 psi. By bed" unit. With a fixed bed unit, preheated oil vapors introducing quench oil a t the proper time, reactions are pass through the bed until coke formation on the catalyst stopped thereby yielding the desired products. The reduces its activity. The oil stream then is switched to reactions which take place break up large molecules, as another case while the bed of catalyst in the spent case is well as joining smaller ones together, yielding a stock of regenerated by blowing air through it to burn off the higher octane number than the original feed stock. coke. Improvements have been made over the original Thermal reforming has largely been replaced by the
I
844
MARINE ENGINEERING
'
PBROLEUM FUELS Table 1
Fuel Test Methods
'I'EBT METHODO m a UED TO CHAIUCTEWE F ~ FOR L HIGH MEDIUM-AND GAS SPEED LOW-SPEED T~BINE DIESEL DIEBEL BOILER Combustion: - cetane number. ........... ASTM D-613 diesel index. .............. IP 21 cetane index. .............ASTM D-976 luminometer.. ............ ASTM Dl740 heat of combustion. ....... ASTM D-240 Conradson carbon residue ASTM D-189 Rambottom carbon residue ASTM D-524 Volatility-Density: distillation.. .............. ASTM D-86 gravity.. ..........,...... ASTM D-287 -
TREATER
RECYCLE
c.,
..
STRIPPER
PARA
2ND
S T B E FEED
1
PRODUCT I
X X
-
X X
X
X
X X X -a @
X X
'
X X
' X ,
X
X X
X
X X
X
X
Fluidity: pour point.. .............. ASTM D-97 X X X viscosity.. ................ASTM D-445 X X X X Safety: flash point.. .............. ASTM IT56 X X X X or D-93 CledhewPurity: water.. ASTM D-95 X X X X ash.. ....................ASTM D-482 X X X X sediment. ................ASTM D-473 X X X water and sediment.. ...... ASTM 1796 X Corrosiveness: ............. co per strip.. ............. ASTM Dl30 X X s ur.. ..................ASTM Dl29 X X X X or 1%6 vanadium. Note 1 below X X X X sodium.. Note 1 below X potassium.. .............. Note 1 below X X calcium.. Note 1 below X Note 1 below lead.. Note 1: Them am no industry-wide standard methods; these are determined by,-eemical or spectrographic analytical procedures.
..................
UE
Fig. 3 Hydrowadring flow rhwt
, catalytic reforming process which produces greater
yields of high octane fractions. I n the catalytic reforming process, light naphtha and hydrogen are pumped through a furnace a t temperatures of about 950 F and pressures of 200-600 psi. The mixture then passes through a series of reactors containing a platinum or alumina catalyst. Hydrocracking, which is illustrated by Fig. 3, is another process for hydrocarbon conversion. I n this process, a mixture of hydrocarbon feed and recycle gas a t 500-800 I? and 500-3000 psi is contacted with a catalyst in reactors. Hydrogen-rich gas and unconverted hydrocarbons are recycled to complete the desired reactions. This process may be used to convert a wide variety of feed stocks to products of lower molecular weight, e.g., gasoline and light fuels. Coking is a thermal process in which heavy residuum is converted to nearly pure carbon. While the other processes described in the foregoing are operated to minimize coke formation, this process is designed to manufacture coke a t a high temperature (900 F) but low pressure: The coke is formed in drums and light volatile fractions flow to a fractionating column for separation. The coke is removed from the drums as a solid. Polymerization processes are designed to join light hydrocarbons to form heavier ones. This is opposite to the cracking processes. The light hydrocarbons in the gas from catalytic cracking are joined in the presence of specific catalysts to produce high-octane-number gasoline
blending stocks. Alkylation is a similar process in which slightly heavier hydrocarbons are combined, in the presence of a catalyst, to form high-octane-number gasoline blending stocks. Many other miscellaneous processes are used in lightproducts refining. One other process of particular interest is catalytic hydrodesulfurization for treating kerosines, gasolines, and diesel fuels. I n this process, the oil to be desulfurized plus hydrogen is heated to 600-700 F under pressuFs of 250-800 psi. The stream flows through a catalyst in a reactor, where the sulfur is converted to hydrogen sulfide. Hydrogen and light ends are stripped from the oil and the effluent hydrogen sulfide is absorbed in a diethanolamine solution. F i l refining ateps include stabilization of gasolines to maintain the required volatility, gas recovery processes, "sweetening" processes to convert foul-smelling sulfur compounds, and caustic washes to control the phenolic content of the finished product. The various stocks and cuts from the processes described are blel'ded to produce finished products having the desired properties for the intended applications. Additives are employed to enhance existing properties and/or impart new ones, e.g., ethyl fluid to increase gasoline octane number, oxidation inhibitors to impart storage stability, and others. 1.3 Fuel Properties. The fuel products obtained by refining crude oil range from the lightest LPG (petroleum gases liquefied by pressure or refrigeration) through aviation and automotive gasolines (90 F to
............... ................. ................ ...................
430 F boiling range), kerosine and jet fuel (300 to 550 F boiling range), household heating oils (400 to 650 F boiling range), diesel fuels (400 to 750 F boiling range), down to residual fuels (600 to 1200 F boiling range). Each of the fuel products must have certain properties to malie it suitable for its intended application. These properties are defined by certain tests of a chemical and physical nature. The American Society for Testing and Materials (ASTM) and British Institute of Petroleum (IP) methods for testing petroleum products are most generally used and recognized over the world. There are other societies which also issue test methods, but to treat all of these would become repetitious. The test methods in common and widespread use are indicated in Table 1. Note that the test methods are grouped in categories broadly denoting the performance property with which they are associated. A brief description is given in the following of each of these test methods together with the significance of the values obtained. a. Gravity ASTM D-1298. By use of a glass hydrometer, the density, specific gravity, or API gravity may be determined. Values are measured on the scale of a hydrometer a t convenient temperatures, with readings of density being reduced to 15 C, and
readings of specific gravity and API gravity to 60 F by means of international standard tables [I].' By means of these tables, values in one system are readily converted to either of the other two. It is, of course, necessary to know the density of a fuel to compute the weight of fuel in a tank of known volume. Since API gravity bears a known relationship to water, i t is useful in estimating the water separation characteristics of a fuel. For example, since water has an API gravity of 10 deg, lighter fuels (greater than 10 deg) will sepa~ateeasily. Heavier fuels (lower than 10 deg) will separate only after heating, which raises the API gravity of the oil and reduces its viscosity. The specific gravity of a fuel is a broad indicator of the fuel composition; therefore, it is used as a parameter in various empirical relations, such as diesel index and heat of combustion. b. Pour Point ASTM- D-97. The pour point test is used to determine the temperature a t which a fuel can be poured under closely prescribed conditions. A fuel sample is placed in a cylindrical jar; after preliminary heating, the sample is cooled a t a specified rate. At Numbera in brackets designate Referencee at end of chapter.
MARINE ENGINEERING Table 2
PETROLEUM FUELS
Approximate Viscosity Equivalents at the Same Temperature
(For conversion of viscosities at the same temperature. Actual equivalents vary sli htly with temperature. Values in this table are average equivalents and ma be use8 for conversion within the uaual temperature range of viscosity memurement o?fYeh without significant error. Values in italics are.extrapolatlorn beyond the normal working range of the particular instrument used in determining viscosity.) Saybolt Sa bolt Redwood Redwood Kinematic Umved &l No. 1 No. 2 Engler (centistokes) (seconds) (seconds) (seconds) (seconds) (degrees)
30 32 34 36 38 40 42 44 46 48
50 55 60 65 70 75 80 85 90
141 150 159 168 177 186 195 204 213 223 232 255 278 301 324 347 370 393 417
19 19 90
U 9l 98 83 94
2 26 28 30 33 35 37 40 42 44
124
132 140 148 156 164 172 180 188 197 205 225 245 266 286 306 327 347 367
-
16 16 17 17 18 19 90
rn
dl 93 $6 97 89 31 33
36 37
4.07 4.32 4.57 4.82 5.08 5.34 5.59 5.84 6.10 6.36 6.63 7.24 7.90 8.55 9.21 9.87 10.5 11.2 11.8
5-deg-F increments, an examination of flow character- equivalent values can be obtained from conversion tables istics is made and the lowest temperature at which for determinations made at the same temperature, as movement of the oil is observed is recorded as the pour indicated in Table 2. The ASTM D-445 method of assessing viecosity point. As a guide, bunker fuels in mme areas of the (kinematic viscosity) involves measuring the time world have pour points of about 80 F. The pour point is important in that the fuel must be required for a fixed volume of fuel a t a given temperature
Table 2 (continued) Kinematic (centistokes) 200 220 240 260 280
Saybolt Universal (seconds) 926 '1018 1111 1204 1296
Saybolt Furol (seconds) 95 104 * 113 123 132
Redwood No. 1 (seconds) 816 898 980 **lo61 1143
Redwood No. 2 (seconds) 88 90 98 106 114
Engler (degrees) 26 29 32 34 37
380 1760 179 1551 155 50 400 1852 188 1633 163 53 420 1944 197 1714 171 55 440 2037 207 1796 180 58 61 460 2129 216 1877 188 480 2222 226 1960 196 63 500 2314 235 2041 204 66 225 72 2245 550 2546 258 . 245 79 600 2777 282 650 3008 306 265 86 700 3240 329 2857 286 92 750 3472 352 3062 306 99 800 3703 376 3266 327 105 850 3935 400 3470 347 112 900 4166 423 3674 367 118 446 3878 388 125 950 4398 408 132 1000 4629 470 4082 517 4490 449 145 1100 5092 564 4898 490 158 1200 5555 611 5307 531 171 1300 6018 658 5715 572 184 1400 6481 705 6123 612 197 1500 6944 752 6531 663 211 1600 7406 790 6939 694 224 7869 1700 846 7348 735 237 1800 8332 1900 893 7756 776 250 8795 2000 9258 940 8164 816 883 For viscosities greater than 1000 sec the Furol instrument is usually em loyed. ** For viscositiesgreater than 1000aec the Redwood NO.2instrumentis uaudy employed.
