Developments in Petroleum Science, 19A
surface operations in petroleum productlon, I
FURTHER TITLES IN THIS SERIES 1 A.G. COLLINS GEOCHEMISTRY OF OILFIELD WATERS 2 W.H.FERTL ABNORMAL FORMATION PRESSURES 3 A.P. SZILAS PRODUCTION AND TRANSPORT OF OIL AND GAS 4 C.E.B. CONYBEARE GEOMORPHOLOGY O F OIL AND GAS FIELDS IN SANDSTONE BODIES T.F. YEN and G.V. CHILINGARIAN (Editors) 5 OIL SHALE 6 D.W. PEACEMAN FUNDAMENTALS OF NUMERICAL RESERVOIR SIMULATION G.V. CHILINGARIAN and T.F. YEN (Editors) 7 BITUMENS, ASPHALTS AND TAR SANDS 8 L.P. DAKE FUNDAMENTALS O F RESERVOIR ENGINEERING 9 K.MAGARA COMPACTION AND FLUID MIGRATION 10 M.T. SILVIA and E.A. ROBINSON DECONVOLUTION OF GEOPHYSICAL TIME SERIES IN THE EXF'LOR .ATION FOR OIL AND NATURAL GAS 11 G.V. CHILINGARIAN and P. VORABUTR DRILLING AND DRILLING FLUIDS 12 T.D. VAN GOLF-RACHT FUNDAMENTALS OF FRACTURED RESERVOIR ENGINEER.ING 13 J. FAYERS (Editor ENHANCED o n RECOVER4 14 G.MOZES Editor) PARAFFIN PRODUCkS 15A 0 . S E R R A FUNDAMENTALS OF WELL-LOG INTERPRETATION 1. THE ACQUISITION OF LOGGING DATA
15B O.SERRA FUNDAMENTALS OF WELL-LOG INTERPRETATION 2. THE INTERPRETATION OF LOGGING DATA
16 R.E. CHAPMAN PETROLEUM GEOLOGY 17A E.C. DONALDSON, G.V. CHILINGARIAN and T.F. YEN ENHANCED OIL RECOVERY, I FUNDAMENTALS AND ANALYSES
18A A.P. SZILAS PRODUCTION AND TRANSPORT O F OIL AND GAS A. FLOW MECHANICS A N D PRODUCTION
second completely revised edition 19A G.V. CHILINGARIAN, J.O. ROBERTSON Jr. and S. KUMAR SURFACE OPERATIONS IN PETROLEUM PRODUCTION, I 19B G.V. CHILINGARIAN, J.O. ROBERTSON Jr. and S. KUMAR SURFACE OPERATIONS IN PETROLEUM PRODUCTION, I1 20 A.J. DIKKERS GEOLOGY IN PETROLEUM PRODUCTION
Developments in Petroleum Science, 19A
surface operations in petroleum production, I G.V. CHILINGARIAN Petroleum Engineering Department, University of Southern California, University Park, Los Angeles, CA 90089-121 1 , U.S.A. J.O. ROBERTSON, Jr. Earth Engineering, Inc., 4244 Live Oak St., Cudahy, CA 90201, U.S.A. S. KUMAR Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-121 1 , U.S.A. Associate Editors: T.A. BERTNESS and C.M. BEESON Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211 , U.S.A. with contributions from: Moayed Yousif Al-Bassam Axelson, Inc. Donald U. Bessler William G. Carter Erle C. Donaldson Clarence Dunbar Bruce A. Eckerson Kern H. Guppy W.B. Hatcher Arnold L. Johnson
Donald G. Knox Frank J. Lockhart Dawood Momeni R.L. Pettefer W.J. Powers K.M. Sasseen Varec, Inc. L.C. Waterman Phil Wilson
ELSEVIER AMSTERDAM - OXFORD - NEW YORK - TOKYO 1987
ELSEVIER SCIENCE PUBLISHERS B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands
Distributors for the United States and Canada: ELSEVIER SCIENCE PUBLISHING COMPANY INC. 52,Vanderbilt Avenue New York, NY 10017, U.S.A.
Library of Congca C
~at.
~
~
I
I
~
Chilingarisn, George V., 1929Surface operations in petroleum production. (kvelopments in petroleum science ; 19) Includes bibliographies and index. 1. Petroleum engineering. I. Robertson, John 0 . 11. Kumar, 6. (SanJay), 1960111. T i t l e . IV. S e r i e s .
.
TN871.C495 1966
622
' .336
ISBN 0-444-42473-3 (U.S. :
V.
06-29126 1)
ISBN 0-444-42473-3 (Vol. 19A) ISBN 0-444-41625-0 (Series) 0
Elsevier Science Publishers B.V.,1987
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording o r otherwise, without the prior written permission of the publisher, Elsevier Science Publishers B.V./Science & Technology Division, P.O. Box 330,1000 AH Amsterdam, The Netherlands. Special regulations for readers in the USA - This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, including photocopying outside of the USA, should be referred to the publisher. Printed in The Netherlands
This book is dedicated to George Deukmejian, Governor of California, for his support of the petroleum industry.
VI
CONTRIBUTORS M.Y. AL-BASSAM Axelson, Inc. C.M. BEESON
D.U. BESSLER T.A. BERTNESS
W.G. CARTER G.V. CHILINGARIAN
E.C. DONALDSON C. DUNBAR B.A. ECKERSON K.H. GUPPY
W.B. HATCHER A.L. JOHNSON D.G. KNOX
S. KUMAR
F.J. LOCKHART
D. MOMENI
Getty Oil Co., P.O. Box 1, Mina Saud, State of Kuwait P.O. Box 2427, Longview, TX 75606, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Tretolite, Petrolite Corporation, 369 Marshall Avenue, St. Louis, MO 63119, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. (also: Consultant, 14827 La Cuarte Street, Whittier, CA 90605, U.S.A.) Earth Engineering, Inc., 4244 Live Oak Street, Cudahy, CA 90201, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Petroleum & Geology Engineering, University of Oklahoma, Norman, OK 73019, U.S.A. Senior Application Engineer, TRW Reda Pump Divi sion, Bartlesville, Okla., U.S.A. Vanson Engineering Co., 1061-B Kraemer Place, Anaheim, CA 92806, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Texaco, Inc., Taft, CA 93268 Vanson Engineering Co., 1061-B Kraemer Place, Anaheim, CA 92806, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Chemical Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A. Petroleum Engineering Department, University of Southern California, Los Angeles, CA 90089-1211, U.S.A.
VII
R.L. PETTEFER W.J. POWERS J.O. ROBERTSON, Jr.
K.M. SASSEEN Varec, Inc. L.C. WATERMAN P. WILSON
Petrolite Corporation, Petreco Equipment Group, 369 Marshall Avenue, St. Louis, MO 63119, U.S.A. Manager Marketing Services, TRW Reda Pump Division, Bartlesville, Okla., U S A . Earth Engineering, Inc., 4244 Live Oak Street, Cudahy, CA 90201, U.S.A. HTI-Superior, Inc., Berry Industries Company, P.O. Box 3908, Santa Fe Springs, CA 90670, U.S.A. VAREC Division, Emerson Electric Co., 10800 Valley View Street, Cypress, CA 90630, U.S.A. Petrolite Corporation, Petreco Equipment Group, 369 Marshall Avenue, St. Louis, MO 63119, U.S.A. Kobe Inc., 3040 East Slauson Avenue, Huntington Park, CA 90255, U.S.A.
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
IX
PREFACE This book has been developed at a time when oil producers are taking a much closer look at the economy of oilfield operation and redesign of production technology to improve ultimate recovery. The very high cost, and risk, of the search for new oilfields demands the reevaluation of production technology and reservoir engineering to improve the production characteristics of existing oilfields. This book serves the need of field production managers and engineers for a comprehensive reference on petroleum production surface operations and technology. Teachers and students will find that it fills the need for a text on production technology to complement the study of reservoir engineering. The book presents a lucid description of the physical and chemical properties of the fluids encountered by engineers in the field. The properties, methods of separation, measurement, and transportation of these fluids (gases, condensate liquids derived from natural gas, crude oils and oilfield waters) are also presented. Thus these chapters form the basis for the explanation of surface equipment, which exploits physical and chemical properties to process the fluids in the field for their processing and transportation. Oilfield emulsions and the chemistry of their control, as well as the equipment and processes used for emulsion separation in the oilfield, are discussed in the opening chapters of the book. Following a presentation of the fluids and their process technology, a series of chapters present a thorough discussion of every type of surface equipment that is encountered in the myriad aspects of oilfield operations ranging from waterflooding to new enhanced oil recovery techniques. Included are methods for pumping, water control, production logging and corrosion control. The book is so comprehensive that it also includes, as a third general category: well completion and work-over operations, methods for design and operation of underground gas storage, and a review of offshore technology. Thus, this two-volume book on oilfield surface operations offers a complete reference for all surface technology for the manager, engineer, teacher and student. The preparation of this valuable book is the result of cooperation from the experts in petroleum production who have devoted their time to the lucid expression of the knowledge that they have acquired through experience in the evaluation and solution of field problems, and development of economic field processes. Oil production companies have been generous in their cooperation through assistance and encouragement to the authors and permission to publish data, designs and photographs. It is hoped that this book will find a prominent place in the Developments in Petroleum Science Series and that it will be instrumental in the improvement of the global enhancement of oil production and ultimate recovery. Erle C . Donaldson, Ph.D., P.E. Petroleum and Geological Engineering University of Oklahoma
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
XI
CONTENTS CONTRIBUTORS PREFACE
........................................................
..............................................................
VI
1X
Chapter 1. INTRODUCTION TO SURFACE PRODUCTION EQUIPMENT K.M. Sasseen and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gatheringsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
..... .....
1 1 3 4
......
11
- Heater treaters, 6
......... . . . . . . . .. .......................
Sample problems . . . ...... . . Exchanger problem . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solution, 21 Washtank problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field data, 23 - Required, 23 Appendix 1.1 - Some fundamental fluid mechanics concepts and sample problems . . Fundamental equation of fluid statics . . . . . . . . . . . . . . . . . . . . Buoyancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General energy equation. . . . Compressible flow formula
. . . ... . . . . . ... .... . . ..... .. . .. .. ..... . . ........ . . . . . . . . .... . ... . ~
12
21
25
27 28
Solution, 28 Example problem 2. Compressible flow ( n o d e ) Solution, 29 Solution, 31
...... Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Saturated hydrocarbons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unsaturated hydrocarbons .... Naphthene hydrocarbons . ........................................ Aromatic hydrocarbons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Classification of petroleums . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .....
32 32 35 3s 31 38
XI1 Some rules of nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
39 40 41
Chapter 2. PHYSICAL PROPERTIES OF PETROLEUM GASES AND LIQUIDS F.J.Lockhart and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Density of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example2-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Density of liquids . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 2-2 .... Viscosity of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example2-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viscosity of liquids . ..........
Example24
..........................................................
43
44 46 47 47 49 50 50 52 52 55 55 56 56 58
Chapter 3. SEPARATION OF OIL AND GAS F.J. Lockhart, G.V. Chilingaxian and S. Kumar Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Equilibrium flash calculations Basic equilibrium relations for Types of separators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Internal parts of a separator . .... ..................... Factors influencing separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Separatordesi g n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gascapacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Liquidcapacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vessel design . . . . . . . . ... .... Example 3-1, 68 - Example 3-2, 70 Separator design using actual manufacturers’ field test data . . . . . . . . . . . . . . . . . . . . . . . . . . . Stage separation . . . . . . . . . . . . . . . . . .................. Determination of optimum pressure f stage is atmospheric . . . . . , . Determination of optimum pressures for three-stage separation (Whinery-Campbell technique) Methods of successive approximations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isothermal flash for two phases . . . Lockhart-McHenry method of flash-eq Example 3-3. . . . . . . . . . . . . . TheLockhartmethod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example3-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three-phase flash equilibrium . . . . . . . . . . . . . ................... ..... Sample problems and questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 3.1 - Raoult’s, Dalton’s and Henry’s laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 3.1-1 .....
59 60 60 61 63 66 67 67 67 68 71 71 73 73 77 77 78 81 81 83 83 84 85 86
XI11 Appendix 3.11 - Illustration, accessories, gas capacities, settling volumes, and specifications for: (1) vertical low-pressure separators; (2) vertical high-pressure separators; (3) horizontal lowpressure separators; (4) horizontal high-pressure separators; and ( 5 ) spherical separators . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
87 108
Chapter 4. OIL FIELD EMULSIONS AND THEIR ELECTRICAL RESOLUTION L.C. Waterman, R.L. Pettefer and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theory of emulsions
109
Electrical treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electric dehydrators .... Automated dehydration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample questions . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . .... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
115 118 121 123
124
Chapter 5 . CHEMICAL RESOLUTION OF PETROLEUM EMULSIONS D.U. Bessler and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nature of emulsions Role of the emulsifier. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stability of emulsions Crude oil production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reduced-temperature treating Chemical resolution process . . Action of demulsifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating procedures Chemical injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Agitation Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Settling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical systems Free water knock - Vertical heater treater, 136 - Horizontal heater treater, 136 - Electrical dehydration, 136 Resolution of oil-in-water emulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trouble shooting . Chemicalinjection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Freewaterknockout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flowsplitters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heat exchangers Gunbarrels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heatertreaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemelectric treater . . . . . . .... Producedfluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
125 125 125 126 126 128 130 131 132 133 133 133 134 134 134
136 137 137 138 139 139 140 140 140 141 142
XIV Waste oil treating systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Treatmenttanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
............ ............ Stokes’law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 5.1 - Derivation of Stokes’ law equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . .
142 142 143 144 144
145 146 146 147
Chapter 6. VAPOR RECOVERY Varec. G.V. Chilingarian and S. Kumar Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evaporation loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
......................
......................
149 149
Equipment required Design of vapor recovery systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storage pressures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vent valve pressure settings . . . . . . . .
158 161
...................................
165
Fast payouts from vapor recovery systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example (A) - System payout from vapor recovery only
166
Valve flow capacity
................ ................
......................................................... Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
......
168 169 169 169 170
Chapter 7. NATURAL GAS AND NATURAL GAS LIQUIDS B.A. Eckerson. A.L. Johnson and G.V. Chilinganan Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Naturalgas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gas processing plants . . . . . . . . . . . . . . . . . . . . . . . . . Gas specifications . ........................
Acid gas content Water content
.......
........ .....................
177
182
182
183
xv Carbon dioxide and air . . . . . . . . . . . . . . . . . . . . . . . . Hydrogen sulfide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Specific gravity ..................................... ........... Heating value . . . . . . . . . . . . . . . . . . . . . . . . Gas measureme ............................ Natural gas liquids .................... . . . . . . . . . . . . . . .
183 183 184 184 184 184 184
Ethane, 185 - Propane, 185 - Butane, 185 - Butane-pr gasoline, 185
.................................................... ......................
............................
185 185 185 188 189 189 190 191 191 193 198 198 200 201 202 204 204
..................................................... .......................
205 206
...........
.......................
........... Adsorption . . . . . . . . . . . . . . . . . . . . . . . . Refrigeration processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Absorption . . . . .
................................................. .................. Ethylene recovery plant problem
Description of natural gaso
...................................
209 210
Chapter 8. OIL AND GAS TRANSPORT S. Kumar and G.V. Chilinganan
.............. Introduction . . . . . . . . . . . . . . . . . .............. Fundamentals of flow in pipes . . . . . Allowable working pressure of pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................ Horsepower . . . . . . . . . . . . . . . . . . . . . . . Friction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Friction head loss in fittings and connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles of pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pumping mechanisms. . ............................................... . . .. .... Measurement of perform e . . . . . . . . . . . . . . . . . . . . . . . . . Oil pipeline transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................ Example 8-1 . . . . . . . . . . . . . . . . . . . . Solution . . . . . . . . . . . . . . . . . . Increasing flow capacity of pipelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1) Complete loop, 226 - (2) Partial loop, 226 Example8-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....
211 211 212 213 213 214 219 219 219 221 224 224 224 227
XVI Solution, 227 - Booster pump stations, 228 - Branching pipelines, 228 Nonisothermalflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fundamentals of heat transfer . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . , . . . , . . . . , . . Conduction, 231 - Convection, 232 Application of heat transfer concepts to buried pipelines . . . . . . . . . . . . . . . . . . . . . . . . . , . (1) Estimation of thermal constants from soil properties, 234 - (2) Estimation of thermal constants by direct measurement, 234 Examples-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solution . . . . . . . . . . . . . . . . . , . . . . . . . . . Steady-state flow in buried pipelines . . . . . . . . . . . . . . . . . . . , . . . . . , . . . . . . . . . . . . , . . Examples-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solution ................................... Viscosity, 244 - Density, 246 - Kinematic viscosity, 247 - Specific heat, 247 Thermal conductivity, 247 Example 8-5 ............................................ Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transient (unsteady state) flow of oil in buried pipelines Heating up of a cold line by introduction of hot oil, Transportation of heavy oils in pipelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pipeline transportation of natural gas Physical properties of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gas compressibility, 254 - Density, 256 - Viscosity, 258 - Specific heat, 259 Gasflowfundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Weymouth approximation, 262 Mean pressure evaluation ..................................... The hydrate point for hyd Gas transmission systems . . . . . . . . . , . . System of parallel lines, 266 - Lines in series, 267 Example8-6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
................................................ Sampleproblems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
230 230 234
236 236 237 238 239
247 248 249 253 254 254 261 263 265 266 268 268 269 269 273 273 275 276
Chapter 9. DESIGN OF FLOWING WELL SYSTEMS S. Kumar, K.H. Guppy and G.V. Chilingaian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reservoir fluid flow . . . . . . . . . . . . . . Example 9-1, 280 - Find, 280 - Solution, 280 - Example 9-2, 282 - Find, 282 -
................................................. Example 9-4, 289 - Find, 289 - Solution, 289 Multiphase flow in directional wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 9-5, 290 - Solution, 290
279 280
286 286 289 290
XVII Horizontal flow in surface flowlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Horizontal flow correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Working pressure traverse curves for horizontal flow . . . . . . . . . . . . Example 9-6, 292 - Solution, 292 Inclined or hilly terrain multiphase flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 9-7, 293 - Solution, 293 Flowthroughchokes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 9-8, 295 - Determine, 295 - Solution, 295 The overall production system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 9-9, 297 - Solution, 297 Vertical correlation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Horizontal correlation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample problems . , . . . . . . . . . . . . . . . . . . .............................. Appendix9.I . . . . . . . . . . . . . . . . . . . . . . . . .. Appendix 9.11 - Introduction to chokes . , . . .. References . . . . . . . . ...................................................
.
291 291 291 292 294 296 298 298 299 300 323 326
Chapter 10. WELL TESTING S. Kumar and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Drillstem testing . . . . . . . . . . Component parts of a conv ............................... Anchor shoe, 329 - Perforated anchor pipe, 329 - Pressure recorders, 329 - Packers, 331 - Equalizing valve,331 - Tester valve, 332 - Choke, 332 - Shut-in valve, 332 Circulating valve, 332 - Other components, 332 Drillstem test procedure . . . . . . . . . . . . . . Qualitative drillstem test interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Buildup and drawdown test fundamentals . . . . . . . . . . . . . Solution to the diffusivity equation for infinite reservoirs Pseudosteady state flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Radius of drainage and stabilization time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Drawdown test ................ Multiple-rate drawdown testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 10-1,351 - Given, 352 - Solution, 352 Pressure buildup test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 10-2, 355 - Solution, 355 Buildup following a long producing time .... Equivalent producing time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Drawdown and buildup tests in gas wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
................
327 327 329
333 336 340 343 344 346 347 350 353 356 356 357 359 365
testing, 367 - Vertical interference tests, 368 - Injection and fall-off tests, 368 - Well test analysis in the presence of a gas phase, 369 Sample questions and problems . . . . . . . . . . . References ...
369 370
Chapter 11. PRODUCTION LOGGING S. Kumar and G.V. Chilingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Logging devices
373 373
XVIII High-resolution thermometer . . . . . . ..................................... Gradiomanometer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... Inflatable packer flowmeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Continuousflowmeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fullbore spinner flowmeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Caliper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ........... Manometer . . . . . . . . . . . . . . . . . . . . . . . . . . .. ........... Radioactive tracer surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1) Velocity-shot method. 380 - (2) Timed runs (controlled-time method). 381 - (3) Differential injection method. 383 Miscellaneous tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interpretation of flowmeter logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Monophasicflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In-situ spinner calibration. 385 Polyphasicflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
......................... s. 387 - Triphasic flow types. 388 Flowparameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spinner response in diphasic flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interpretation of the gradiomanometer . . . . . . . . . . . . . . . . . . . . . . . Temperature surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic conditions . ............ ... ....................... Static conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1) Lost circulation. 393 - (2) Cementing. 393 - (3) Fracturing. 395 - (4) Pr 396 - (5) Fluid injection. 397 Appendix 11.1 - Production logging (field examples) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......................... Example 11.1-1 - PCT survey in a gas well . . . . . . . Example 11.1-2 - Completion evaluation: initial flow profile . . . . . . . . . Conclusion. 402 Example 11.1-3 - Evaluation of completion: monitoring of acidizing . . . . Example 11.1-4 - Evaluation of completion: monitoring the perforati completion in a gas well . . . . . Example 11.1.5 - Diagnosis of ater . . . . . . . . . . . . . . . . . . . Example 11.1-6 - Diagnosis of a well problem: gas channeling behind liner . . . . . . . . . . . . . Example 11.1-7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Production logs. 411 .Quick-look interpretation. 412 - Quantitative interpretation. 412 - Conclusion. 412 Sample problems and questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
374 374 376 377 377 378 380 380
383 384 384 386 387 388 389 390 391 393 393
401 401 401 403 404 408 408 411
412 413
Chapter 12. GAS LIFT J.O. Robertson, Jr., G.V. Chilinganan. W.G. Carter and S. Kumar
Introduction Review of gas lift fundamentals Pressure gradients . . . . . . . Derivation of pressure at bottom of gas column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Energy utilized in lifting fluids Types of gas expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isothermal expansion. 420 . Adiabatic expansion. 421 . Polytropic expansion. 421 Volume of gas necessary for gas lift . . .
415 416 416 416 420 420 421
XIX Gas lift efficiencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) Leakage, 422 - (b) Entrance and discharge losses, 422 (d) Friction losses, 422 - (e) Back pressure at discharge, 422 Kick-off pressure (without valves) . . . . . . . Gas volumes necessary for gas lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fluid velocity in eductor tube . . . . . . . . . . . . . . . . . . . . Average density of fluid in eductor tube . . . . . . . . . . . . . Principles and methods of gas lift Gas lift terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Types of gas lift . . . .......................................
431
Spacing between gas lift valves . . . . Continuous flow gas lift (unbalanced Example 12-1 . . . . Given, 446 - Solution, 447 Intermittent gas lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Kick-off valves . . . . . . . . . . . . . . . . . . . . ............. Flow type valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential valves .............................. Mechanically-controlled valves . . . . . . . . . . . . . . . Pressure valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
453 454
......................................
456
.....................
460
..............................
461
Sample problems and questions References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter 13. PLUNGER LIFT C.M. Beeson, D.G. Knox, M. Al-Bassam and G.V. Chilinganan Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History . . . . . . ..................... Free-cycling plunger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cycle-controlled expanding plunger .................................... Equipment developments . . . . . . . . . . . . . . . . . . . . . . . . . . . Type M plunger lift Christmas tree . . . . . . . . . . . . . . . . . . . Turbulent-seal plunger ...................... Tandem plunger . . . . . . . . . . . . . . . . . . . . . Segmented retractable pads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Removable footpiece .............................. Taylor Type K cycle controller . . . . . . . . . . . . . . . . . . . . .