23
for the most part, require a maximum viscosity of 90 SUS at the nozzle for proper spray. Figure 4 shows the viscosity-temperature relationships for typical bunker fuels as applied to boiler usage [3]. I t should be noted that fuel oil preheaters should be placed on the discharge side of supply pumps so that the fuel is heated under pressure, thereby avoiding the f.lashiig of entrained water or light components with a consequent vapor lock. d. Distillation ASTM D-86. A distillation test is used to determine the volatility characteristics, or boiling range, of a fuel. A measured sample is heated in a flaskand the vapors are condensed and jreclaimed in a graduated cylinder. The temperature of initial boiling, incremental percents of liquid recovery, and final boiling points are recorded. Measurements are also made of the nonvaporiaed residue and losses. The temperature curve depicts the boiling points of the multitude of hydrocarbons of which the fuel is composed. While the test can be applied to all fuels, it is primarily used to characterize the lighter products (i.e., jet fuel, light diesel oil, and light fuel oils). The initial boiling point and the 10-percent recovery
point are commonly associated with the cold-starting characteristics of high-speed diesel engines and the ignition properties of jet engine fuels and light boiler fuels. The h a 1 boiling point and the last 10-percent recovery point are related to smoking in diesels and boilers. As mentioned later, the 5Gpercent recovery temperature is used as a parameter in calculating the cetane number [2,3]. Many other relations in petroleum manufacture and blendin4 utilize distillation data. SufEce it to say that engine and boiler manufacturers state volatility require ments for their equipment on the basis of this simple yet extremely meaningful test. . e. Flash Point ASTM D-56 and D-93. Two methods are commonly employed to deterrnipe the flash point of fuels; one is the D-56 tagclosed flash-point test for fuels with a flash point less than 175 F, and the other is the D-93 Pensky-Martens closed cup test. In both tests the fuel sample is heated in a closed cup and a test flame is periodically passed through the vapors above the liquid level. The lowest temperature a t which the vapors ignite is the flash point. The value determined is indicative of the maximum temperature a t which the fuel
MARINE ENGINEERING
PErROLEUM FUELS Table 3 Specific Gravity and Heating Values of Fuel Oils HIGHER HEATING VALUE LOWER HEATING VALUE
GEAVXTY AT 60/60 F
UPPER RANGE OF VISCOSITY FOR BUNKER FUELS
Deg API
>
70
Specific Gravitv
(at conatant volume) Btu per Btu Gal st ~ e Lb r 60F
-
(at comtmit preeaure) Btu per Btu Gal at ~ e Lb r 60 F
-
60 55
--
50
-
45
-
40
-
NOTE: CURVES SHOW AVERAGE VALUES ONLY
10
30
M
70
90
110nLL130
150
170
190
210
230
250
270
290
TEMPERATURE, DEGREES FAHRENHEIT Fig. 4 Viieasitptempcrature characbrtrtio for a range of typical bunker fuel oils
can be stored or handled without the risk of explosive vapors. A 150 F minimum flash point is the generally accepted minimum value for fuels used aboard ship. While this minimum is specified by the various European classification societies, the U.S. Coast Guard has revised coastal shipboard requirements to permit fuels having a minimum flash point of 140 F. f. Luminometer Number ASTM D-1740. A luminometer test is used to measure the flame radiation characteristicsof a fuel and, therefore, is indicative of the combustion quality of jet or distillate products used in gas turbines. Fuel flame radiation characteristics are determined by burning the fuel in a particular lamp and measuring the flame radiation through optical and photocell equipment. The burning process in gas turbine combustors is primarily determined by various factors of design. However, the lufinometer number is an indication of the amount of energy in the form of thermal radiation and is, therefore, a specification requirement usually stipulated
by manufacturers of aircraft-type gas turbines to avoid engine overheating and combustor daniage. g. Heat of Combustion ASTM D-240. A heat of combustion test is conducted to determine the heat generated during the combustion of a fuel under rigidly controlled conditions. A weighed sample of fuel is '8iurned in an oxygen bomb calorimeter and the heat of combustion is computed from temperature observa%ions before, during, and after combustian. Corrections are made for thermochemical and heat-transfer effects. The results. are reported either in calories per gram or Btu (British thermal units) per pound. The total heat of combustion (also called gross, or higher, heating value) includes the latent heat of the water vapor formed during the combustion of hydrogen. The net heat of combustion (also known as the lower heating value) is the total heat of combustion minus the latent heat of the water vapor formed during combustion. Use of the latter value is sometimes preferred (in Europe) because it is reasoned that the latent heat of the water formed during combustion is unrecoverable in practical applications.
For example, a diesel engine's exhaust is a t temperatures well above the boiling point of water. While the use of the net heating value can be justified, thermal efficiencies in the United States are customarily based on the higher heating value for two reasons: (1) fuel higher heating values may .be determined by means of a calorimeter to an accuracy of 0.5 percent, whereas net heating values must be calculated fram the gross value and the amount of hydrogen in the fuel oil, and (2) it has become customary in the petroleum industry to quote higher heating values. Higher and lower heating values of fuels ranging in gravity from 5 deg to 49 deg API are approximately as shown in Table 3 [4]. For improved accuracy the heating values shown in Table 3 should be reduced by a percentage equal to the sum of the percentage of sulfur, water, and sediment actually in the fuel in question.
h. Cetane Number ASTM D-613. The cetane number of a diesel fuel is a measure of its ignition quality. The test is performed in a specially constructed singlecylinder,diesel engine arranged so that the compression ratio can be altered while it is in operation. In essence, the test consists of varying the compression ratio for the sample fuel, and blends of known reference fuels, to obtain the same ignition delay period, i.e., the time interval between injectim and ignition. The reference fuels are normal cetane, which has an arbitrary value of 100, and alpha-methylnaphthalene which has a value of zero. The blend matching the sample fuel is reported as the cetane number of the sample fuel. Higher cetane numbers denote shorter ignition delay periods. This has particular significance in high-speed diesel engines where shorter ignition delay periods are necessary to complete the combustion cycle. Slow-speed,
MARINE ENGINEERING
large-bore engines, on the other hand, can utilize fuels of longer ignition delay characteristics (i.e., lower cetane number) because of the greater combustion time available. The cetane number required for a fuel is a design
placed in a furnace that is maintained a t 1020 F. After a specified period, the bulb containing the residue is cooled and weighed. The results are reported in the same manner as above; however, the values obtained
Since the special engine for this test is expensive and the test requires a skilled operator, other means (as described in the following) are commonly employed to estimate the cetane number. i. Diesel Index IP-21. The diesel index is a method %hatis used to approximate the cetane number through the determination of two properties which have a correlation with the cetane number. These two properties are the aniline point and the API gravity. The aniline point (ASTM D-1012) is the temperature in degrees F a t which specified quantities of aniline and the test sample separate after first being heated and put into the solution. The temperature of solution is indicative of the paraffinic composition of the fuel which in turn is related to the cetane number. Generally speaking, the more paraffinic the fuel, the higher thecetanenumber. The aniline point and the API gravity are used to classify the composition of a fuel by means of the diesel index. The following equation is used to calculate the
given in the test descriptions in the ASTM manual. m. Ash Content ASTM D-482. The ash content test is used to obtain an assessment of the incombustibles in a fuel. The residue or ash remaining after test consists of rust, noncombustible dirt, and also small amounts of constituents such as vanadium, calcium, sodium, and potassium. Excessive ash constituents in a fuel can result in the clogging of fuel nozzles or injectors, cylinder and piston ring wear in diesel engines, blade abrasion or corrosion in gas turbines, slagging in boilers, or corrosion of tubes in superheaters. The fuel should be treated by either centrifuging, filtration, or both to minimize the amount of ash in the fuel before it is burned in diesel engines or gas turbines; the fuel ash content is generally not as critical with boilers of proper design. Vanadium and sodium form corrosive compounds a t high temperatures (i.e., approximately 1250 F) which can rapidly attack gas turbine blades, diesel engine valve faces and piston heads, and high-temperature areas in boilers. Vanadium exists in soluble form in petroleum. Sodium generally is the result of saltwater contamination either during delivery or by residual salt water in bunker tanks. Unfortunately, certain combinations of sodium and vanadium form eutectic compounds which have relatively low melting points; consequently, in cases where eutectic compounds are formed, corrosion occurs a t lowered temperatures. The mechanism for this corrosive attack has been discussed extensively in the literature [5, 61. Many crude oils contain significant amounts of vanadium; therefore, the international ship operator cannot avoid obtaining residual fuels with a high vanadium content a t some ports in the world. Generally, corrosion is controlled by the diesel engine designer through metallurgy, exhaust valve cooling, exhaust valve rotation, and piston cooling. In boilers, ash corrosion is also controlled by design, metallurgy, and temperature control. Some fuel-oil additives have been reported to offer a degre'e of protection against ash corrosion. The control of vanadium/sodium corrosion in gas turbines is critical because of the high turbine bladg temperatures. Aircraft-type turbines require very low concentrations of such materials as indicated in Table 4. Industrial-type turbines can utilize fuels with somewhat higher vanadium/sodium content by operating a t lower blade temperatures but a t a penalty in the thermal efficiency of the cycle. n. Water ASTM D-95. Water contamination in storage, transportation, or ship tanks is determined by a distillation test utilizing a water-immiscible, volatile solvent. The fuel is heated under a reflux condenser , with a water-immiscible solvent which co-distills with
diesel index
=
(aniline point OF)(OAPI gravity) 100
j. Cetane Index ASTM D-976- The citane index is another method for estimating the cetane number of diesel fuels based on physical properties- It makes use of the API gravity and the mid-boiling-point temperature for 50 percent recovered in the ASTM Distillation D-86. The equation includes a number of complex factors which can be reduced to the alignment chart illustrated by Fig. 5. k. Carbon Residue ASTM D-189. The carbon residue test is used to determine the amount of carbon residue left after evaporation and pyrolysis of a fuel oil. A weighed sample is placed in a crucible and subjected to severe heating during a fixed period. After cooling, the crucible is reweighed and the remaining residue is calculated as a percentage of the original sample. For light fuels and diesel fuels, the carbon residue is run on the 10 percent heavy-end fraction obtained from a distillation of the sample; the results are expressed in weight percent, specifying whether the test was made on ti whole sample or a 10 percent distillation fraction. The carbon residue is an indication of the tendency of a fuel to form deposits in 'burners or in diesel engine combustion chambers. 1. Ramsbottom Carbon Residue ASTM D-524. A Ramsbottom carbon residue test may also be used to determine the amount of carbon residue left after evaporation and pyrolysis of a fuel oil. A weighed sample, contained in a special open-end glass bulb, is
85 1
I'ErRVLtUM tUtLS
- OC
BASED ON E0UAtK)MS :
CALC. C.I. = 0.49083 + 1.06577 (XI -0.001os~t ( x )* X 97.833 [LOG M1O.B P T . ' , ~ F ) ~ + P ~ O ( W IAPI) lLOGMIPBPl.,DFl +0.01247 UPI)~ 423.51 (LOG MID. B PT.,'FI -4.7800 [API) +419.59
-
MID*F a. PT.