467 468 46 8 468 469 469 473 474 474 474 474
xx Time cycle controller with attachments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electronic controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early prediction methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Well data for cycle-controlled expanding plunger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Least squares equations for plunger lift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Description of plunger lift and need for equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Net operating pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gas-liquid ratio gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pressurebuildup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maximum production rate . . . . . . . . . . . . . . . . . . . . . ........ Method of obtaining equations for gas and pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wells with 2.in . plungers .................... ...................... Wells with 2.. in. plungers . . . . . . . . . . . . . . . . . . . ...................... Method of obtaining equations for maximum production rate . . . . . . . . . . . . . . . . . . . . . . . . . . .... Constructing plunger lift nomographs . . . . . . . . . . . . . . Need for nomographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Method of constructing nomographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estimating average pressure at point where gas enters tubing. 488 - Derivation of operating line for 2.in . plunger, 489 - Derivation of operating line for 2f.in . plunger, 490 - Mathematical operations by nomographs. 490 - Derivation of supplementary operating line for 2.in . plunger. 493 - Derivation of supplementary operating line for 2f.h . plunger, 494 Method of testing effects of various well conditions . . . . . . . . . . . . . . . . . . . . . . . . . Tailpipe, 495 - Water cut, 496 - Oil gravity, 496 How to use plunger lift nomographs ..................... Purpose of nomographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nomograph instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Step I, 497 - Step 11, 498 - Step 111, 498 - Step IV, 498 - Step V.A, 498 - Step V.B, 500 Plunger lift nomographs and examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typesofexamplewells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nomograph examples . . . ................. Accuracy expected from res ............................................. Plunger lift applications . . . . . . . . . . . ............... Use of equations and figures without Types of oil wells suitable for plunger lift ..................... General considerations, 510 - Capac How to determine if a well is a possible plunger lift candidate . . . . . . . . . . . . . . . . . . . . . . . Use of Figs. 13-10 and 13.11, 513 - Well depends entirely on formation gas, 515 - Well has gas available from some outside source, 515 Intermittent flowing and plunger lift system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Equipping wells at start for intermittent flowing and plunger lift (free piston) system . . . . . . . Tubing program, 515 - Christmas tree, 516- Well starting, 516 - Cycle controller for intermittent flowing, 516 Advantages of intermittent flowing and plunger lift system . . . . . . . . . . . . . . . . . . . . . . Gas well applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Predicting plunger lift performance . . . . . . . . . . . . . . Need to use tubing bottom as reference depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example of change in productivity index with depth . . . . Obtaining static and index consistent with operating Changes in operating line with changes in static and index computed from operating conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
475 475 475 476 476 476 483 484 484 484 484 485 485 486 487 488 488 488
495 496 496 497
505 505 505 507 510 510 510
512
515 515
516 517 517 517 518 518 520
XXI
Actual field example of changes in predictions from operating line . . . . . . . . . . . . . . . . . . . . Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample problems and questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
521 527 528 528 529
Chapter 14. SUCKER-ROD PUMPING D. Momeni, G.V. Chilingarian, W.B. Hatcher and Axelson
General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , Subsurface pumps . . . .................................. Evaluation and selection of pumps . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . Selection of pump bore size . . Selection of pump setting depth Selection of pump types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (1) Casing pump, 552 - (2) Tubing pump, 554 - (3) Rod pump with travelling barrel, 555 - (4) Rod pump with stationary barrel and bottom hold-down, 556 - (5) Rod pump with stationary barrel and top hold-down, 559 - (6) Rod pump with stationary barrel and top and bottom hold-down, 559 Theoretical analysis in sucker-rod design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Design of the sucker-rod string .... Example 14-1 . . . . . . . . . . . . .... Solution, 569 Rod motion analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . Example14-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solution, 570 Effective plunger stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pump-size determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example 14-3 . . .......... Solution, 575 Polished rod loads calculation Counterbalance design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Torque calculation . . . Prime mover horsepo Example14-4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solution, 582 API recommended design procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... Dynamometer cards (dynagraphs) . . . . . . Visual diagnosis of operating conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pumping efficiency determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Problem well testing . . . Energyoptimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surface efficiency . . Subsurface efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pump-off controls and timers . . . . ............... Selectionofmaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corrosion . . . . . . . . . . . . (1) Pitting and concentr corrosion, 622 - (4) Galvanic corrosion, 623 - (5) Sulfide stress cracking, 623 - (6) Corrosion fatigue, 623 - (7) Material strength, 623 - Inhibitors, 625
531 531 532 539 539 539 551 551
560 561 568 569 570 570 574 575 576 578 580 581 582 583 593 595 611 611 611 619 619 621 621 622 622
XXII Installation and operation , , , . . . . . , . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample problems and questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 14.1 - Useful formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 14.11 - Pumping unit design calculations , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
626 626 628 629 633
Chapter 15. HYDRAULIC SUBSURFACE PUMPS P. Wilson and G.V. Chlingarian Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subsurface pumps-piston type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example problem 15-1 Pump selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . , . . . . . . . . . . . Example, 665 Example problem 15-2 Given, 667 - Find, 667 - Solution, 668 Example problem 15-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Given, 669 - Find, 669 - Find, 669 - Solution, 669 Classproblems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tubing arrangements . Wellheads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Control manifolds Power fluid cleaning system
Example problem 15-4 ........................................ Given, 708 - Assume, 709 - Find, 709 - Solution, 710 Class problems . . . . . . . . . . .............. . . . ..... Bottomhole pressure calculations Sample problem and questions . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
635 638 656 660 667 669 670 670 680 680 681 681 683 683 684 685 687 693 693 708 719 720 726 736
Chapter 16. ELECTRIC SUBMERGIBLE PUMPS W.J. Powers, C. Dunbar and G.V. Chilingarian
Applications
................................... ......................................
...................... ..............................................
............................ Power cable . . . . . . . . . . . . . . . . . . . . . Motor flat cable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
737 737 738 738 740 740 743 743 743 748 749
XXIII Switchboard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transformer .............. Wellhead . . . . . . . . . . . . . . Junction box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Accessory options ..................................... - Variable speed drive, 754 Pressure-sensin Selection data and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical data . . .......................... Production data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.......................... Electric power supply . . . . . . . . . . . . . . . . .
150 752 152 153 153 756 756 758 760 160 760
Handling, installation and operation . . .
......
.....................................
175 116
Fluidpump-off . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SUBJECTINDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surface Operations in Petroleum Production, I1 Contents
Chapter 1. Introduction (and an Appendix by D.D. Coleman) by G.V. CHILINGARIAN and J.O. ROBERTSON, Jr. Chapter 2. Flow rate measurement by T.R. SIFFERMAN Chapter 3. The manufacture, chemistry and classification of oilwell cements and additives by J.O. ROBERTSON, Jr., G.V. CHILINGARIAN and S. KUMAR Chapter 4. Fracturing by J.O. ROBERTSON, Jr., G.V. CHILINGARIAN and S . KUMAR Chapter 5. Acidizing oilwells by J.O. ROBERTSON, Jr. and G.V. CHILINGARIAN Chapter 6 . Gravel packing by W.B. HATCHER, G.V. CHILINGARIAN and J.R. SOLUM
801 805
XXIV Chapter 7. Steam enhanced oil recovery by J.P. FANARITIS AND G.V. CHILINGARIAN Chapter 8. Corrosion in drilling and producing operations by T.A. BERTNESS, G.V. CHILINGARIAN and M. AL-BASSAM Chapter 9. Water quality for subsurface injection by C.C. WRIGHT and G.V. CHILINGARIAN Chapter 10. Offshore technology by S. KUMAR and G.V. CHILINGARIAN Chapter 11. Pollution control by K.M. SASSEEN, G.V. CHILINGARIAN and J.P. BRADY Chapter 12. Underground storage of gas and oil by A. ALI AZUN, G.V. CHILINGARIAN and S. KUMAR Appendix A. Technology of testing petroleum products and sample experiments by G.V. CHILINGARIAN, J.O. ROBERTSON, Jr. and C.M. BEESON Appendix B. Conversion of units by J.O. ROBERTSON, Jr. and G.V. CHILINGARIAN Reference Index Subject Index
1
Chapter I
INTRODUCTION TO SURFACE PRODUCTION EQUIPMENT K.M. SASSEEN, GEORGE V. CHILINGARIAN, AND JOHN 0. ROBERTSON JR. INTRODUCTION
During the mid 1800’s, there was a thriving salt producing industry in the U.S.A. (West Virginia and Pennsylvania), based on the evaporation of natural brines to recover salt. Crude oil was a troublesome contaminant that would often accompany the produced brine. It was skimmed off in the evaporation pools and discarded. Many enterprising salt producers, however, bottled the oil and sold it at “medicine shows”. A picture of a fierce Indian often appeared on the label, attesting to the universal curative properties of this “rock oil from the bowels of the earth”. A sample of the oil was sent to Yale University for analysis and distillation revealed the presence of some valuable properties. Based on this information, a syndicate was formed in Pennsylvania to promote the drilling of an oil well. The Drake well was drilled and completed in 1859 by a group of salt water well drillers from West Virginia, and marked the beginning of the American petroleum industry. Thus, the roles of contaminant and product have been reversed in the case of salt water and oil, which have been associated from the beginning. Since then the oil industry has progressed steadily until today (1985) the U.S.A. alone consumes around 16 million barrels of oil per day. All phases of petroleum technology have kept pace with this expansion through the never-ending search for better and more efficient methods. Production techniques have advanced from the very crude wooden troughs and pipes used in the early development of the industry to the modern complex gathering systems, staged separation, and treating plants. Transportation has evolved from the wooden barrels filled at the wellhead to a system of pipelines and tank trucks connecting all parts of the country. In this book, the writers are concerned with a phase of production classified as “surface operations in petroleum production”. This classification includes all equipment and operations from the wellhead to the refinery (Fig. 1-1). GATHERING SYSTEM
The gathering system (see Fig. 1-1) consists primarily of pipes, valves, and fittings necessary to connect the wellhead to the separation equipment. The gathering system may contain one or more lines with branches to each well, or it may consist of separate lines from each well, which are connected to a group header or test
N
I---
4. STEAM INJECTION
SrVlY GENERATOR w m WNDOIOOR FLYSCRUBBING FLYASH HANDLING
CW OIL
i
r
I AWT SKID UNIT
@ GATHERING
SYSTEM
.-
I
VAPOR RECOVERY
CLEAN OIL
SHIPPING PUMP
I
@
PIPELINE
I1
TRUCKING MANUAL -GAUGING
-
TO PIPELINE
r
r
4
-
-
STORAGE TANKS
REJECT
LACT UNIT
-@
Fig. 1-1. Schematic flow diagram of surface production equipment. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs, California.)
3
header system, as distance and distribution dictates. Accessory items might include: (1) gross-production meters; (2) automatic well test units with programmers and computer readouts communicated to remote office locations or production headquarters; (3) corrosion inhibitor and chemical injection equipment; (4) automatic routing valves; and (5) production limiting devices. In some cases, gas is removed from the casinghead into a field gas-gathering system. Some of these operate under vacuum to 28 in. of Hg. Orifice meters may be installed at each well, or one meter may be used to measure the entire gas production from a lease. Development of oilwell steam injection procedures has added to the complexity of some field gathering systems; however, new development facilities, although more complex, often tend to simplify the overall design for good continuity and operation. Oilwell steam injection facilities have been installed in a number of low-gravity oilfields to stimulate oil production. Steam stimulation, in several installations, increased the net oil production of an existing well up to ten times the production prior to steaming operations and created new well drilling activity in many existing fields. Among other details that determine the feasibility of steam operations are considerations which involve the subsequent steam stimulation of the well after the initial two or three steaming phases, in terms of a production decline to a level approaching the production prior to steaming, or a stable increased production due to the continued steam stimulation. A steam injection system primarily consists of: (1) a steam generator; (2) a water conditioning system for feedwater to the steam generator; (3) a main steam header to various oil production-steam injection manifolds that are locate! in the central part of the wells to be steamed; (4) a series of lines to each individual well; (5) a common test line from the above-mentioned manifold; (6) a purge line from the manifold; and (7) a group oil line from the manifold back to the dehydration facility. The line to each well from the oil production-steam injection manifold is often a dual purpose line (in lieu of a single steam line and a single oil line) used for steaming the well for a certain period of time. In this case, the oil production returns through the same line to the manifold, where the production is then routed to the group header. The manifold may contain twenty or more wells, each of which can be designed with automatic diverter valves in order to: (1) direct the flow of an individual well to the well test line; (2) route the flow to an automatic well test (AWT) unit, completely panel programmed; (3) return the flow to the maingroup header; and (4) after a purge period, program the next well for testing, etc. In the case of oil-burning type steam generators, ancillary equipment might include scrubber systems for the removal of fly ash and sulfur dioxide from the generator combustion gas, to control air pollution and maintain standards set by the local environmental agencies. TREATING SECTION
The treating section (see Fig. 1-1)consists of some method of dehydration, such as using washtanks, heater treaters, or electrical dehydrators. The principal purpose
4
of the treating section is to remove water, sand, and other contaminants from the oil. In most cases, the waste water must be cleaned to meet the requirements of the local water quality board. Often the water is further processed for waterflood applications or for reuse as steam generator feedwater in some locations where the water has proper chemical composition and properties. Oil enters the treating section from the separators, where it has been essentially degassed, and flows to the dehydration equipment. Dehydration may be accomplished by one or a combination of several methods ranging from simple tank settling to complex methods. In general, dehydration equipment can be divided into three classes: gravity, electrical, and chemical, or a combination thereof. Gravity dehydration
Washtanks, heater treaters, centrifuges, etc., are included in the gravity class. As implied, the principal force involved in the separation of oil and water is gravity (separation in accordance with Stokes’ law). Centrifuges add mechanical force to aid gravity settling; however, due to high cost and low capacity, for the most part, this equipment is no longer a viable consideration for processing. Washtanks A washtank (Fig. 1-2) is a large tank equipped with a spreader, oil draw-off, level control, and low-pressure separator. Oil enters the low-pressure gas separator (gas boot), situated on top of the tank, and is conducted (higher elevation head) to the bottom of the tank by means of a large-diameter balance column attached to a spreader near the bottom of the tank. The operating principle of the spreader design is to allow the oil-water flow to break up into smaller oil globules through the small vertical louvered sections of the spreader skirt and thereby cause coalescence of the minute particles, up into the oil pad with retention time being a factor. Inasmuch as heat lowers the specific gravity and viscosity of the crude oil, it plays an important part in the operation of a washtank. A level control (weir balance operation or electronic sensor) maintains the oil-water interface at a desired height, usually in the midsection of the tank. Because the spreader is located below the interface, the input oil is forced to rise vertically through a water bath before entering the oil layer. Water settles out of the oil under the force of gravity and clean oil is skimmed from the surface of the oil layer. Uniform distribution of oil is of primary importance in a washtank. Excessive channeling rapidly reduces the overall capacity and operation of this type of dehydrator. Heat is often required to reduce the viscosity of the oil to a value that will promote gravity settling. In some cases, heat applied internally creates convection currents in the water bath that may seriously interfere with proper operation of a washtank. Application of heat to the oil stream prior to entrance to the washtank will correct this condition. Injection of emulsion-breaking chemicals at the wellhead or into the oil stream at the treating section aids in the resolution of emulsions. The
5
II
0 W = OIL 80 PSI AIR ELECTRIC PROGRAMMER
WATER 6 =OAS S =BASIC SEDIMENT W =WATER C.O= CLEAN OIL ST= STEAM C =CONDENSATE
TRERYOPANE-'
Fig. 1-2.Automated wash tank. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs, California.)
detrimental effects of channeling and convection currents led to the development of a new configuration of spreaders and baffles for the washtank. In this design, the spreader takes the form of a horizontal slotted pipe instead of a circular pan. The oil is allowed to rise in a small sector of the tank area confined by a vertical baffle extending from the bottom to above the oil level. After rising vertically through the water bath, the oil enters the oil layer which flows uniformly in a horizontal circular path to an oil skimmer. The skimmer is located on the opposite side of the vertical baffle from the spreader. Segmented baffles maintain a uniform flow in the circular path. Heat is applied externally to the oil stream by a heat exchanger or inline heater. The oil-water interface is maintained at the desired level by the same means as those used in a conventional washtank, i.e., a weir box or electronic controller. Capacities of washtanks may be based on three to four barrels of clean oil per square foot of cross-sectional area per day, when the viscosity of the crude has been reduced to 100 SSU * or less. A difference in density must exist between the oil and water for proper washtank operation. Washtanks are not considered as sophisticated process equipment items, but they are (and will remain) viable dehydration units for many production facilities, where the API gravity is in the middle-low range. They are low in the initial installation cost per barrel of oil throughput; and once they are set and adjusted for the design parameters, they operate practically maintenance-free.