TEMP.RANGE 374-410 410-446 44s 482 48z-sle 51 8- 554 554-590 5 9 0 - 626 626-662 862-698
CORRECTION@ EACH lo MU DIFF. FDR IN PRESSURE
-
1.19 1.24 I .28 1.32 1.37
a
TEMPERATURE FOR 50% RECOVERED AT 760 M Y ~ p B A R O Y ETRIC PRESSURE.
@
TO BE ADDED FOR PRESSURES BELOW
760
MY H g OR
SUBTRACTED FOR
PRESSURES ABOVE 760 Y Y
ng
EXAMPLE: MID API B.PT. GRAVITY SSO*F 33.0AT 700 MM
HQ
CORRECTION FOR MID B.PT=I.IS x e rn 7.14 CORRECTED MID B PT.550DF 7.14.F 557.144. CALC. CETANE INDEX (NOMOGRAPH) m51.4 CALC.CETANE INDEX (FORMULA)=II.~!I
+
I
5 Calculated Wan0 index (ASTM)
OF
852
MARINE ENGINEERING
rnpe!efi
4
'(W) pse?
'BPI
rnpXq mdd '(q mnpp3 )
PETROLEUM FUELS
,,
In
In
+
I*Xq mad '(X *N) mnrssslod mid mn!Pos I@eM Xq mdd '(A) arn!+wg IdV %P 'XI!At=D
xsgq
El
El
B
0,
!
h
=w i
OD
3
'BYU:
4
09 u)
am
2
9 El
'BW
:
h
09
3
s
h
d
h
a
3
=W i
h
U!rI
:
'9
El
b
% I
the water in the sample. Condensed solvent and water are continuously separated in a trap, the water settling in the graduated section of the trap and the solvent returning to the still. The volume of water is read directly from the graduations etched on the trap. , Water should be removed from the fuel before being supplied to either diesel or gas turbine engines. This applies particularly to salt water which could cause injector clogging and corrosion in diesels and blade deposits and corrosion in gas turbines. Steam boilers are more tolerant but excessive saltwater contamination in the fuel results in sodium slag in the boiler. The water can be removed by proper centrifuging or filtration procedures. o. Sediment ASTM D-473. The fuel can contain sediment, which is insoluble suspended matter and tarry polymerized hydrocarbons which are borderline in solubility. The amount of such material is determined by employing a solvent which selectively dissolves the unchanged hydrocarbons leaving the sediment in a device for measurement. In this test the fuel sample is placed in a refractory thimble which is suspended in a flask. Toluene in the flask is heated and condensed so that it drips through the thimble until all the oil has been extracted. The sediment remaining in the thimble is weighed and the result reported as a weight percent of the original sample. Diesel injectors and gas turbine nozzles havk little tolerance for sediment in a fuel. Moreover, polymerized materials break down under heat, forming lacquer and tarry deposits which impair movement of parts or interfere with spray patterns. Centrifuging or the use of filters is necessary to remove such materials. p. Water and Sediment ASTM D-17%. This test measures both water content and sediment and is often used as a substitute procedure for the two tests described in the foregoing. Two equal samples of the fuel are placed in special centrifuge tubes which are inserted in a laboratory centrifuge. The centrifuge is then operated under specified conditions of temperature, centrifugal force, and time. A solvent is also employed to facilitate the separation of water and sediment from the fuel. The tube is designed with calibrations so that the amount of
853
water and sediment can be read directly a t the conclusion of the centrifuge test. q. Copper Strip Corrosion ASTM D-130. A copper strip corrosion test entails pouring a fuel sample in a tube in which a polished copper strip is immersed. The test tube assembly is placed in the water bath a t 212 F for three hours. At the end of the test period, the copper strip is compared with stafidard strips to evaluate the amount of corrosive attack. The significa~ceof the data is, of course, related to the corrosive rqterials in the fuel which may attack metal parts in fuel storage, handling, or combustion systems [6]. r. Sulfur Content ASTM D-129. To determine the sulfur content of a fuel, a sample of the fuel is placed in a test bomb pressurized with oxygen. After ignition, the interior of the bomb and sample cup, etc., are rinsed with distilled water and the drainings are collected in a beaker. After the addition of bromine water and hydrochloric acid, the contents are evaporated to a given residue. During this evaporation, barium chloride is added to convert the sulfuric acid to barium sulphate. The solution is then passed through a special filter paper which traps the barium sulphate. Then the filter paper is placed in a crucible, ignited, and weighed to determine the weight of barium sulphate remaining. Through calculation, the sulfur content of the original fuel i s finally determined and expressed as a percentage by weieht. since sulfur forms corrosive oxides and acids, it can cause accelerated cylinder wear in diesel engines, corrosion of boiler tubes and surfaces, as well as attack gas turbine blades and interiors. I n boilers, it is essential to maintain the stack gas temperature above the dew point (approximately 250 F) to prevent the formation of sulfuric acid and the attendant violent corrosive attack on metals. Control of corrosive wear in diesel cylinders is obtained by the use of lubricating oils containing alkaline additive materials which neutralize the acids formed [2, 61. The sulfur content of fuels is subject to regulation insofar as land use is concerned, since it plays a part in atmospheric pollution. However, international ships have been exempted from landside regulations, in view of the relatively small pollution effect they have.
Section 2 ~ u e Procurement l 2.1 Fuel Requirements. It should be evident that a degree of judgment must be exercised in relating the significance of the fuel tests previously described to engine or boiler performance. For example, the same, amount of ash-forming material in a fuel may be of no consequence in a boiler, of little significance in a lowspeed large-bore diesel, injurious in a high-speed diesel, and disastrous in some types of gas turbines. Conse-
quently, the ash content test is only a means of assessing the quantity of ash-forming material that is present in a fuel as determined by the rigorously defined procedure of the test method. The other fuel tests, by and large, provide empirical data of similar usefulness. I n order to define the kind of fuel an engine or boiler can use, it is necessary to relate these fuel properties to the design and performance requirements of the power
.
11
854
MARINE ENGINEERING Table 5 Limiting Requirements for Diesel Fuel Oilsa (ASTM) Pour Water Csrbon Flaah Point and hidue Point, deg' Bediment, on 10 Anh, deg F percent percent percent F (deg by Residuum, by (deg C) C) volume percent weight
Grade of Diesel Fuel Oil
No.
A volatile distillate fuel oil for engine8 in service requuling frequent opeed m d Iondl Lchanges. el oil of lower engines in inheavy mobile
i
A fuel oil for low- and meNo' 4-D (dium-epeed engines.
)
Min 100 o r legal (37.8) 125 or legel (51.7)
Max b
+,.F2",";fiBB, ~ i s ~ ~at~100 i tFy d e g y ( d e g C) (37'8 I' sulfur, 90 percent point
copper ~ e t m Kinematic, centi- percent Stnp Numstokes (or Saybolt by Cqrro- bed Universal, eec) weight slon
Max Tr808
Max 0.15
Max 0.01
Min
...
Max 550 (287.8)
Min 1.4
Max 2.5 (34.4)
Max 0.50
Max No.3
Min
0.10
0.35
0.02
640c (282.2)
(338)
640
2 (32.6)
4.3 (40.1)
1.0
No.3
40.
408
'1".",r?
(54.4)
a To meet special operating conditione modifications of individual limiting requirements may be agreed upon between purchaeer seller and sup lier. b For cold-weather o eration, the pou;point should be specified 10 deg F (5.6 deg C) below the ambient temperature a t which {he en&ne is t o i e operated
except where fuel oil heating h i t i e n are provided. F e n pour point less than 0 F (-17.8 C) is specified, the minimum visconity ahall be 1.8 cs (32.0 sec, Saybolt Universal) and the minimum 90 percent point shall be waived. d Where cetane number by Method D 613 Test for Ignition Quality of Diesel Fuels by the Cetane Method, is not available, Calculated Cetane Index may be used BB an approximation. Where there is disagre&nent, Method D 613 shall be the referee method. * Low-atmoepheric temperaturea se well as engine opemtion a t high altitude8 may require use of fuels with higher cetane ratings.
plant. The power-plant designer should design the plant such that the fuel burned represents the best compromise between fuel costs, initial costs, operating costs, and maintenance costs. The more expensive fuels are the light distillates (i.e., similar to kerosine and home heating oil), while the cheapest is heavy residual fuel. The latter is jokingly described as "liquid road paving." Obviously, a power plant operating on residual fuel would have a marked advantage over one requiring a light distillate fuel, provided all other costs were the same. Power-plant designers have established fuel requirements which may be very detailed in some cases and broad in others. Furthermore, requirements may be different for the same power plant in one type of service versus another. For the most part, power plants are designed to employ the fuels available in the intended areas of application. For example, high-speed diesels in inland rivers or coastal fishing service can make use of landside automotive-quality diesel fuel. On the other hand, ships in international service will not find such fuels available a t most ports. Even if potentially available a t some ports, the low-volume storage and delivery expenses plus customs duties for nonbonded fuel would make the cost to the ship operator prohibitively high. 2.2 Fuel Speciflcations. Considering the large number of power-plant builders in the world, the multitude of types (and models) produced, and the fact that fuel requirements may differ according to service conditions, it is quite understandable that there are no universally accepted fuel specifications. However, to provide guidance to power-plant builders and users, the American Society for Testing and Materials has published classifications for diesel, heating/industrial, and marine gas turbine fuels [I]. The ASTM D-975 classification of diesel fuels, Table 5, lists three classes of diesel fuels ranging from a light volatile distillate or kerosinetype fuel (No. 1-D), through a heavier distillate (No. %D), to an even heavier distillate or distillate-residual blend (No. 4-D). The test limits are intended to portray the
1
maximum (or minimum) acceptable value for each characteristic. It would be highly unlikely that any commercial fuel would possess the maximum (or minimum) limits for all characteristics. The classification limits, therefore, are extremes and should be considered as the minimum-quality fuel in each class. The ASTM D-396 classification of heating or fuel oils, Table 6, lists six classes of fuel oils ranging from kerosine to heavy residual. Note that No. 1, 2, and 4 are quite close in their characteristics to No. 1-D, 2-D, and 4-D diesel fuels. The ASTM tentative classification for gas turbine fuels (non-aircraft) lists four classes of fuels as shown in Table 4. Note that a specific limit is imposed on the content of vanadium, calcium, lead, and combined sodium+andpotassium since these are related to blade corrosion in some engines operating a t high inlet gas temperatures [5, 61. Other societies, associations, and government agencies have published standards and specificationswhich largely apply to inland services in the countries where they were originated. However, bonded supplies of ships' bunker fuels are free of most "local" regulations. Large volume buyers, such as the U.S. Navy, have established their own specifications in order to obtain the optimum fuels meeting their particular operating and storage requirements. I n most cases, deliveries tqe made ex refinery into a fleet oiler, or in commercial tanker loads, to fleet storage. Obviously, particular specification requirements can be met under such conditions of purchase [7-111. 2.3 Fuel Distribution System. Over the years a fuel distribution system has developed for ships in international trade which essentially provides three basic fuels plus blends of two of the three to comply with any intermediate viscosity level desired. The fuel distribution system has evolved as a natural solution to the problem of supplying adequatequality fuel a t minimum cost a t all the ports of the world where ships
I
!
856
PETROLEUM FUELS
MARINE ENGINEERING
require bunkers. The cost of fuel is a major ship operating expense; therefore, a major emphasis is placed on economy. Large-volume bunkerings are also involved; the average ship takes on about 3300 barrels (42 gal. each) of fuel a t a time. Bunkering, therefore, is performed a t refineries or bulk s t o ~ g terminals, e or by fuel barges shuttling between vessels in port and storage terminals. By these means, transportation, storage, and delivery costs are minimized to the benefit of the ship operator [4]. Since each type of fuel requires separate transportation, storage, and delivery facilities, it is obvious that the fewer types of fuel handled by the fuel suppliers, the greater the economieswhich can be achieved to the benefit of the ship owner. As mentioned in the foregoing, the general practice of major fuel suppliers is to offer three basic grades of fuel oils, and a range of different viscosity blends produced by blending two of these basic grades in the required proportions. The three basic fuels are: a. Gas Oil, In general, gas oil is a light distillate with a 45 or higher cetane number, 150 F flash point minimum, contains no residual (it is, therefore, very low in carbon residue), and is virtually free of ash-forming s ASTM classifications materials. It a ~ ~ r o x i m a t ethe for No. 2-D (~iesel),No. 2 (Fuel), and No. ZGT (Gas Turbine). b. Marine Diesel. I n most areas west of Suez the product known as marine diesel is a 100-percent distillate of higher boiling range than gas oil, cetane number of 40 or higher, and 150 F flash point minimum. It has a low carbon residue and very little ash-forming material. While this product leaves the refinery as a distillate, it is carried in "dirty bottoms" to achieve minimum transportation and delivery costs. "Dirty bottoms" is a term applying to tankers or barges which carry residual fuels as well as distillates. Since their tanks may contain some remnants of the previous delivery (below pump suction), residual fuel contamination results and the distillate becomes dark or black in color. The fuel shows some carbon residue and ash content as a result of this "handing" contamination. This fuel approximates the ASTM classifications for No. 2-D (Diesel), No. 2 (Fuel), and No. 2-GT (Gas Turbine). I n the Middle and Far East the product known as marine diesel fuel generally contains 5 to 10 percent residual fuel blended a t the refinery. This practice resulted from past regulations in a number of countries requiring industrial and marine fuel to be black in color to differentiate it from automotive fuels for tax and customs purposes. The presence of the residual component results in higher carbon residue and ash values than found in marine diesel fuels originating west of the Suez. Other properties are not affected to any appreciable extent. Marine diesel fuels originating in the Middle and Far East approximate the ASTM classifications for No. 4-D (Diesel), No. 4 (Fuel), and No. 3-GT (Gas Turbine). c. Residual Fuel. The third basic fuel is commonly known as residual fuel (or Bunker C) and is blended
3
Table 7 Common Names for Petroleum Fuels MOSTCOMMON NAMF, OTHERN ~ E B Gas Oil Light Diesel Oil Light M+ne Di-1 Fuel Marine I h w l Light Marine Diesel Medium Marine Diesel Medium Fuel Oil Marine Diesel Heavy Di&l Oil Marine Diesel Fuel Oil Marine Diesel Heavy Marine Diesel O i l Blended Fuel Heavy Fuel Oil High Viscosity Fuel Oil (followed by a number, Intermediate Bunker Fuel Oil e.g., 400, 1000, 2000, Intermediate Fuel MD 400,600, etc. etc. inhoating Light Fuel Oil 200, 400, etc. +dwPod No. 1 Thin Fuel v~scoslt of blend 0100 Thin Fuel Oil Residual Fuel Bunlccir Fuel Bunker . - Fuel . .Oil - . Bunker C Fuel Bunker C, P8400 Bunker C. Boiler Fuel Boiler Fuel Oil Heavy Fuel Oil Heavy Fuel, PS-400 Heavy No. 2 Fuel Oil Marine Fuel Oil No. 5 Fuel 011 Number 6 Number 6 Fuel Oil
d
- -
from residuum left from the various processes in a refinery and distillate to a viscosity suitable for pumping and handing aboard ship. This product is a black, viscous material, with a gravity in some cases heavier than water. Carbon residue and ash content are considerably greater than for the other fuels described in the foregoing. This fuel approximates the ASTM classifications for No. 6 (fuel) and No. 4 G T (Gas Turbine). A series of fuels may be made by blending the second and third basic fuels (marine diesel and residual fuel), and these blends are available a t many world ports to accommodate the large low-speed diesel ship. The operator specifies the viscosity desired and in-line blenders mix the two basic fuels during the bunkering operation. Viscosities of the blends range from 200 seconds to 3500 seconds Redwood No. 1 a t 100 F. At bunkering ports which do not have in-line blenders,. batch blends are made in tanks ordelivery barges, or several viscosity blends are carried in storage. The supplier carefully selects components in making these blends to avoid incompatible mixtures. The price of the blended fuel varies according to the amount of the more costly marine diesel fuel used in the blend. Various considerations, such as fuel heating, pumping, fuel treatment, injection equipment, and engine design, have an influence in determining the most economical blend for a ship. Ships driven by large low-speed diesels generally use a fuel of 1000-1500 seconds Redwood No. 1 viscosity a t 100 F. The characteristics of a blend will reflect for the most
part those of each component but not necessarily in direct relation to the blend proportions. For example, the sulfur content will be a linear proportion while the viscosity is nonlinear. Except for broad approximations, blend curves for each property are used to determine-the desired ratios. This type of information can be readily obtained from petroleum suppliers. Just as there are no universally accepted specifications for marine fuels, marine fuels are known by a variety of names throughout the world. This is an understandable result of language and usage, but often causes misunderstanding; not to mention deliveries of the wrong product. Competition between petroleum suppliers adds to the problem, since each company promotes "brand names" for its products as part of its sales program. Table 7 is a listing of the more common names used for marine fuels; it is interesting to note the diversity of names which the same fuel product can have [4]. References
2 Burning a Wide Range of Fuels in Diesel EnginesVol. 11, Society of Automotive Engineers, Progress in Technology Series, 1967. 3 "Diesel Fuel Oils, " Special Technical Publication No. 413, ASTM, 1967. 4 Technical Bulletins supplied by major oil companies; e.g., Esso, Gqf, Mobil, Shell. 5 K. A. Troth, "SymIfosium on Diesel Fuel Oil," Tram. ASTM, 1966. 6 "Hot Cgrrosion Problems Amckted with Gas Turbines," 8pecia1 Technical Publication No. 421, ASTM, 1967. 7 "Fuel Oil, Burner," Military Specification M I L F-859, (NATO Symbol F-77). 8 "Fuel Oil, Diesel, Marine," Military Specification MILF-16884, (NATO Symbol F-75). 9 "Fuel Oil, Burner," Federal Specification W-F815. 10 Tuel, Navy Distillate, " Military Specification MIGF-24397. - --- 11 "Fuel, Reference, and Standard Distillate, " Military Specification MILF-24376. -