* Saybolt Universal seconds.
6
Heater treaters
A heater treater is a pressure vessel operating on the same basic principle as the washtank. Heater treaters may be vertical or horizontal and are often direct-fired, although indirect-heated types are available. The gravity separation principle is essentially the same as that in the washtank, the main difference being that the heater treater operates under pressure. Distribution, convection currents, viscosity, and density difference affect the heater treater operation in the same manner as they do the operation of a washtank. Similar to the washtank, a heater treater removes impurities from the produced crude oil. The most prevalent impurity is the produced water, which usually contains dissolved salts. Other impurities are sand, silt, metallic oxides, hydrogen sulphide, and various minerals. All of these constituents may deposit in the storage tanks, pipelines and, finally, in the refinery process equipment. Resulting corrosion, erosion, and plugging is detrimental to the efficiency of operation of various equipment. Because of this, the pipeline-refinery BSW (basic sediment and water) acceptability limit on crude oil shipments is 3%. The gas also must be separated from the crude oil. All treaters utilize gravity to separate the lighter from the heavier components: gas is the lightest, followed by oil and then water, with solids being generally the heaviest components. Time and gravity are the two main factors involved in the separation of various components. The closer the specific weights of the components are, the longer it will take for gravity separation to occur. In addition, oil coats and saturates the other components changing their specific weights. Coating with oil affects the weight of small water droplets more than that of the larger ones. Another major deterrent to gravity separation is high viscosity of the crude oil. The thicker the oil, the longer the duration of separation process. Whereas gravity separation is the basic method used to remove the impurities from the crude oil, it is generally necessary to apply additional treating processes to speed up the separation including: (1) chemicals to break emulsions, (2) heaters to reduce the viscosity of the oil, and (3) coalescing processes to enlarge the water droplets. Treaters incorporate some or all these means to speed up the separation process. A treater utilizes the gravity separation process, which is speeded up by enlarging the water droplets and/or reducing the viscosity of the oil, often using chemicals that can help break emulsions. All of these techniques must be applied in a systematic manner that will conserve energy, minimize costs, and accomplish the task as quickly and efficiently as possible. Operation of a typical heavy-duty thermal treater (Fig. 1-3) is based on the best principle of gravity separation by maintaining a horizontal flow pattern throughout the entire processing scheme. The constant force of flow resisting the settlement of water droplets is minimized, as opposed to treaters with vertical flow patterns. Efficient uniform heating is assured by the stream-flow distribution of the fluid around the heating elements. The stream-flow (channeled flow) heating pattern results in a maximum heating efficiency and attains positive equilibrium at the
Fig. 1-3. Superior horizontal downflow emulsion heater treater. (Courtesy of Superior Tank and Construction. Division of Rheern Manufacturing Co.. Los Angeles, California.) 4
8
oil-gas and water-oil interfaces. With the stream-flow pattern, free water is separated and bypassed around the heating elements. The crude oil emulsion is heated sequentially, with the entrained gas and water being removed at the earliest possible moment in the heating process. This insures a minimum energy loss from the heating constituents above that necessary to achieve separation. The crude oil emulsion is heated directly by the heating elements being immersed in the oil phase. Heating in the oil constant phase reduces scaling and coking of the heating element, Free and entrained gas, which is released by the heating process, is removed ahead of the coalescing and settling section, thus preventing the agit’ation that would be created by its separation in the settling area. All of the water is collected at the bottom of the treater and discharged from a single control valve at the back of the treater. This reduces the number of control valves. In addition, the water experiencing a greater residence time is discharged cleaner. The gas-free crude oil flows from the heating section into a coalescing section that contains louvered plates to achieve coalescence and insure uniform distribution throughout the settling section. The louvered plates impart a slight pressure drop that causes the oil to distribute across the full cross-section of the vessel. The baffles are designed to force a change of direction that impacts the water onto the plate. The cutting edges of the louvers break the oil film present around the water droplets, allowing the water to escape and coalesce into larger droplets that can settle more quickly. The horizontal flow reduces the resistance to separation and settlement. All this results in the delivery of the maximum volume of clean oil with the minimum expenditure for energy and chemicals. The thermal-electric treater (Fig. 1-4)combines the best principles of thermal treating with electrostatic and chemical treating. Flow pattern in the thermal-electric treater is an adaptation of the stream-flow (channeled flow) pattern discussed above. Gas removal is accomplished at the warmest part of the treater to prevent gas breakout in the coalescing section. Free water bypasses the heating element and is not heated, thereby minimizing fuel consumption. Small water droplets are coalesced in passing through a high-energy electrically charged field. Precisely engineered distribution and collection headers assure uniform flow through the electrical field. The thermal-electric treater is equipped with a single-point quick adjusting electrical grid spacing apparatus for precise electrical current regulation. The adjustment can be made while the unit is in operation as simply as closing a gate valve. The initial gas removal occurs at the top of the vessel as the emulsion enters the vessel. As the emulsion flows through the first compartment, free water is bypassed around the heating section and is not heated more than necessary to accomplish its separation. Efficient, uniform heating is assured by stream-flow distribution of the fluid around the firetubes. The uniform stream-flow (channeled flow) heating pattern provides maximum heating efficiency and equilibrium at the fluid-gas interface. Gas evolved during the heating is removed at the warmest part of the system preventing its breakout later in the process. The closely controlled sequential
GAS EQUALIZER
GAS CONDUIT OUTLET DISTRIBUTOR
REAMFLOW BAFFLES
SURGE SECTEN
N DISTRIBUTOR
Fig. 1-4. Thermal electric treater. (Courtesy of Superior Tank and Construction. Division of Rheem Manufacturing Co., L~~ ~
~California.) ~
~
l \D
~
~
10
heating assures heating only to the point that is required to attain the desired amount of separation. The gas-free oil first flows through a surge section and then into the electrical coalescing section. Presence of a system of distribution spreaders results in uniform flow through both the length and width of the vessel. An overhead, clean oil collection header system gives rise to a uniform outflow of clean oil. This complements the bottom distributor system to provide minimum velocities and an even distribution throughout the coalescing section. The tendency for fluids to channel is virtually eliminated, and good distribution assures effective coalescing and maximum capacity. Coalescing of the small water drops dispersed in the oil is accomplished by the high-voltage alternating electrical field. As the emulsion rises through the field, the water droplets acquire an electrical charge. As a result, they rapidly move about repelling, attracting and colliding with one another. The action is energetic and ADJUSTABLE WATER SIPHON\
1
SAFETY REULF
r""
our
Fig. 1-5. Superior vertical emulsion heater treater. (Courtesy of Superior Tank and Construction, Division of Rheem Manufacturing Co.. Los Angeles, California.)
effective because all the water droplets acquire a charge, regardless of size. The droplets collide with sufficient energy to overcome the emulsifying forces, combining into larger drops. This growth in mass allows fast gravity settling of the larger drops into the water phase. Vertical emulsion heater treaters have been successfully employed with the higher-gravity crude oils ( > 30" API) in some areas (Fig. 1-5). Electrical dehydrators The operation of electrical dehydrators is based on the well-known principle of Cottrell. The oil-water emulsion is heated to reduce viscosity and is then exposed to a high-voltage alternating electric current field. Inasmuch as the water particles are charged, the alternating electric field increases the random motion or displacement, thus aiding the coalescence of the small water particles. Gravity separation occurs when the small water particles coalesce into large drops. Chemicals may be added to aid in emulsion resolution. Cleaning costs are usually greater with this type of equipment than they are with the gravity settling type. (See Fig. 1-4.) Chemical dehydration Chemical treatment usually is used in combination with one of the gravity settling class of equipment. In the case of stable emulsion, conditions at the interfacial film can be altered to produce equilibrium with applied stress. If the equilibrium is upset by any change in conditions or stresses that occur at the boundary, the film will collapse and the emulsion will become unstable. In principle, chemical treatment through the addition of surface-active agents, alters the chemical composition at the interfacial film to such an extent, that the emulsion becomes unstable. Heat produces stress on the film that further renders the emulsion unstable. STORAGE TANKS
Storage tanks are usually of bolted construction up to the 10,000-barrel size. Welded steel tanks are used extensively in the larger sizes. In some fields, small welded steel tanks in bolted tank sizes up to 500 bbl are used. API Specifications 12B and 6 5 0 cover bolted and welded steel tanks, respectively, whereas API 12G describes welded aluminum tanks. Bolted tanks have the advantage of being easily transported and relocated. Bolted steel tank components can also be factory sandblasted and precoated with various epoxy coatings to assure quality control. ACCESSORY EQUIPMENT
Accessory equipment includes the equipment that is not basically necessary to convey the oil from the wellhead to the pipeline or other means of transportation.
12
Vapor recovery, water treatment, and automatic custody transfer are included in this group. Vapor recovery
Tank vapors were allowed to vent to the atmosphere in the past. With the introduction of gasoline plants and the ever-increasing demand for liquefied petroleum gas (LPG) and natural gasoline, these vapors became an important source of revenue. Initial pay-out periods of a few months are not uncommon (see Chapter 6 ) . A recovery system consists of a network of piping to collect tank vapors, a control system to maintain constant pressure on the tanks, and a compressor to increase the pressure of the vapors from atmospheric to the gas-gathering line pressure (Fig. 1-6). Large-diameter, thin-wall pipe is used to collect the vapors because very large volumes at low pressures are involved. In most cases, the total pressure differential cannot exceed 0.5 in. of water. The pressure control system uses two very sensitive regulators, one to remove excess vapors from the tanks and one to admit natural gas for repressuring during demand periods. A pressure control system of this type can maintain the tank pressures within 0.1 in. of water. A slight positive pressure is maintained in the tanks to prevent the admission of air. The compression section can consist of a blower or compressor, depending upon the gasline pressure. In some cases, the gasline is under vacuum, eliminating the compression section entirely. The blower or compressor can operate continuously, using an unloader controller to load or unload the compressor on demand condi-
VAPOR HEADER
TO SHIPPING PUMP
&*
H
GAS
SCRVBBER
PRESSURE TRAP
AIR COOLED COHPRESSOQ
A
;j TO GAS SYSTEM
STACK
VAPOR RECOVERY COMPRESSOR SKID
Fig. 1-6.Vapor recovery flow sheet. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs, California.)
13
tions. A meter is often included to measure the volume of vapors discharged into the gasline. The writers recommend thorough investigation of vapor recovery possibilities on all tank farms. Many small tank batteries have yielded important revenues after installation of vapor recovery systems, because the storage tanks constitute the final stage of gas-oil separation with a high liquid content in the gas. State air-quality boards in the U.S.A. and other countries have issued many regulations regarding the requirements for vapor recovery installations on tank farms. Wastewater treatment
Disposal of wastewater is a very important problem. In order to prevent contamination of domestic water supplies, wastewater cannot be discharged into unlined sumps or streams in many places. Discharge of wastewater into sewers or into the ocean without proper treatment is prohibited by state and local authorities in the U.S.A. and in many other countries. Treatment consists of reducing the oil and sediment content of the wastewater to the established limits. In some cases, chemical treatment is required for control of basic oxygen demand (BOD), microorganisms, and certain gases. The flotation cell systems are most successful when it is required to clean wastewater to a higher degree than the usual gravity settling basins permit, especially for waterflood systems where the effluent from the flotation cell is processed through filters prior to water injection (Figs. 1-7 and 1-8). In waterflooding, the treatment of wastewater is very critical and complex. The water not only must be thoroughly cleaned of oil and sediment, but also must be chemically stabilized to prevent scale formation. Precautions must be exercised to prevent any type of deposition in the injection well.
I
R E C O M I I I O 6 f w m w m n WILLWIT
Fig. 1-7. Fully pressurized flotation cell: wastewater system (P and I flow diagram). (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs, California.)
14
Fig. 1-8. Dual flotation cell: wastewater system. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs. California.).