1 Petroleum Products-Fuels, Solvents, Lubricating Oils, Cutting Oils, Gease, ASTM Standards, part 17.
INDEX
Acidic solutions, 559 Active fin stabilizers automatic controls, 602 design calculations, 600 fin tilting gear, 602 location, 601 types of units, 601 Air conditioning defined, 710 objectives, 710, 711 Air conditioning systems air handling systems; see Air handling systems controls, 725 cooling coil load, 727 cooling coils, 763 design criteria, 726 factory assembled units, 765 internal cooling loads, 726 load calculations, 729 overall considerations, 23 refrigeration equipment; see Refrigeration equipment types, 718-721 Air ejectors; see Ejectors Air handling systems air distribution, 758 air flow, 745 dampers, 760 duct design criteria, 746 duct design methods, 746 duct elbow losses, 749-751 duct fitting losses, 749
Anchor windlasses (continued) powering calculations, 587 requirements, 578 special design considerations, 581 types, 579 Assurance engineering, 5 Attemperators, 111 Automation applications, 796,809 boilers, 94, 797, 800 cargooil systems, 808 console, 621 control center, 621, 795, 802, 807 control consoles, 804, 807 control system characteristics, 800 defined, 791 design considerations, 24 diesel propulsion plants, 276, 307, 804 gas turbine propulsion plants, 808 information systems, 795, 802, 807, 808 logic diagram, 798 manning levels, 799 mimic diagrams, 796, 808 regulatory body requirements, 797 safety, 797 steam propulsion plant, 797 systems, 791 Auxiliary turbines, 59, 201, 203 efficiency,59 steam rate, 59, 201
'
fan pressure, 745-747 fan pressure calculations, 755, 756 refrigerated space air distribution, 763 system pressure calculations, 755,756 a
ventilation terminals, 754, 759, 760 Air heaters, 85,108 Anchor windlasses brake design, 582 brake design calculations, 583 brake test data, 583-585 overall considerations, 23 power units, 586
Boilers (continued) attemperators, 111 automation, 94, 797, 800 bent-tube, 79 burner clearances, 100 burner selection, 99 burners, 116,797 care, 127 casing design, 115 circulation, 93, 112 cleaning, 127 combustion air, 97 combustion systems, 21 construction, 114 controls, 120 definitions, 86 design criteria, 91, 92, 94 design problem, 120-125 desuperheaters, 111 drums, 114 duty cycle, 93 economizers, 22, 85, 108, 109, 125 efficiency, 98, 100 example calculations, 120-125 feed pumps, 22, 28, 403,417, 688 feedwater test sample cooler, 491, 517 feedwater treatment, 126 firing methods, 90 flue, 78 forced circulation, 82 fuel analysis, 97 fuel combustion, 92, 95 fuel-oil piping systems, 694 fuels, 90 furnace calculations, 121-123 furnace design, 43, 101 furnace design limitations, 103 furnace exit gas temperature, 43, 101, 122-123 furnace heat absorption rates, 102 heat recovery equipment, 48 heat transfer in furnaces, 43 heat transfer in tube banks, 46 heating surface calculations, 123 high-temperature corrosion, 107 history of, 78
Basic solutions, 559 Batteries, storage, 634, 835 Bearings diesel engine, 292 friction coefficient, 772 hydrostatic, 781 journal, 197, 239, 325, 379 characteristics, 778 externally pressurized, 781 nonpressurized, 780 pressure fed, 778-780 machining-type failures, 774 power loss, 772 pressure gradients, 771 rolling contact, 239, 779, 783-785 thrust; see Thrust bearings types, 770 viscous effects, 771 Boilers air heaters, 85, 108 arrangement, 20, 671
%
I
I
instrumentation, layout, 121 120 main feed pumps, 22,28,403,417,688 main feed system, 688 materials of construction, 824 mountings, 117 number of, 20 oil burners, 116 once-through, 84 operation, 127 physical requirements, 114 piping, 682-690 pressure drops, 93 radiant heat absorbing surface, 102
Boilers (continued) regulations, 92 reheat, 81 reheaters, 108 safety valves, 118 Scotch, 78 sectional header, 78, 104 d selection of, 90 slagging, 107 smoke indicflor, 119 space requirements, 91 steam conditions, 19, 90 steam drum b d e s , 112, 114 storage, 128 supercharged, 84 superheaters, 105-108 terms, 86 tube banks, 104 tube metal temperatures, 102, 106 two-drum, 80 waste heat, 84 water level indicators, 118 water treatment, 125-127 weight, 91 Boiling-water nuclear reactors, 155 Bow thrusters design details, 597 effect of submergence, 598 performance characteristics, 598 types, 597 Bracket fans, 639 Brakes, electric, 632 Brayton cycle, 206 Brine coolers, 767 Brine cooling systems, 745 Brines, 715 Bushnell, David, 3 Cable ship Long Lines, 348 Cables, see Wiring Campbell diagram, 186, 233 Canberra, passenger ship, 355 Capstans arrangements, 594 electric details, 595 mechgnical boom rigging details,namenclature, 594 588 Cargo Cargo handling systems, 23 Cargo pumps, 420,705 Cargo-oil systems automation, 808 piping systems, 705-709 Cargo winches electrical details, 590 hydraulic detaila, 592 mechanical details, 589 ratings, 589 tmes, 589
860
INDEX
MARINE ENGINEERING
Cathodic protection, 533,645,646,811-813,816,817,835 Centrifugal pumps; see Pumps, centrifugal Chilled-water system, 740 Chyssi, tanker, 371 CZermont, steamhip, 2, 78 Clutches, 327 Combined main propulsion cycles, 11 CODAG (combined diesel and gas), 211 CODOG (combined diesel or gas), 211 COGAG (combined gas and gas), 211 COCAS (combined gas and steam), 75,173,211 COGOG (combined gas or gas), 211 COSAG (combined steam and gas), 173, 211 STAG (combined steam and gas), 173, 211 Combustion constants, 96 Compressors centrifugal, 444 fundamentals, 441 liquid sealing ring, 434, 561 performance, 441 reciprocating, 443 rotary, 443 types, 440 Condensers air cooler design, 460 air inleakage, 455, 460 air removal, 455, 460 arrangements, 450 auxiliary, 450 bypass systems, 464 condensate deaeration, 462 condensate purity control, 470 condensate reheating, 462 condensate removal, 455 condensate returns, 463 condensate storage, 456 condensing pressure, 454 condensing space, 456 condensing surface arrangement, 457 cooling water passes, 465 cooling water system pressure losses, 467 corrosion, 471 design conditions, 455 design criteria, 476 design examples, 476478 divided circulation;466 establishment of size, 476 feedwater makeup, 464 heat transfer condensing, 478 conduction, 479 convection, 479 overall, 480 heat-transfer coefficient, 475, 480, 486 heat-transfer equations, 475 logarithmic mean temperature difference, 474, 475 main, 22, 450 materials of construction, 471, 829 nomenclature, 473
Condensers (continued) performance characteristics, 451 performance determination, 477 performance evaluation, 486 performance prediction, 478,486 performance standards, 476 piping, 675 piping systems, 690 pressure around tube banks, 482 pressure losses, 467, 480 pressure losses in tube banks, 485 shell expansion, 469 shell structure, 453, 470 steam distribution, 458, 464 steam dump, 464 steam flow distribution, 458 steam space pressure losses, 480 steam velocities, 456 supporting arrangements, 453 surface, 450 temperature distribution, 452 terms, 473 tube drag coefficient,459 tube expansion, 469 tube sheets, 467 tube spacing, 467 tube support plates, 468 tube-to-tube-sheet joints, 468 tube vibration, 458 tube water velocities, 473 waterboxes, 466 Constant-tension mooring winches arrangements, 595, 596 electrical details, 596 mechanical details, 596 Contaminated drain inspection tank, 526 Contaminated evaporators, 491, 524 Contaminated steam system, 524 Control systems command input, 793 configurations, 791 control medium, 793 design of, 794 hardware, 793 operation, 792 proportional band, 793 % types of, 791 Corrosion crevice, 814, 836 galvanic, 472, 488, 811-813, 837, 828, in seawater, 811-817, 827, 829 of metals, 810 types of, 811 Cost effectiveness, 5 Couplings electrical, 360 reduction gear, 326 shafting, 377 Creep, of materials, 821-823, 836