In the flotation cell system, air or gas injection is utilized upstream of the main process pump, The air or gas together with a chemical, if required, is thoroughly mixed inside the pump. The fluid discharged from the pump enters the retention tank where bubbles are collapsed under 2-3 atm of pressure and are driven into true solution. As the water enters the flotation cell at this point, the air comes out of solution due to the pressure decrease to, i.e., atmospheric. Thus, small sludge and oil particles become floatable and move to the top where the rotating skimming arm sweeps the oil sludge into a compartment for removal. The same drive shaft also rotates a bottom grit scraper arm for the separate removal of settled solids from the grit collecting box. Clean water effluent is either directed to the outlet system or recycled back to a surge tank, on low incoming flow, as the main process pump runs continuously. Lease automatic custody transfer (LACT)
Automatic custody transfer simply means automatic gauging, sampling, and shipping of oil from the lease tank farm to the pipeline (Fig. 1-9). Two methods are in current use: the meter type and the volumetric dump type. Other basic equipment, such as automatic cut determination device, temperature recorder, and sampling device are essentially standard for both types. In both methods, oil from storage is pumped through the water content (water cut) determination device where the oil is either routed on through the system or bypassed back to the dehydration equipment,
15
TOPVILW
h
l l 0 L VIEW
I -PUMP 1-STRAINER
3-DIVERT VALVE 4 - A I R ELIWINAfOR
5 -METER 6-SAMPLER 7 -PROVER MANIFOLD 8 -BACK PRESSURE REGULITOR
Fig. 1-9. LACT unit. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa’Fe Springs, California.)
depending on the water content of the oil. The oil flows to a deaerator and then through the meter where the volume is determined. A temperature compensation device corrects the meter with an electrically driven sampler. The electrically driven sampler can be actuated by a pulse from the meter. Float switches on the storage tank start and stop the unit as required. The volumetric dump type unit operates in the same manner as the meter unit, except that two calibrated tanks are alternately filled and dumped, each dump registering on a counter. Both types lend themselves readily to telemetering results to control centers. Data, such as temperature, pressure, water cut, and volumetric rate of flow, are easily transmitted and recorded at remote locations. Installation of LACT, where applicable, results in a saving in labor cost and a uniform scheduling for pipeline operation. This type of product handling is replacing the old manual gauging, sampling, and shipping methods. Remote-reading tank level gauges should be included in this section. Gauges are available to determine, record, and transmit, for remote reading, the level in any selected tank of a battery. SEPARATORS
Oil and gas separators are used in most petroleum production operations. The primary function of these separators is to produce gas-free liquid and liquid-free gas. Basically there are three types of separators: vertical, horizontal, and spherical. Each type has proven features particular to various field applications. It is often recommended to use the horizontal separator where the gas/oil ratio is high; the
16
vertical separator where the gas/oil ratio is low; and the spherical separator for an intermediate range of the gas/oil ratios. Integrated with the natural forces utilized to induce mechanical separation, the interface area between the liquid and gas is considered an important factor for oil and gas separation. For extremely high gas/oil ratios, therefore, the horizontal separator provides the maximum interface area. Some designs of vertical and spherical separators include a secondary interface chamber in the lower oil reservoir of the separator. The interior of a separator is divided into three compartments: (1) the primary separator chamber in the middle (2) the secondary separation chamber at the top, and (3) the oil reservoir at the bottom. The standard oil and gas separator accessory components are as follows: (1) a back-pressure gas regulator, (2) an oil-level control valve (diaphragm-operated or mechanical float linkage dump type), (3) a pressure gauge assembly, (4) a gauge glass assembly, ( 5 ) a float flange assembly, and (6) a safety relief valve. It is often advisable.to include a flanged safety disc head with a pressure setting higher than the relief valve to protect against high overpressuring. In addition to the various types of stationary separators for field processing, there are portable test separators (complete with gas and oil meter runs) for preliminary tests at the wellsite. Composite volumetric metering type separators are also available. The criterion for the design of oil and gas separators and separation systems is the economical separation of the gaseous and liquid phases of crude oil and distillate, contingent on the characteristics and conditions of any particular oilfield. The factors which affect oil and gas separation and stage separation are wellhead or first-stage pressure, temperature, wellstream composition, gas/oil ratio, and gravity, all of which are relative and present variable conditions of operation. Stage separation processing is utilized to separate high-pressure wellstream gas-oil mixtures into gas and liquid phases by two or more equilibrium flashes at consecutively lower pressures. The washtank or stocktank is considered as one stage in the separation process. Staging is necessary to increase washtank or stocktank recovery and to remove the bulk of vapors formed when the pressure is decreased. This precludes the entry and evolution of large quantities of vapor into the tank, which would cause a “rolling” action and be detrimental to the efficiency of operation, especially in a washtank. Operating conditions are established to separate the gas from the wellstream at the optimum pressure and temperature in order to induce efficient mechanical separation within the separator and thereby produce a more stabilized flow of crude oil to the washtank or treater. Separators are designed to utilize natural forces and conditions in order to facilitate the mechanical separation of oil and gas. These include: (1) centrifugal force-a mixed stream of oil and gas is subjected to a whirling motion and the heavier oil is thereby forced to the outside and away from the lighter gas; (2) high velocity-the amount of centrifugal force depends upon the velocity of the whirling stream; (3) gas expansion-decreases the gas density and thus induces the heavier oil particles to fall out; and (4) impingement
17
+
GAS OUTLET
IN IT IA L SE PARATlO N
SECONDARY
FLUID INLET
/QUIESCENT
ZONE
LIQUID OUTLET
VORTEX BREAKER
Fig. 1-10, Vertical oil and gas separator. (Courtesy of Superior Tank and Construction, Division of Rheem Manufacturing Co., Los Angeles, California.)
contact-scrubbing and removal of oil mist from the gas flow near the gas outlet, utiIizing surface contact and directional change in flow. In order to accomplish good mechanical separation, the separator should perform
INITIAL SEPARATION SECONDARY SEPARATION MIST EXTRACTION
I
I
WAVE
/
/ U BREAKER W R T E X BREAKER
LIQUID ACCUMULATION
Fig. 1-11. Horizontal oil and gas separator. (Courtesy of Superior Tank and Construction, Division of Rheem Manufacturing Co., Los Angeles, California.)
18
* INLET
MIST EXTRACT I ON,
INITIAL SEPARATION
SECONDARY SEPARATION
LIQUID OUTLET QUIESCENT
ZONE VORTEX
BREAKER
C - - G I \ S OUTLET
Fig. 1-12. Spherical oil and gas separator. (Courtesy of Superior Tank and Construction, Division of Rheem Manufacturing Co., Los Angeles, California.)
the following four basic functions, utilizing the above-mentioned phenomena (Figs. 1-10, 1-11, and 1-12). (1) Initial separation for diverting the bulk of the free liquid immediately, by specially designed tangential inlets in vertical and spherical vessels, and deflection plates in horizontal vessels. (2) Secondary separation for removing the small liquid droplets. The principles involved are reduction in velocity and gravity settling. The vessel diameter is generally established by this phase of the separation process. Separators incorporate scientifically designed internal components to impart laminar flow and to insure utilization of the full cross-sectional area. The science of particle dynamics must be thoroughly understood and properly utilized in the design. (3) Mist extraction for removing the mist-like liquid particles that are tightly entrained in the gas stream. Generally, impingement-type devices are utilized in this section of the oil and gas separator. The principles of surface contact, adhesion, and agglomeration are applied in the design of mist eliminators. (4) Collection and disposition of the liquids in a manner that will not expose the liquids to re-entrainment. In order to accomplish this, a large quiescent zone is provided, which is removed from the high-velocity gas flow. Baffles are placed in such a manner that a “surf spray” effect on the liquid surface is prevented. Gas from separators is metered and transferred to sales gasline systems or process plants. It csn also be further processed for field gas injection systems.
GAS PROCESSING AND CONDITIONING
The two basic reasons for processing or conditioning natural gas are: (1) removal of impurities that could cause problems in transportation, distribution, and final
19
TABLE 1-1 Removal processes for hydrogen sulfide Name
Reaction
Girbotol a Phenolate a Phosphate a Sodium carbonate (vacuum) Seaboard Lime Iron oxide Caustic soda (pH = 9.5) Caustic soda (pH = 7.0) IronitemSponge"
2RNH2 + H2S (RNH3)ZS NaOC6H, H,S + NaHS + C 6 H , 0 H K 3 P 0 4 +H,S+KHS+K,HPO, Na,CO, + H,S * NaHCO, +NaHS Na,CO, + H,S --L NaHCO, +NaHS Ca(OH), + H , S + CaS+2H20 FeO + H, S 4 FeS + H,O
a
Regenerated by steaming.
+
*
2NaOH+H,S+Na2S+2H,0 N a O H + H , S + NaHS+H,O Fe,04+4H,S-. 3 F e S + 4 H 2 0 + S environment FeS + S + FeS,
Regenerated by vacuum steaming.
Regenerated by air blowing.
use; and/or (2) the modification of the characteristics of a natural gas to achieve the most efficient utilization. The most common impurities are carbon dioxide (CO,), hydrogen sulfide (H2S), and water (H,O). The latter is present as a liquid and a vapor. First, the liquid water can generally be removed in a separator or scrubber which is an integral part of the gas processing system. The content of water vapor is then reduced by bringing the gas into contact with a solid or liquid desiccant. Hydrogen sulfide (H2S) can be removed by many methods (Table 1-1). In chemical adsorption processes, the H,S reacts chemically (combines) with a liquid absorbent. Regeneration of the absorbent is accomplished by: (1) the addition of heat to reverse the chemical reaction, or (2) oxidation of hydrogen sulfide to elemental sulfur by air. The removal of CO, also can be accomplished by chemical absorption. Water vapor can be removed from natural gas by the glycol absorption dehydration process (GAD) (Fig. 1-13). Gas to be dehydrated enters the absorber (A) near the bottom and rises through a series of bubble trays where it is contacted with glycol. The glycol absorbs the water vapor from the gas and dehydrated gas is discharged from the absorber. As the glycol travels downward through the absorber, its water content increases. The water-rich glycol is discharged from the bottom of the absorber through heat exchangers (B) to the regenerator fractionation-distillation column (C). The water-rich glycol is heated in the reboiler (D), causing the water to be released in a vapor form. The reconcentrated glycol flows to the glycol storage tank (E) where it exchanges heat with the water-rich glycol in the heat exchanger. The reconcentrated glycol flows through an additional glycol-to-glycol
20
FLOW DIAGRAM : GAD-GLYCOL ABSORPTION DEHYDRATION (GAS DEHYDRATION UNIT)
Fig. 1-13. The glycol absorption-dehydrationprocess. (Courtesy of Superior Tank and Construction, Division of Rheem Manufacturing Co., Los Angeles, California.)
Fig. 1-14. Automated treating facilities. (Courtesy of HTI Superior, Inc., a Berry Industries Company, Santa Fe Springs, California.)
heat exchanger (F) to the glycol pump (G). Then it is pumped through a gas to glycol heat exchanger (H) and into the absorber. The efficient application of heat exchange and conservative heat flux rates assure exceptional fuel economy and high overall performances. A summarized, overall view of the surface production equipment is presented in Fig. 1-14. SAMPLE PROBLEMS
Exchanger problem
Crude oil is heated in exchangers by the counter-current flow of fuel oil: Measured at 60°F, the exchangers handle 15,500 gal/hr of the crude and 8350 gal/hr of the fuel oil. The gravity and UOP characterization factor of the crude are 40" API and 11.3, whereas those for the fuel oil are 25" API and 11.3, respectively. The fuel oil is cooled from 675" to 225"F, and the crude enters the exchangers at 100°F. Compute the heat lost by the fuel oil in Btu/hr. Estimate the outlet temperature of the crude oil, assuming all of the heat lost by the fuel oil is picked up by crude oil. T, = 675
*
O F
0, =
T2 = 2 2 5 "F
8 3 5 0 gal / h
Kw=11.3
25O A P I
G=?
4
6'2 = 15.500 gol/h Kw = 11.3 40OAPI
T3 = 100 'F
Scheme 1-1.
Solution: Heat lost by the fuel oil in Btu/hr = gal/hr x lb/gal X Btu/"F X AT X C = 8360 X 7.529 x 0.650 x (675 - 225) X 0.975 = 17.95 X lo6 Btu/hr. Specific heat, cp: in Btu/"F, and correction factor, C , for UOP characterization factor can be obtained from charts provided by Nelson (1949), for example Fig. 1-15. Assuming T4 = 350"F, the heat gained by the crude oil in Btu/hr is equal to: 15,500 x 6.870 x 0.625 x (350 - 100) X 0.975 = 13.7 x lo6 Btu/hr. Thus, the outlet temperature T4 must be greater than 350°F. Assuming T4 = 400"F, the heat gained by the crude oil is equal to: 15,500 X 6.870 x 0.570 x (400 - 100) x 0.975 = 17.7 x lo6 Btu/hr. Assuming T4 = 450"F, the heat gained by the crude oil is equal to: 15,500 X 6.870 x 0.580 x (450 - 100) X 0.975 = 21.0 X lo6 Btu/hr. From Fig. 1-16, the correct outlet temperature T4 is equal to 403°F. See p. 47, Chapter 2, for definition of characterization factor.
22 1.0
2 0.9 k
3 00 I-
2
07
W
L
0 0.6 k 0.5 fn
0.4
a3
200
0
400
600
000
1000
TEMPERATURE, OF
Fig. 1-15. Specific heats of Mid-Continent liquid oils with a correction ctor for other bases of oils. (After Nelson, 1949, fig. 16, p. 136; courtesy of McGraw-Hill Book Company, Inc., New York, N . Y . )
350 10
12
14
16
18
20
22
HEAT GAINED, Btu/hr x
Fig. 1-16. Solution for exchanger problem.
GROSS PRODUCTION
60 PSIG STEAM CONDENSATE WATER
Fig. 1-17. Diagram for washtank problem.