Deaerating feedwater heaters, 21, 27, 70, 519, 690 Dealurninisation, 815 Deck machinery; see Hull machinery Deepwell pumpa, 420, 708 Desuperheaters, 111 Dezincification, 815, 836 Diesel, Dr. Rudolf, 3 Diesel engines acceleration characteristics, 254 accessories, 275, 306 . air-cooled, 247 air filters, 272 air intake silencers, 272 arrangement aboard ship, 14, 258, 308 attached pumps, 253 automation, 276, 307,804 bore, 246, 280 bottom dead center, 246 brake mean effective pressure (BMEP), 246, 281 compression ratio, 246 connecting rod, 291 controls, 274, 307 cooling systems, 268, 294 cooling water treatment, 269 crankcase explosions, 277 crankshaft, 292 crankshaft bearings, 292 crosshead, 291 cylinder arrangements, 247 cylinder design, 286 cylinder ratings, 280 defined, 246 direct reversing, 249 displacement, 246 dual-fuel operation, 305 exhaust mufflers, 273 exljaust system, 272 fishook curves, 252 four-stroke, 247 fuel coalescer, 264 fuel consumption, 252, 281 fuel filters, 264 fuel injection, 297-299 fuel oil system, 263 fuel purifier, 264 fuels, 261, 854 gaseous fuel operation, 305 governors, 274 heat balance, 281 heat losses, 268 heavy fuel operation, 303-305 idling speed, 253 ignition delay, 246 inner dead center, 246 installation items, 277 instrumentation, 275, 306 intake system, 272 light-load operation, 253 liquid-cooled, 247
Diesel engines (continued) lubricating oil consumption, 267, 281 lubricating oil coolers, 268 lubricating oil filters, 267 lubricating oil system, 265 lubricating oils, 265 lubrication, 296 maintenance, 256, 264, 269, 269, 272, 277, 306-307 marine requirements, 250 naturally mphated, 248 number of, 259 opposed piston, 247, 282, 284 performance characteristics, 251 physical characteristics, 256, 281 piston and piston rod, 288-291 piston speed, 247, 280 power output, 251 principal structure, 281-283 propulsion plants; see Propulsion plants ratings, 255 repair parts, 306 reversing, 257, 299-303, 308, 807 safety features, 277 scavenged, 248, 282 scavenging systems, 283,285-287, 293 selection of, 258 shipboard applications, 257 shipboard installation, 308 speed classifications, 250 starting, 249, 254, 273, 299, 301-302, 307 stroke, 246, 280 stroke-to-bore ratio, 280 supercharged, 248, 292 tests, 256 top dead center, 246 torque characteristics, 251 torsional vibration, 277 two-stroke, 247 types of, 247 unidirectional, 249 waste heat utilization, 73-75, 270, 281, 555 weights, 16, 280 Diesel-generator sets, 611 Distilling plants ' acid cleaning, 558 air eduptors, 561 air ejectors, 533, 559 automation, 561,801 basket evaporators, 510, 543 brine recirculation systems, 536 chemical feed treatment, 557 cold shocking, 530, 540, 541 condensate-cooled plant, 538 . design calculations, 552 distillate purity, 530 early designs, 530 economy, 555 electrodialysis, 549 feedwater heater, 533
1 I'
862
Distilling plants (continued) , flash evaporators, 531,552 fouling resistance, 550, 552 for harbor duty, 536 heat sources, 555 heat transfer, 550 heat-transfer coefficient, 551 instrumentation, 536 logarithmic mean temperature difference, 550,551 materials of construction, 556, 829 membrane processes, 547 once-through systems, 537 overall heat-transfer coefficient, 553 scale control, 557 spray-film evaporators, 538, 542, 544 submerged tube, 530 temperature losses, 552 thin-film evaporators, 538 vacuum equipment, 558 vacuum pumps, 561 Drinking water cooling, 734 Economizers, 22, 85, 108, 109, 125
for distilling plants, 533, 559 performance, 444 Electric generators accessories, 610, 611 characteristics, 607 electrical characteristics, 613 emergency generators, 613 gas turbine driven, 612 installation, 607
special arrangements, 613 steam turbine driven, 201,607
INDEX
MARINE ENGINEERING
Electric plants (continued) generating capacity, 607 generating plants, 201 lighting; see Lighting list of motors and controls, 622-624 motor control equipment; see Motor control equipment motor-generator sets, 634 motors; see Motors panels; see Panels power; see Electric power switchboards; see Switchboards rules and regulations, 605 wiring, see Wiring Electric power branch circuits, 645 circuit proteotixe devices, 652 circuit selectivity, 652 circuits, 649 distribution, 643-646, 648-653 emergency, 646-648 equipment, 621 fault-current analysis, 652 feeders, 643 list of feeders and mains, 651, 652 panels, 645 short-circuit analysis, 652 special features, 645 voltage drop, 648-649 wire current capacity, 649 wire size, 648 wiring, 664 Electric propulsion drives a-c, d-c systems, 337 alternating-current drives, 336 amperage, 335 applications, 334,343,348, 355, 356 cable ship Long Lines, 348 design features, 341, 353 diesel a-c systems, 356 diesel d-c systems, 339 direct-current drives, 335 efficiency,337,338,341,353,355 icebreaker Glacier, 335,346 icebreaker Lenin, 335 insulation, 341, 354 oceanographic survey ship Wilkeg 343 passenger ship, Canberra, 355 physical characteristics, 342,349,355 power rating, 335, 336, 337, 353 reversing characteristics, 349,351,357-359 significant features, 334 submarine tender Hunley, 356 T2 tankers, 355 turbine a-c systems, 348 turbine d-c systems, 347 types of systems, 335 voltage, 335, 337,338,341, 353
Electrodialysis, 549 Evaporators; see Distilling plants 1
I
Fans, 756 Fatigue data presentation, 818-820 factors affecting fatigue life, 820 of metals, 817 stress-strain relationships, 817 Feedwater heaters high pressure, 490, 521 low pressure, 490, 518 Feedwater system, 21,688 Final design, development of, 33 Fire hose, 703, 705 Fire hydrants, 703 Fire protection, 703 Flash evaporators, 531, 552 Flue gas conductivity, 39 sensible heat, 39 specific heat, 39
fundamentals, 436 laws of similitude, 438 performance, 438
combustion constants, 96 ash content, 13, 804,850 bunker fuels, 848 carbon residue, 850 cetane number, 261,849 cetane index, 850 corrosion test, 858 diesel engine, 261, 854 diesel index, 850 distillation processes, 842 distillation test, 847 distribution systems, 854 flash point, 262,847 fuel specifications, 852, 854, 855 gas turbine, 850,852 heat of combustion, 95-97,261,304,848,849 lurninometer number, 848 names for, 856 pour point, 262, 845 properties, 13, 844
Fuels, petroleum (continued) Ramsbottom carbon residue test, 850 sediment, 853 specific gravity, 304, 845 sulfur content, 14, 262, 304, 853 test methods, 845 types, 13 , ' viscosity, 262, 304, 846-848 Galvanic corrosion, 472,488, 811413,837, 828,830 Gas turbine generator sets, 612 Gas turbines accessories, 218 air bleed, 212 air inlet, 213 arrangement, 213 arrangement aboard ship, 15 automation, 808 axial-flow compressors blading, 226 performance, 225 rotor design, 227 stator design, 228 basic considerations, 206 bearings, 239 centrifugal compressors diffuser design, 223 impeller design, 223 , performance, 222 rotor design, 224 combined cycles, 75, 173, 211 combustion systems configurations, 235 design objectives, 236 design parameters, 236 , fuel n o z z l ~237 ignition systems, 239 mechanical details, 237 compression ratio, 207 controls, 219 cycle performance, 206 effect of ambient conditions on performance, 209 effect of duct losses on performance, 209 exhaust collector, 214 exhaust silencers, 219 fuel consumption, 207 fuel control, 221 fuel nozzles, 237 fuels, 212, 850, 852, 854 ignition system, 239 inlet air filters, 218 inlet silencers, 219 installation, 212 intercooling cycle, 207 lubrication systems, 241 maintenance, 213 materials of construction, 831
INDEX MARINE ENGINEERING
Gas turbines (continued) nozzle construction, 231 nozzle design, 231 operation, 213 overspeed protection, 222 power control, 221 propulsion plants; see Propulsion plants recuperator, 216 reduction gearing, 217 regenerative cycle, 207 regenerator, 216 reheat cycle, 207 reversing considerations, 217 simple cycle, 207 single shaft, 213 specific horsepower, 208 specific speed, 229 specific weight, 208 speed measurement and control, 219 starting, 212, 218, 239 structural arrangement, 216 temperature control, 221 thrust bearings, 240 turbine aerodynamic design, 230 turbine rotor design, 233 turbine velocity diagrams, 230