Washtank problem It is required to design a heating system for an oilfield production process dehydration washtank (Fig. 1-17). Field data: (1) 1800 bbl/day of 12" API oil and 500 bbl/day of water at 110°F inlet temperature. (2) Tank shell insulated with 2 in. fiberglass; 2250 sq ft shell surface; 28 Btu/sq ft/hr radiation loss; 60°F air temperature. (3) Each internal heating pane has 42 sq ft of surface; overall coefficient of heat transfer is 35. (4) 60-psig steam is available for heating. ( 5 ) The required washtank dehydration operating temperature is 200°F. Required: (a) Calculate the washtank heat load. (b) Find the logarithmic mean temperature difference (see Nelson, 1949, p. 478). (c) Find the number of heat panes required for the heat load. (d) Calculate the condensate return load, lb/hr. (e) How much gas (1000 Btu/cu ft) will be required for the heating system if the steam boiler efficiency is 70%. ( f ) In lieu of gas, how much fuel oil (145,000 Btu/gal) would be utilized if the efficiency is 62%.
APPENDIX 1.1-SOME PROBLEMS
FUNDAMENTAL FLUID MECHANICS CONCEPTS AND SAMPLE
Fundamental equation of fluid statics The fundamental equation of fluid statics states that pressure increases with depth, the increment per unit length being equal to the weight per unit volume (Binder, 1962, p. 13): d p = - p g dz
(1.I-1)
where d p is increment in pressure; dz is increment in depth (L is a vertical distance measured positively in the direction of decreasing pressure); p is density (mass per unit volume); and g is gravitational acceleration. The minus sign indicates that pressure decreases with increasing z. The above relationship can be clearly understood on examining Fig. 1.1-1, which shows vertical forces on the infinitesimal element in the body of a static fluid. In this figure, dA represents an infinitesimal cross-sectional area, p is the pressure on the top surface of the element and
24
BODY OF A FLUID Fig. 1.1-1. Schematic diagram of vertical forces on an infinitesimal element in body- of any fluid. (Modified after Binder, 1962, fig. 2-2, p. 13.)
+ d p ) is the pressure on the bottom surface. Inasmuch as the pressure is due to the fluid weight, the weight of the element ( p g d z d A) is balanced by the force due to pressure difference ( d p dA):
(p
d p d A = -pg d z dA
(1.I-2)
or : d p = -pg dz
In integral form, the above equation can be expressed as follows (see Fig. 1.1-1):
[‘%=
- i 2 d z = -(z2.-zl)
(1.I-3)
If p is assumed to be constant, eq. 1.1-3 becomes: (1.1-4)
or : Ap = yh
(1.I-5)
where h is the difference in depth between two points, which is commonly referred to as the “pressure head”; and y ( = p g ) is the specific weight. On expressing y in lb/cu ft and h in ft, pressure difference Ap is found in lb/sq ft.
25
Buoyancy
When a body is completely or partly immersed in a static fluid, there is an upward vertical buoyant force on this body equal in magnitude to the weight of displaced fluid. This force is a resultant of all forces acting on the body by the fluid. The pressure is greater on the parts of the body more deeply immersed. The pressures at different points on the immersed body are independent of the body material. For example, if the same fluid is substituted for the immersed body, this fluid will remain at rest. This means that the buoyant, upward force on the substituted fluid is equal to its weight. If the immersed body is in static equilibrium, the buoyant force and the weight of the body are equal in magnitude and opposite in direction, passing through the center of gravity of the body. For a comprehensive treatment of fluid statics, the reader is referred to an excellent book on fluid mechanics by Binder (1962). General energy equation
The heat added to unit weight of the flowing
The work transferred to (done upon) unit
trance (1) and exit ( 2 ) .
q+---P I U , 778
p2u2 + W = u 2 - u1 + v2’- v2 z2 - Zl 778 778 2g(778) + 778
The total gain in energy by unit weight of the fluid between entrance and exit.
(1.1-6)
where p = pressure in psfa; u = specific volume in ft3/lb; V = velocity in ft/sec; 2 = potential head in ft; q = heat transferred to fluid; p 1 u 1 / 7 7 8= external work in pushing 1 Ib of fluid across the entrance; W = work in ft-lb per lb fluid flowing; u2 - u1 = gain in internal energy; [( V? - V : ) / 2 g ( 7 7 8 ) ] = gain in kinetic energy; and ( Z , - Z , ) / 7 7 8 = gain in potential energy. Point 1 = entrance; point 2 = exit; 1 Btu = 778 ft-lb; u2 - u1 = c,(T, - T,); c, = specific heat at constant volume. Inasmuch as enthalpy = h = u + ( p u ) / 7 7 8 , eq. 1.1-6 becomes:
v;- v; W q + -= h , - h , + 778 2g(778)
+-z,-7782,
(1.I-7)
where h , - h , = cp(T,- Tl); cp = specific heat at constant pressure. For a number of cases the process is adiabatic and change in internal energy is negligible. Thus: PlU,
P2U2
---+-= 778 778
w 778
v2’- v: +-Z 2 - Z , 778
2g(778)
(1.I-8)
26
and each term in the latter equation is in Btu/lb fluid flowing. On multiplying through by 778: (1.I-9) where y = specific weight in lb/ft3 ( l / v ) ; p/y = pressure head in ft; V2/2g velocity head in ft; and 2 = potential head in ft. For frictionless compressible fluid with no work done:
vz’+ Z, = 2 P +v: + Z, = const. fi + Y2
2g
Y1
2g
=
(1.I-10)
which is the well-known Bernoulli’s equation. Derivation of formula for flow through orifice meter
A schematic diagram of incompressible fluid flow through an orifice meter is presented in Fig. 1.1-2. For an ideal flow with no friction losses the following relation will hold true:
Fig. 1.1-2. Schematic diagram of an orifice meter.
+ Pl/Y
v:/2g
+ z,= V2Wg +P,/Y + z,
(1.1-11)
where V = velocity in ft/sec; p = pressure in psfa; y = specific weight in lb/cu ft; and Z = potential head above any datum plane in ft. Inasmuch as volumetric rate of flow (in cu ft/sec) Q = V I A ,= V 2 A 2 : (1.I-12)
V, = V,A,/Al
Substituting eq. 1.1-12 in eq. 1.1-11 and solving for V2: V2=
[
1
2g( Pl/Y - P2/Y + Zl - z2> 1 - (A2/-41)2
1’2
(1.1-13)
27
For an actual flow one has to introduce correction factor for velocity (C,) and correction factor for area (Cc).The latter is termed coefficient of contraction and is equal to A , / A . Thus:
Q = C,C,V,A
(1.1-14)
The term discharge coefficient (C or C,) often is substituted for C,Cc. Another term flow coefficient ( K ) is defined as: K
=
C / [1- ( A , / A , ) , ]
(1.1-15)
1/2
Thus: actual Q = K A [ 2 g ( p I / y - p 2 / y
+ Z, - Z, )]
(1.I-16)
1/2
If A h is manometer deflection in in. of Hg, then: P I / Y -k zl - P 2 / Y - z2
Ah
=
12(sP grHg - sP grf)/sP
grf
(1.1-17)
where sp gr, = specific gravity of fluid flowing. Flow equation for the Venturi meter (Fig. 1.1-3) can be derived similarly; however, C, = 1 in this case, INLET
1
THROAT
Fig. 1.1-3. Schematic diagram of a venturi meter.
Compressible flow formula
For a compressible flow, one can derive the following equation starting with the general energy equation (also see Binder, 1962): (1.148)
28
where G = weight rate of flow in lb/sec, and k = (specific heat at constant pressure)/(specific heat at constant volume) = c,/c,,. As shown in Nelson (1958, p. 211), constant k can be obtained for various hydrocarbons. Example problem 1. Maximum reliable flow
Two reservoirs shown below are connected by a 4-in. 10,000-ft long pipe having friction factor of 0.02. Determine: (1) pump horsepower required to maintain a flow rate of 0.33 cu ft/sec of water ( y = 62.4 lb/cu ft); and (2) the maximum distance x for dependable (reliable) flow.
DATUM PLANE
Fig. 1.1-4.Diagram for example problem 1
Solution: (1) One can use Bernoulli's equation between points 1 and 3: Pl/Y + v / 2 g
+ z 1 + E , = P 3 / Y + G2/2g + z3 + A,, + A,, + h,,
where E, = energy output of the pump, ft-lb/lb of fluid flowing; h,, = head loss due to friction =f(l/d)(V,2/2g), ft-lb/lb; h,, = head loss due to entrance = 0.5V;/2g in the case of sharp entrance, ft-lb/lb; h,, = head loss due to the exit = dissipated kinetic energy ( = V,2/2g); Vp = velocity in the pipe, ft/sec; d = inside diameter of the pipe, ft; I = length of the pipe, ft; y = specific weight of the flowing fluid, lb/ft3; g = gravitational acceleration, ft/sec2; and z = elevation above some datum plane, ft. Inasmuch as velocities at the surface of two reservoirs ( V , and V 3 ) can be considered negligible and pressures p1 and p 2 are atmospheric (0 gage), the above equation reduces to:
E,
= ( z 3 - z,)
+ h,, + h,, + h ,
= ( z 3 - Z,
) + V,2/2g( f l / d
+ 0.5 + 1)
29
Inasmuch as: V, = Q / A = 0.33 cu ft/sec/(~(4/12)~/4) = 3.78 ft/sec, E , = (325 - 175) 3.782/64.4[(0.02)(10000)/(4/12) + 0.5 + l.O)] = 285 ft-lb/lb. Thus, horsepower of the pump is equal to: H P = QyE,/550 = (0.33)(62.4)(285)/550 = 10.6, where 550 ft-lb/sec = 1 H P . (2) For maximum and yet reliable flow of water (i.e., no cavitation), the pressure at the inlet side of the pump ( p 2 ) should be 2/3 of the barometric head of water. With safety factor incorporated, it is equal to - 21 f t of water (= p 2 / y ) . Thus, using Bernoulli’s equation between points 1 and 2: p I / y + V:/2g + z , = p 2 / y V,2/2g z 2 + hi, + A,,, one can solve for unknown distance x , inasmuch as terms p I / y and V:/2g can be neglected. Thus: 175 = -21 + (3.78)2/64.4 + 100 0.02[(~)/(4/12)] (3.782/64.4) + 0.5(3.78)2/64.4 and solving for x : x = 7180 ft.
+
+
+
+
Example problem 2. Compressible flow (nozzle)
A convergent-divergent nozzle is connected to a tank with air, having pressure of 100 psia and temperature of 100’F. The tip diameter (point 3) is equal to two inches, and air discharges to atmosphere ( p i = 14.7 psi and Ti = 60’F). Determine throat diameter (point 2) necessary to maintain maximum flow rate through this nozzle. Adiabatic constant k for air is equal to 1.4. (See Fig. 1.1-5.)
Fig. 1.1-5. Diagram for example problem 2.
Solution: For maximum flow rate, the velocity in the throat must be sonic, because maximum velocity attainable in a convergent nozzle is sonic. Inasmuch as p ; / p l ( = 14.7/100 = 0.147) is less than ( p2/pl)criticd (= {2/(k + l)}k’k-l) = {2/(1.4 1)}1,4’1,4-1= 0.528), velocity V, in the divergent passage will be supersonic.
+
30
T o attain sonic velocity in the throat (point 2), pressure p 2 must be critical: p , = p 1 ( 2 / k l ) k / k -= l 100 x 0.528 = 52.8 psia. The specific weight of air in the tank, yl, is equal to: y1 = p l / R T l = 100 X 144/53.3 X 560 = 0.483 lb/ft3, where gas constant for air, R , is equal to 53.3 and Tl is the absolute temperature in O R ( = OF + 460). Inasmuch as y2/y1 = ( p , / ~ ~ ) 'y2/ ~ = ,(0.528)1/',4X 0.483 = 0.3 lb/cu ft. In order to attain sonic velocity in the throat, temperature in the throat must be critical: T, = T 1 ( 2 / k+ 1) = 560(2/2.4) = 466"R. Thus, velocity V, is equal to V,: V2= V , = c2 = (kgRT,)'/, = (1.4 X 32.2 X 53.3 X 466)1/2= 1060 ft/sec. Temperature at point 3 can be determined from the following equation *: T, = T , ( p , / p l ) k - l / k= 560(0.147)04/1.4 = 320"R, and y3 is equal to: yj = p , / R T , = 14.7 X 144/53.3 x 320 = 0.12 lb/ft3. Velocity at point 3 can be determined on using the following equation:
+
K2 = 64.4 x (100 x 144/0.483)(1.4/0.4)[1 - ( 0 . 1 4 7 ) ~ . ~ / ' . ~ ] Solving for V,: V, = 1700 ft/sec, i.e., supersonic speed. Inasmuch as for adiabatic flow the weight rate of flow in the throat ( W 2 )is equal to the weight rate of flow at the exit (W,): W, = A,V,y, = W, = A,V3y,
(77d;/4 x 144)(1060)(0.3)= ( ~ 2 , / 4x 144)(1700)(0.12), one can solve for throat diameter d,: d, = 1.161 inches. Example problem 3: Compressor problem Air at standard conditions is handled at a rate of 1000 lb/hr by a compressor. Cross-sectional area of inlet is 0.6 f t 2 and that of outlet is 0.11 ft2. Air is compressed to 100 psia and 180"F, and the heat taken from air is 50,000 Btu/hr; cp = 0.239. If the change in elevation is negligible, what is the work done on the air?
31
Solution:
Weight rate of flow: G = A V y lb/sec
where A = cross-sectional area in ft2; V = velocity in ft/sec; and y in lb/ft3.
= specific weight
G = A,T/,Y, = A2V2Y2 V---G - AIY,
1000/3600 (0.60) x (0.07651)
= 6.06
ft/sec
where 0.07651 is the specific weight of standard sea-level air (59'F and 14.7 psia).
=
P2 a
V 2 = - -G A2y2
q+
1oo(144) (53.3) X (640)
-
--
= 0.421 lb/ft3
1000/3600 = 5.97 ft/sec (0.11) X (0.421)
W
=h, - h, + 778
h2 - h,
778
=
Vz'- V: 2gJ
= 0.239(180 - 59) = 29
-50+29+
Btu/lb
(5.97)2 - (6.06)2 (64.4) X (778)
W = 61,600 ft-lb/lb If the answer is desired in HP then one has to use the following equation:
HP =
( W ft-lb/lb)
( G Ib/sec) (550 ft-lb/sec)/HP X
32 APPENDIX 1.11-HYDROCARBONS: PRODUCTS
COMPOSITION OF CRUDE OIL AND PETROLEUM
Introduction
The word “petroleum” comes from the Greek work petra meaning rock and the Latin work oleum meaning oil; thus it literally means rock oil or oil coming from rock. Crude petroleum is composed chiefly of hydrocarbons (compounds containing only carbon and hydrogen, Fig. 1.11-1) together with small amounts of compounds containing sulfur, nitrogen, and oxygen. Hundreds of analyses of samples of crude petroleum from all over the world indicate the range in elemental composition presented in Table 1.11-1. Inasmuch as the hydrocarbons form the bulk of the chemical compounds in petroleum, some simple facts concerning their structure must be understood in order to intelligently interpret what takes place when these materials are broken down (as in the cracking operations), as well as to interpret the physical tests made on the products manufactured from petroleum. For various reasons one must rely chiefly on the physical properties to describe or identify the materials made from petroleum. Hydrocarbons are grouped into families according to the manner in which the carbon and hydrogen are held together in the compound. The principal groups of hydrocarbons considered here are straight and branched chain saturated (paraffinic), unsaturated (olefins and diolefins), aromatic, and naphthenic. Saturated hydrocarbons
Chemistry makes frequent use of the terms “saturated” and “unsaturated’. Saturated means full, cannot take up any more, such as a sponge full of water cannot hold any more water. The prefix “un” means not; so unsaturated means not full, consequently, capable of taking up some more. Carbon has the property of holding four other univalent atoms. When an atom of carbon has combined with
TABLE 1.11-1 Elemental composition of crude peiroleum Element
%
Carbon Hydrogen Sulfur Nitrogen Oxygen
83-87 11-14 0.05-2 0.1-2 none-2
33
HYOROCARBONS ARE MOLECULES C O N S I S T I N G OF HYDROGEN AN0 CARBON. THE CARBON ATOn HAS FOUR BONDS THAT CAN U N I T E W I T H E I T H E R ONE OR H M I E CARBON ATOnS MI W I T H ATOnS OF ANOTHER ELEMENT. A HYDROGEN A T W I HAS ONLY ONE BOND AND CAN NEVER UNITE W I T H MORE THAN ONE OTHER ATOn. .
Fig. 1.11-1. What are hydrocarbons?
four atoms of hydrogen it cannot combine with any more, hence the compound would be called a saturated hydrocarbon. This compound is methane, a gas, which is always present in gas coming from oil wells and has a chemical formula of CH,. If, on the other hand, only two of three atoms of hydrogen are combined with one carbon atom the products are unsaturated hydrocarbons. The general formula for saturated (paraffin) hydrocarbons is CnHzn+*. The carbon atoms are joined by single linkages only, being arranged in a straight chain.
i 7 7 7 i ' l i
H
I
H-C-H
H-C-C-C-C-H
I
I H
H
Methane
I
I
H H Butane
H- C-C-C-C-C-C-C-
I
H
S a t u r a t e d hydrocarbons
H
1 H1 H1 H1 H1 HI HI H Heptane ( s t r a i g h t chain)
Formula 1.11-1.
Simple saturated hydrocarbons with which the petroleum industry is concerned are listed in Table 1.11-11. There are paraffins in which the carbon atoms are not all arranged in a straight line. These are called branched chain or isoparaffins. (See Formula 1.11-2.)
34 TABLE 1.11-11 Saturated hydrocarbons Name
Formula
Boiling point, OF
Specific gravity at 68’F
Methane Ethane Propane Butane Pentane Hexane Heptane Octane
CH.4
- 258.9 - 127.8 - 43.8 31.1 97.0 155.7 209.1 258.1
0.415-164 0.546-RR 0.585 - 4 5 0.601’ 0.627 0.659 0.684 0.703
C2H6
C3HR ca12
Cd-44 C8HIR
*
* Unless superscript is presented (temperature of liquid in “C). Temperature of water to which density is referred is 39.2”F.
H
I
H-5-H
H-C-C-C-H
I
H-C-H
I
H
Isobutane
I
H-C-C-C-C-C-H H
I
H H-C-H
I
I I H , IH
H
H-C-H
I
H
lsooctane
Formula 1.11-2,
The principal hydrocarbons found in natural gas are shown in Fig. 7-3, p. 180. The paraffins are present in aviation fuel to a greater extent than in all other compounds put together (usually at least 60% by volume). The name “paraffin” is derived from a Latin word meaning inactive or inert. The paraffin compounds show the least tendency to unite with other compounds or to attack and dissolve metals, rubber, and other parts of the fuel system. In addition, they have the highest amount of heat energy per pound when compared to the other hydrocarbon compounds. This is due to their capacity to hold the greatest possible amount of hydrogen. The performance numbers are also affected by the arrangement of carbon atoms. If the carbon atoms are arranged in a straight chain, as in normal heptane, performance numbers are low, whereas if arranged in a “branched” chain as in isooctane (2,2,4-trimethylpentane),the performance numbers are greatly improved. The paraffins as a class are more resistant to preignition than any other group of compounds used in aviation fuels, and usually have extremely low freezing points.
35
Unsaturated hydrocarbons
Olefin unsaturated hydrocarbons are represented graphically by a double line between one pair of adjacent carbon atoms. This is to indicate the fact that they are able to take up or combine with some more hydrogen or other elements. The general formula for olefins is C,,Hz,: (See Formula 1.11-3.) H
I
H
I
c=c
C=C-C-H
I HI H
H
I
Ethene
I
H
Propene
Formula 1.11-3.
It is doubtful whether olefin hydrocarbons occur in crude oil, but a high percentage is found in products which are made during cracking operations. The olefins are the least inert of the compounds in aviation gasoline and combine with air or with themselves to form varnish-like or rubber-like materials. The olefins are largely excluded from aviation fuels (less than 5%) also because of their relatively poor performance numbers at lean mixtures. In addition, olefins have somewhat greater solvent power than the chain paraffins, and have the lowest resistance to preignition. Unsaturated hydrocarbons which contain two pairs of double-linked carbon atoms in a straight chain are called diolefins. One of the well-known diolefins is butadiene which is playing so great a part in the synthetic rubber industry.
Butadiene
Formula 1.114
Simple unsaturated hydrocarbons with which the petroleum industry is concerned are listed in Table 1.11-111. Unsaturated hydrocarbons are unstable and are readily attacked and acted upon by strong acids. This is easily understood because in the case of the unsaturated hydrocarbons there is a vacant space on the carbon atom ready to grab up some substance in order to fulfill its property of having four univalent atoms around it. Naphthene hydrocarbons
Cyclic paraffins are saturated compounds having a closed ring structure instead of a straight chain. Their formulas correspond to C,H,,. Cyclopentane and cyclohexane shown below are typical examples (see Formula 1.11-5).
36
\/
H
H
H
H I
" \ /
\/H
:;1
C ti ' H"H
CHI
H H'
Cyclopentone
\
p
' H
H'
H
Cyclohexa ne
Formula 1.11-5.
TABLE 1.11-111 Unsaturated hydrocarbons Name Ethene Propene Butene(1-) Isobutene Pentene(1-)
Formula
Boiling point, "F
Specific gravity
CIH4
- 154.5 - 53.3 20.8 20.1 86.4
0.00126 at 0°C 0.5946 at 68°F 0.6411 at 68°F
c3H6
C,H* C4H8 C5HlO
The naphthene hydrocarbons shown to exist in gasoline are cyclopentane and cyclohexane and their substitution compounds. Simple naphthene hydrocarbons are listed in Table 1.11-IV. The naphthenes can be present to the extent of 30%of the total volume of fuels. Because of their lower hydrogen content, they have less energy per unit weight than the paraffins. The performance numbers of the cyclic paraffins are better at rich mixtures than at lean mixtures, and vary from good to bad. The solvent power of cyclic paraffins is greater than that of chain paraffins.
TABLE 1.11-IV Naphthene hydrocarbons Name Cyclopropane Cyclobutane Cyclopentane Cyclohexane Cycloheptane
* In 'C.
Formula C3H6 C4H8 C5H10
C,H,2 C,H,4
Boiling point, O F
Specific gravity at 68'F
- 26.9 55.4 120.7 177.4 245.1
0.720-79 * 0.69' * 0.745 0.780 0.8099
37
Aromatic hydrocarbons
Aromatic compounds are hydrocarbons which contain a benzene ring nucleus in their structure. The general formula for this type of chemical compounds is CnH2,,-6.This group of compounds has a characteristic aroma or smell. They react readily with concentrated sulfuric acid, a property which differentiates them from the paraffin and naphthene hydrocarbons. The percentage of carbon to hydrogen is greater in aromatic compounds than in the other hydrocarbons discussed previously. Here the carbon percentage is approximately 92 and hydrogen 8, by weight. They are characterized also by relatively high specific gravity, thus having a greater weight in pounds per gallon than saturated, unsaturated, and naphthenic hydrocarbons of a similar boiling point. Aromatic hydrocarbons are found in various crude oils and many are produced as a result of “cracking” crude oil. Although it is not possible to have them present to as great an extent as the cyclic paraffins, the aromatics are the second most important group of compounds in aviation fuels. Inasmuch as they contain considerably less energy. per unit weight than chain paraffins, the amount permitted in aviation gasoline does not exceed about 20% by volume. The aromatics are particularly desirable because of their performance numbers at rich mixtures. The aromatics are almost as inert as the chain paraffins in respect to combining with other compounds, and they have high stability. They have powerful solvent tendencies, however, and either dissolve or cause swelling of rubber and rubber-like substances. The preignition resistance of the aromatics as a class is also distinctly inferior to that of the paraffins. Examples of the structural arrangement of carbon and hydrogen in aromatic compounds are shown below: H
I
H-C-H
H
I
H-C
H-C
/‘\C-H
1I I \ rH C
I Benzene
I
H
Toluene
Formula 1.11-6.
Several members of the aromatic (benzene series) hydrocarbon family are listed in Table 1.11-V.
38 TABLE 1.11-V Aromatic hydrocarbons Name Benzene Toluene Ethylbenzene Ortho xylene o-Methyl ethylbenzene
Formula
Boiling point, “F
Specific gravity at 68°F
‘bHb
176.2 231.3 277.0 291.4 323.6
0.879 0.8669 0.8672 0.8802 0.873
C,H” CXHIO CXH1, CP,,
Classification of petroleums
Many methods of classification of crude oils have been devised. Systems based on a superficial inspection involving some physical property, such as specific gravity, are easily applied and specific gravity is actually used to a large extent in expressing the quality of crude oils. A more rational basis of classification is found in some expression of the composition of the oils. In American practice, crude oils long have been roughly
Fig. 1.11-2. Chemical composition of some crude oils plotted on a triangular diagram. (After Nelson, 1949, p. 87, fig. 7.)
39 AROMATIC HC+ NSO COMPOUNDS
AROMATIC INTERMEDIATE 01
PARAFFINIC OILS OILS
N+I SO-ALKANES (PARAFFINS)
80
60
50
40
2o CYCLO-ALKANES (NAPHTHENES)
Fig. 1.11-3. Ternary diagram showing the composition of six classes of crude oils from 541 oil fields. (After Tissot and Welte, 1978, p. 373.)
classified as (a) paraffin base, (b) naphthenic or asphaltic base, and (c) if they contain both paraffin and asphalt, mixed base (Fig. 1.11-2). This system was derived on the basis of differences in the nature of the lubricating oil portion of the crude after a nondestructive distillation was made. In other words, the crude oil is carefully distilled and the portion boiling in the lubricating oil range is examined. If this portion is waxy and has the physical properties of paraffins, the crude oil is termed paraffin base. On the other hand, if the lubricating oil portion contains little or no wax and does contain asphaltic material, the crude oil is termed asphaltic. Crude oils, lubricating oil fraction of which contains both paraffins and asphalt, are termed mixed base. Using this basis for the classification of petroleum, it has been found that crude oil occumng in various sections of the United States can also be classified. The Pennsylvania type of crude oil is paraffinic. This type of crude oil is found in the eastern states such as Pennsylvania, West Virginia, New York, Michigan, and Ohio. Midcontinent type of crude oil is mixed base and is found in Kansas, Oklahoma, all of Texas except the Gulf coastal area, northern Louisiana, and Arkansas. It includes also eastern Colorado and parts of New Mexico and Arizona. The Gulf Coast type of crude is asphaltic and naphthenic in nature and is found in the area lying in southern Louisiana and southern Texas. Tissot and Welte (1978) classification of crude oils is presented in Fig. 1.11-3. Some rules of nomenclature
The most important rules of nomenclature can be summarized as follows: (1) The stem name of an alkane corresponds to the longest carbon-to-carbon chain present. (2) The carbon atoms in the longest chain are numbered to indicate the location of attached groups. In order to permit the smallest numbers to be used in the name, the numbering should start from the end closest to the attached group.