Governors, 184, 201 Great Eastern, steamship, 2 Heat balance calculations, 19, 61-73, 76, 90 Heat exchangers applications, 490, 514 bpiler feedwater test sample coolers, 491, 517 characteristics, 490,491 construction details, 495 contaminated drain inspection tank, 526 contaminated evaporators, 491, 524 deaerating feedwater heaters, 21,27,70,519,690 design calculations, 504, 506, 508, 511 design considerations, 488 design data requirements, 496 design features, 490, 491 desuperheater test sample cooler, 517 double-tube, double-tube-sheet, 328 double-tube-sheet, 528 effective flow area, 501, 502, 504 fin-tube designs, 510-514 flow arrangements, 498 flow b a e s , 493, 495 fuel-oil heaters, 491,514
,/
Heat exchangers (continued) gland leak-off condensers, 491,524 head joint, 492, 493, 524, 526, 528, 529 heat-transfer calculations, 504, 506, 511 heat-transfer fdm coefficient,498-500, 503, 510, 511 heat-transfer relationships, 496 high-pressure feedwater heaters, 490, 521 logarithmic mean temperature difference, 497 low-pressure feedwater heaters, '490, 518 lubricating-oil coolers, 490, 514 lubricating-oil purifier heaters, 517 overall heat-transfer coefficient, 497, 504, 511 materials of construction, 490, 491, 829 pressure drops, 505-509,512-514 shell, 493 tank cleaning system drain coolers, 490, 518 tank cleaning system heaters, 490,518 thermal expansion provisions, 495, 496 tube pitch, 494,495 tube sheet, 493, 528, 529 tube-to-tube-sheet joint, 489, 528, 529 unfired steam generators, 491, 524 water heaters, 491, 526 Heat transfer, 39 conduction, 40 convection, 41 overall, 42 radiation, 41 Heating equipment direct radiation heaters, 765 electric heating, 763 heating coils, 763 unit heaters, 765 Heating and ventilation systems air handling systems; see Air handling systems design criteria, 726 ' heating equipment; see Heating equipment heating loads, 728 overall consideratione, 23 piping systems, 734-740 types of, 721 Hot-water heating systems, 735, 740 Hull machinery active fin stabilizers; see Active fin stabilizers anchor windlasses; see Anchor windlasses bearings, 565 bow thrusters; see Bow thrusters capstans, 594 cargo winches; see Cargo winches centralized hydraulic systems, 568 constant-tension mooping winches, 595 electrical details, 569 gears, 565 hydraulic details, 566 mechanical details, 564 procurement of, 34 special thrust devices, 597 steering gears; see Steering gears topping winches, 593 %
Hull machinery (continued) types of drives, 564 vang winches, 593 warping winches, 595 Human engineering, 795 Hunley, submarine tender, 356 Hydraulic systems, 566
I I
i
Ice-making load, 734 IZe-&-France, passenger liner, 310 Insulation, thermal, 715 Interior communications alarm systems, 655456 communication systems, 656-657 control system, 657 indicating systems, 654-655
Jarnestown, tanker, 371
Lubricating-oil systems (continued) lubricating-oil purifier, 27, 517, 788 piping, 692, 785-790 pumps, 787 reduction .gear, 326, 692,785-790 steam turbines, 196,692,785-790 strainers, 787 tanks, 786 types, 785 valves, 788 Lubricating-ail viscosity index, 265,775 Lubrication boundary, 772 fluid film, 770 lubricant characteristics, 265, 775-777 lubricants, 265, 774 mixed film, 772 pressure formation, 770 /
Machinery space access, 675 hrangement, 13, 33, 670 Labyrinth seals, 53, 54, 194, 241 Latent heat, 713 Main feed pumps, 22, 28, 403, 417, 688 Main propulsion system requirements, 8 Maintainability, 5, 12 Marine engineering branch circuits, 642 distribution, 640 defined, 1 emergency, 646-648 early history, 2 emergency distribution, 647 Materials applications, 824 creep, 821423,836 high temperature, 821 low temperature, 834 metallurgical terms, 835 of construction, 810 of various spaces, 636, 637 Membrane processes, 547 Memrmmac, ironclad, 3,78 searchlights, 639 Mollier chart, 164 signal lights, 638 voltage drop, 648449 Monitor, ironclad, 3, 78 Mooring winches, 595 wire size, 648 Motor control equipment a-c electrical characteristics, 628 controller panels, 627 Logarithmic mean temperature difference, 42,474,475,497, 550, 551 d-c motor starters, 629-632 Logistic support, 5 master switches, 628 protection features, 628, 629 Long Lines, cable ship, 348 Longitudinal vibration; see Shafting, longitudinal vibraspeed-regulating rheostats, 628 starter panels, 627 Lubricant characteristics, 265, 775-777 Motor-generator sets, 634 Lubricant properties, 265, 774 Motors a-c electrical characterisltics, 626 Lubricants, 265, 774 ambient temperature, 625 Lubricating-oil coolers, 27, 490, 514, 788 alternating current, 626 Lubricating-oil purifier heaters, 27, 517,788 Lubricating-oil systems bearings, 625 condition monitoring, 789 d-c electrical characteristics, 627 designs, 626 direct current, 627 duty ratings, 626 enclosures, 621
MARINE ENGINEERING
I
INDEX
I
list of, 622-624 mechanical characteristics, 621 space heaters, 625 speed reduction, 626 terminal boxes, 625 ventilation, 621 wound-rotor, 626, 627 Navigation lights, 638 Navigational systems applications, 660
air conditioning, 718 Nuclear reactors applications, 149 basic fundamentals, 130 boiling-water reactors, 155 Consolidated Nuclear Steam Generator, 151
gas-cooled reactors, 140, 156 Geiger counters, 138 health physics, 146 ionization chambers, 137, 147 materials of cdnstruction, 826 neutron detection, 138 nodssionable material conversion, 136 organic-moderated reactor, 157
radidactive waste disposal, 146, 151 radiological safety, 145 reactor arrangement, 138 reactor controls, 138, 142, 150, 153
Nuclear reactors (continued) reactor coolants, 140 reflector, 138 safety, 144, 153 SavanncJ1,146, 148, 149 shielding, 139, 143, 145, 150, 153 UNIMOD pressurized-water reactor, 153 Obiapo, tanker, 371 Observation Island, cargo ship, 371 Operability, 5 Operations research, 5 Otto Hahn, nuclear ship, 151
controller, 627 distribution, 620 emergency, 647-648 lighting, 641 power, 645 starter, 627 test, 621 types of, 620 Parametric studies, 7 Perfect gaa laws, 38 Petroleum fuels; see Fuels, petroleum Piping arrangements, 676 design details, 676 design velocities, 678, 679 determination of pipe sizes, 678 diagrams, 34 fittings, 681 flange attachments, 681 flexibility, 679 pressure losses, 678, 679 safety valves, 688 sea connections, 682 valves, 68'1 wall thickness, 679 working plans, 34 Piping systems air escapes, 700 auxiliary exhaust systems, 686 = auxiliary steam, 825 ballast systems, 696,703,705 bilge and ballast systems, 696 boiler pressure auxiliary steam systems, 683 brine cooling systems, 745 cargo-oil systems, 705 chilled-water systems, 740 condensate systems, 690 contaminated steam system, 524 drain systems, 690 fire main systems, 703
Propulsion plants (continued) fuel consumption, 16 fuel types, 13 machinery arrangement, 13,33, 670 maintainability, 13 operating personnel, 17 rating limitations, 18 reliability, 12, 22 reversing capability, 17 selection of, 41 space and pangement requirements, 13 steam conditions, 19-20, 90, 159 weight requirements, 13, 16 Propulsors efficiency of, 10 selection of, 9 Psychrometric chart, 711-712 Psychrometry, 711 Pumps ballast, 421, 696 basic fundamentals, 401 bilge, 28, 421, 696 cargo, 420,705 centrifugal cavitation, 413 clearances, 365 characteristics, 414 contrarotating, 10,385 efficiency, 408,411 head, 4Q3 controllable and reversible pitch, 10, 11, 217, 260, 385, head-capacity curves, 406 laws of similitude, 408410 detachable blade, 385 losses, 408 efficiency of, 10 net positive suction head, 413-421 fixed pitch, 385 performance curves, 410 fully cavitating, 10, 386 specific capacity, 409 specific speed, 409,412,413 suction lift, 413 . suction specific speed, 413 type% 401 velocity diagrams, 405 viscosity effects, 412 condensate, 28,415417,455,690 condenser circulating, 28,415, 690 deepwell, 420, 708 rpm-torque curves, 351 drain, 415-417 selection of number of blades, 369, 386 feed, 22, 28, 403, 417, 688 feed booster, 419 general service,.421 thrust eccentricity, 367, 368, 371, 373 jet propulsion, 421 Troost B Series, 9, 10 liquid ring, 434, 561 main feed, 22, 28, 403, 417, 688 materials, 421 power axial piston, 431 arrangement, 14, 15, 20 fixed-stroke, 428 automation, 797, 804, 808 radial piston, 429
Piping systems (continued) freshwater systems, 698, 827 fuel-oil filling, transfer, and oily ballast systems, 694 fuel-oil service systems, 694 hot-water heating systems, 735, 740 lubricating-oil systems, 692, 785-790 main feed systems, 688 main steam systems, 682, 825 materids of construction, 825,827 plumbing vents, 701 overflows, 702 refrigeration systems, 740-745 safety valve and escape piping, 687 seawater systems, 690, 827 sounding arrangements, 702 steam heating systems, 734 turbine bleed systems, 684 vents, 700,701 Preliminary design considerations, 18 Preliminary design spiral, 8 Pressurized-water nuclear reactor, 149,151,153,171 Producibility analyses, 5
establishment of rating, 10 fractional power performance, 17
variable stroke, 429 primary coolant, 421
INDEX
MARINE ENGINEERING
Pumps (continued) reciprocating steam
steam consumption, 422
internal gear, 433 liquid piston, 434, 561
Radio communication, 659 Recuperator, 216 Reduction gears accessories, 330 applications, 331
for contrarotating propellers, 332 for hull machinery, 565 gear wheels, 323 involute geometry, 314 journal bearings, 325 K factor, 314,324 locked-train, 311, 331 lubrication, 317, 325, 326, 329 lubricating-oil systems, 326, 692, 785-790 manufacturing methods, 311 materials of construction, 833 nested gears, 311 nodal drive, 321, 389, 390 pinion deflection, 318, 319 pinion wheels, 323 planetary, 311, 332 reversing, 327, 332 size approximations, 317 slow-speed gear misalignment, 320
Reduction gears (continued) thrust bearings, 325 timing, 311 tooth bending strength, 315 contact, 319 contact prwure, 313 design factors, 313 diagonal loading factors, 316 pitch, 311, 318 scoring, 316 stresses, 313 unit loading, 316 turning gear, 199, 200, 330 weight estimates, 331 Refrigerants, 715 Refrigerated cargo, 724 Refrigerated cargo containers, 724 Refrigerated compartment cooling loads, 730 Refrigerated compartment design criteria, 730-734 Refrigerated stores, 723, 740 Refrigeration absorption cycle, 714 compression cycle, 713-715 procem of, 713 Refrigeration equipment absorption systems, 768 brine coolers, 767 centrifugal compressors, 767 condensers, 766 defrosting facilities, 767 evaporators, 766 liquid chilling systems, 767 receivers, 766 reciprocating compressors, 766 steam-jet systems, 768 Refrigeration piping systems, 740-745 Regeqerators, 216 Reheat turbines arrangement, 72, 170 gas reheat, 167 state line, 72 steam conditions, 72, 166 steam reheat, 170 turbine performance, 73, 166 Reliability, 5, 12, 22 Resonance changers, 394, 783 -. Reverie, steam yacht, 78 Reverse osmosis, 547 Reynolds number, 409 Robinson, victory ship, 371 Safety, 5, 797 Savannah, nuclear ship, 146, 148, 149, 171 Savannah, steamer, 2 Scoops, 690 Sea trials, 35 Searchlights, 639
Sensible heat, 713 Sensible heat factor, 711-713 Sensible heat of g a s , 39 Sensitivity analyses, 7 Series 60 (model tests), 9
balance, 378 bearing locations, 373 bearing reaction iduence numbers, 374, 378 bearing reaction locations, 370 bearings, 379 bending stresses, 370 bent shafts, 379 cold rolling, 372, 377 computation of diameter, 372
line shaft bearings, 379 location of main engines, 365 longitudinal vibration acceptable limits, 397 damping, 395 determination of natural frequencies, 393 excitation changer, factors, 395 resonance 394,783 vibratory torque calculations, 396 materials of construction, 372, 832 nodal drives, 321, 389, 390 number of bearings, 373 propeller-to-shaft interface, 375 protection from seawater, 377 radial loads, 372 shaft rake, 366, 675 shaft withdrawal, 366 thrust load, 369 torsional loads, 369 torsional vibration, 321 acceptable limits, 392 analysis models, 389 damping, 391 excitation, 390 mode shapes, 389, 390 modes of, 388 natural frequencies, 389 nodal drives, 321, 389, 390 vibratory torque calculations, 392 vibratory loads, 367 weighing bearing reactions, 376, 378
Shafting (continued) whirling vibration acceptable limits, 398 determination of natural frequencies, 397 mode shapes, 397 Ship requirements, 7 Ship resistance, 9 Ships service air compressor4 27,440-444 Ship system constraints, 6 Shock resistance, 24 Signal lights, 638 Silicon controlled rectifiers, 335, 337, 339, 343, 614 Similitude, laws of, 408-410 Solar heat gain, 726-728 Sonar, 662 Sound conditioning, 718 Specific power, for vehicles, 6 Specific speed, 229, 409, 412, 413 Specifications preparation of, 31 ship, 31 Spray-film evaporators, 538, 542, 544 Steam cycles high-performance cycles, 72 regenerative cycle, 65 simple cycle, 61 Steam heating systems, 734 Steam turbines arrangement aboard ship, 14 astern operation, 200 astern turbine losses, 56 ' automation, 800 bar-lift control valves, 182 blade design, 185 blade erosion, 171, 187 blade fastenings, 187, 189 blade material, 186 blade shrouds, 188 blade vibration, 185 blading, 51 684-686 bleed steam, boiler pressure control, 183 bypass valves, 183 casings, 193 choked flow, 165 conditiqn curve, 164 cross-compound, 178 159 design considerations, diaphragms, 191 double-flow, 180 emergency operation, 200 erosion, 171, 187 exhaust losses, 55, 56, 63, 164, 165 exhaust vacuum, 160 external moisture separation, 172 external thermodynamics, 55 extraction of steam, 163 extraction pressures, 58
!r
I
INDEX
MARINE ENGINEERING
Steam turbines (continued) gland leakage, 55, 63 for nuclear cyoles, 171 gland seal system, 195, 524 governors, 184, ?01,610 hand control valves, 180
low oil pressure protection, 184 lubricating-oil systems, 196, 692, 785-790
piping, 682487 propulsion plants; see Propulsion plants reaction blading, 189 reaction stages, 174 . reduced-pder operation, 59 reheat turbines; see Reheat turbines rotor balance, 189,190 rotor critical speed, 189 rotor design, 189, 190 rotor m a t e d , 190 single-cylinder, 178 single-stage auxiliary, 203 stage efficiencyp54 stage pressures, 163 state line, 54,. 55,56,63,72,164 steam conditions, 19, 159 steam induction, 164 steam rate, 55, 59, 63, 161, 164, 165, 610 throttle valve, 180 thrust bearings, 197 torque and speed characteristics, 165 turbine control, 180 valve design, 184 vector diagrams, 51
Steering gears (continued) types, 570 vane type, 571 Steering system, 23 Stern tube bearings, 382 Stress corrosion, 816 Strouhsl number, 459 Strut bearings, 382 Superheaters, 105-108 Sustained sea speed, 10 Switchboards applications, 615 arrangement of, 615,640 characteristics, 615 circuit protective devices, 619 emergency, 616, 640, 647-648 load-center, 619, 641 main generator and distribution, 615, 616 one-line diagram, 617-619 ratings, 615 requirements, 614 selectivity, 619 types, 615 Systems analysis, 3, 4, 5 Systems engineering, 4 T2 tankers, ,355 Tank cleaning system drain coolers, 490, 518 Tank cleaning system heaters, 490, 518 Taylor power coefficient, 9 Taylor's gtandard Series, 9 Tests, 35 Thermal insulation, 715 Thermodynamic fundamentals, 38 Thermodynamics of steam cycles, 61 Thrust bearings, 197, 240, 325, 379 resonance changers, 783 thrust meters, 783 types, 781-783 Thrust devices jet type, 600 trainable type, 597, 599 tunnel type, 597 types, 597 Thrust meters, 783 Topping winches, 593 Torsional vibration; see Shafting, tossional vibration Transformers, 633 Turbine-generator sets, 26,59-61,201,607, 610 Turbines, auxiliary; see Auxiliary turbines Turbines, early history, 310 Turbinia, gunboat, 3 Turtle, submaxine, 3
power units, 575 ram group design, 573 Rapson-slide type, 573 rudder torque rating, 571
Vacuum pumps, 434,547, 561 Vang winches, 593 Vapor-compression distilling plants, 542-547
Vapor migration, 716 Ventilation systems air handling systems; see Air handling systems design criteria, 726 overall considerations, 23
Vespasian, turbine ship, 310 Virginia, ironclad, 78 Viscosity of lubricants, 265, 775 Voltage drop, 648-649 Wampanoag, cruiser, 78 Warping winches, 595 Waste heat from diesels, 73-75, 270, 281, 555
Waste heat (continued) from gas turbines, 75 Water consumption rates, 700 Water heaters, 491, 526 m i l i n g vibration; see Shafting, whirling vibration Wilkes, oceanographic survey ship, 343 Wiring applications, 663 connections, 669 determination of lengths, 668 grounding, 669 installation of, 664 methods, 663 s t f i g tubes, 665 terminals, 669 Wiring appliances, 639 Work study, 5 Working plans, 33,34