40
(3) Prefixes are used to specify the attached group; numbers are used to denote carbons to which groups are attached. (4) The generic name for open-chain hydrocarbons with one double bond is alkene. The specific name is derived from the name of corresponding alkane by changing -ane to -ene. If two double bonds are present, the generic name becomes alkadiene (-diene ending). ( 5 ) The stem name of 60th alkene and alkadiene corresponds to the longest carbon-to-carbon chain containing the double bond. Examples H
I
H-F-H
a . 2-Methylpentane
Formula 1.11-7.
b. 2,5
- Dimethylheptane
Formula 1.11-8.
H
1
C
H-C-C-C-C-C-C-H
I
H
l
H
I
l
H
H
c. 2 -Ethyl
- 1 -pent ene
Formula 1.11-9.
H-C=C-C=C-H
d.
1,3-Butadiene
Formula 1.11-10.
I
H
41 REFERENCES Binder, R.C., 1962. Fluid Mechanics. Prentice-Hall, Englewood Cliffs, New Jersey, N.J., 4th ed., 453 pp. Chilingar, G.V., 1956. The TechnologV of Testing Petroleum Products. USAF F'ubl., 2nd ed., 76 pp. Chilingar, G.V., 1960. Cenozoic-type and Paleozoic-type oils, by A.A. Kartsev, A Review. Compass Sigma Gamma Epsilon, 37(4): 331-336. Chilingar, G.V. and Beeson, C.M., 1969. Surface Operations in Petroleum Production. Am. Elsevier, New York, N.Y., 397 pp. Frick, T.C. (Editor), 1962. Petroleum Production Handbook, Vol. 1. McGraw-Hill, New York, N.Y. Nelson, W.L., 1949. Petroleum Refinery Engineering. McGraw-Hill, New York, N.Y., 3rd ed., 830 pp. Nelson, W.L., 1958. Petroleum Refinery Engineering. McGraw-Hill, New York, N.Y., 4th ed., 960 pp. Tissot, B.P. and Welte, D.H., 1978. Petroleum Formation and Occurrence. A New Approach to Oil and Gas Exploration. Springer, Berlin-Heidelberg-New York, 538 pp.
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
43
Chapter 2
PHYSICAL PROPERTIES OF PETROLEUM GASES AND LIQUIDS FRANK J. LOCKHART and GEORGE V. CHILINGARIAN
INTRODUCTION
Surface operations in petroleum production involve hydrocarbon gases and liquids with a smaller content of nonhydrocarbon compounds. Although, solid phases are sometimes present, they are avoided whenever possible. Natural gas and natural gasoline are primarily mixtures of the lighter hydrocarbons with varying amounts of nonhydrocarbons such as water, carbon dioxide, and hydrogen sulfide. The heavier mixtures such as crude oil consist of a myriad of TABLE 2-1 Some physical constants of light hydrocarbons and other components of petroleum fluids (Abstracted from Natural Gas Processors Suppliers Association, 1981) Compound
Methane Ethane Propane Isobutane +Butane Isopentane n-Pentane n-Hexane n-Heptane n-Octane n-Nonane n-Decane Nitrogen Oxygen Air Carbon dioxide Hydrogen sulfide Water
Formula
CH4 C2H, C,H, C4Hlo C4H,, CSHl2 CSHI2 C6H,, C,H,, C,H,, C,H, CIoH22 N2 0 2
N2+02 CO, H2S H2 0
Molecular weight
16.0 30.1 44.1 58.1 58.1 72.2 72.2 86.2 100.2 114.2 128.3 142.3 28.0 32.0 29.0 44.0 34.1 18.0
Boiling point at 1 atm (OF) -258.7 -127.4 -43.7 10.7 31.1 82.1 96.9 155.7 209.2 258.2 303.5 345.5 -320.4 -297.3 -317.8 -109.3 -76.6 212.0
Vapor pressure at 100°F (psis)
-
-
188.0 72.4 51.5 20.4 15.6 5.0 1.62 0.54 0.18 0.06
-
387.1 0.95
Critical temperature
(OF)
Critical pressure (psia)
Liquid density at 60°F ( g m )
- 116.7 90.1 206.0 275.0 206.6 369.0 385.6 453.6 512.7 564.1 610.5 651.6 - 232.7 181.2 221.4 87.9 212.6 705.5
-
667.8 707.8 616.3 529.1 550.7 490.4 488.6 436.9 396.8 360.6 331.8 304.4 493.0 736.9 546.9 1071.0 1036.0 3208.0
0.356 0.508 0.563 0.584 0.624 0.631 0.664 0.680 0.707 0.722 0.734
-
0.818 0.787
1.OOO
44
higher boiling hydrocarbons, too many in number for individual identification, and various compounds containing sulfur, nitrogen, and oxygen. Certain physical characteristics of these hydrocarbon mixtures are needed in order to design new equipment and to understand and predict the performance of existing equipment. Many of these physical properties are given in the standard references such as Nelson (1958), American Petroleum Institute (1966), and Natural Gas Processors Suppliers Association (1981). Some physical constants of the lighter hydrocarbons and other components of petroleum gases and liquids are given in Table 2-1. Properties of special importance, i.e., density, viscosity, and vapor-liquid equilibrium ratios, are discussed in this chapter.
DENSITY OF GASES
The ideal gas relationship presented below is convenient and usually acceptable when dealing with gases at low pressures, say less than 300 psia:
pu
=NRT
(assuming 1 lb of gas)
(2-1)
and y = l/u = p / N R T = Mp/RT
(2-2)
where p = pressure, psia; u = specific volume, ft3/lb; N = number of pound moles; R = gas constant, 10.73 (psia X ft3/lb-mol X OR); T = absolute temperature,’R (= OF 460); y = specific weight, lb/ft3; and M = molecular weight. As the pressure is increased, eqs. 2-1 and 2-2 become gradually less accurate. Errors of about 2% at atmospheric pressure become as great as 500% at higher pressures. The simplest method of compensating for such errors is the introduction of a multiplying factor on the right hand side of eq. 2-1, and in the denominator of eq. 2-2. This dimensionless factor, Z , is called the “compressibility factor” and was determined empirically. It is probably accurate to within about 5 % for most gases. Thus, eqs. 2-1 and 2-2 will become:
+
pu = Z N R T
(2-3)
and
MP
Y=ZRT
(2-4)
The compressibility factor is determined by the composition, temperature, and
45
9
I
PSEUDO REDUCED PRESSURE 5 3 4
2
6
7
8.
N
PSEUDO REDUCED PRESSURE PR
Fig. 2-1. Compressibility factor for natural gases. (From Natural Gas Processors Suppliers Association, 1981, fig. 16-3, p. 16-8.After Brown et al., 1941 in: Natural Gasoline Supply Men’s Association, 1957.)
46
pressure of the gas. For a single compound, Z is a function of reduced temperature, T,, and reduced pressure, p,, as shown in Fig. 2-1.
and
where T = temperature of gas, absolute units (OF + 460); T, = critical temperature of the compound, absolute units (OF + 460), from Table 2-1; p = pressure of gas, absolute units (psig 14.7); and pc = critical pressure of the compound, absolute units (psig + 14.7), from Table 2-1. Any units of temperature or pressure may be used, provided that the same absolute units are used for T and T,, and for p and p,. For mixtures of gases, Fig. 2-1 can also be used upon calculating the “pseudoreduced” temperature and pressure:
+
n
pPr= ~
/
~ = p P c/
CYi Pci 1
The pseudo-critical temperature (pressure) of a gas mixture is calculated as the molecular average critical temperature (pressure) of the components. Example 2-1
Calculate the density of a gas mixture containing 90% methane, 8% ethane, and 2% propane at 55’F and 1100 psig. Note that gas compositions are usually given in volume percents, which are equal to mole percents. Composition
Mole fraction Y
CH, C2H6 C3H8
0.90 0.08 0.02 1.oo
Molecular weight M
yM
16 30
44
r,
YTC
343
550
17.7
666
Pc
YPC
(PW
(OR)
366
668 708 616
670
41
+ 460)/366 = 1.41 Using eq. 2-8: ppr= (1100 + 14.7)/670 = 1.66 Using eq. 2-7: Tpr= ( 5 5
From Fig. 2-1 : Z = 0.81 Using eq. 2-4: y
MP
= -=
ZRT
(17*7)(1100 + 14'7) (0.81)(10.73)(55 + 460)
= 4.41
lb/ft3
DENSITY OF LIQUIDS
The approximate density of petroleum liquids at elevated temperatures and pressures can be obtained from Fig. 2-2. This figure relates the specific gravity (equal numerically to the density in g/ml) to temperature and pressure as a function of any two of the following three parameters: "API, Watson K, characterization factor, and mean average boiling point. In general, the first two parameters are known rather than the last one: "API =
l4IS
SG (60°F)
- 131.5
or
and
K,
= characterization
factor =
3K SG (60°F)
(2-10)
The characterization factor K, is a factor which classifies crude oils and petroleum liquids roughly as paraffin-predominant (K,= 12.0-12.5), naphthene-predominant (K,= 11-12), and aromatic-predominant (K,= 10-11). T, = average boiling point, O R. Example 2-2
What is the density (g/ml) of a petroleum liquid at 500°F and 1000 psia, if the oil gravity is 40 "API and K, = 11.0? Using eq. 2-9, 40 "API corresponds to a density of 0.825 g/ml at 60°F. In Fig. 2-2, the dashed line shows the density to be 0.608 g/ml at 500°F and 1000 psia.
~Poo:
1.05
900I.00
EXAMPLE - A T 500 F A 40 API OIL.K,*110,
800-
@
@
UPS A SP OR OF 0.SOB AT 1.000 PSlA
6
mo -
600
-
IA.
W
K
a
z K
W
a
I
w
c
@/I
m;',
-
\
\ \
\ \
400-
\ \ \
1
300-
2
em-
\ \ \
\ \
100-
0-
Fig. 2-2. Specific gravity of petroleum liquids. (After Ritter et al., 1958, fig. 4, p. 230; courtesy of Petroleum Refiner.)
49
Extending the dashed line of the example to the 500 psia line, the density is 0.590 g/ml at 500°F and 500 psia. VISCOSITY OF GASES
The viscosity of petroleum gases at low pressures (say below a reduced pressure of 0.6) may be obtained from Fig. 2-3. The pressure limitation is in most cases about 400 psia. The effect of pressure may be obtained from Fig. 2-4.
TEMPERATURE, OF
Fig. 2-3. Relationship between viscosity and temperature. (After Bicher and Katz, 1944; American Petroleum Institute, 1966, fig. llB3.1, p. 11-73.)
REDUCED PRESSURE, p,
Fig. 2-4. Effect of pressure on viscosity of petroleum gases. (Modified after Carr et al.. 1955; American Petroleum Institute, 1966, fig. llB4.1, p. 11-75.) p1 = viscosity at 55'F and 14.7 psia.
50
Example 2-3
What is the viscosity of the petroleum gas given in Example 2-l? From Example 2-1, the average molecular weight = 17.7, the pseudo-reduced temperature = 1.41, and pseudo-reduced pressure = 1.66. From Fig. 2-3, the viscosity at 14.7 psia and 55°F is equal to 0.0104 cP. From Fig. 2-4, p / p I = 1.21. Thus, the viscosity at 55°F and 1100 psig = (1.21)(0.0104)= 0.0126 cP. VISCOSITY OF LIQUIDS
There are no accepted correlations of the viscosity of petroleum liquids as a function of temperature. It is usual procedure to plot the viscosity versus temperature on ASTM viscosity paper (four charts, low and high ranges, for universal and kinematic viscosities are available from the American Society for Testing Materials, Philadelphia, Pa.). In order to use this method it is necessary to know the viscosity at two temperatures in order to define the straight line. Extrapolation beyond these two experimental points is not recommended: at lower temperatures, there is danger of getting into the solid region, whereas at higher temperatures there is danger of getting into a region of cracking or thermal decomposition.
TEMPERATURE,
*F
Fig. 2-5. Effect of temperature on viscosity of some crude oils.
51
Alternate methods which are probably less accurate include the use of the semi-logarithmic equation: log 7 = a ( l / T )
+ k,
(2-11)
where 7 = kinematic viscosity in centistokes; T = absolute temperature, (OF + 460); and a, k , = constants. This method was revived by Amin and Maddox (1980). Also, the logarithmic relationship can be used: log 7 = b log
t
+k,
(2-12)
where t = temperature,"F, and 6 , k , = constants. Equations 2-11 and 2-12 often do not yield straight lines. The use of eq. 2-12 is shown in Fig. 2-5 from data on crude oils given by Nelson (1946a,b, 1954, 1958). Although Fig. 2-5 indicates a rough relation between viscosity and specific gravity (or OAPI), this is definitely not sufficient to eliminate the need for experimental data to define the relationship. Most viscosity data, including those in Fig. 2-5, were determined at atmospheric pressure.
8
8
CWARACTERlZATlOW FACTOR.
PRESSURE,
PBlA
Yw
Fig. 2-6. Effect of pressure on the viscosity of petroleum liquids at relatively low reduced temperatures. (After Lockhart and Lenoir, 1961, fig 1, p. 209; courtesy of Petroleum Refiner.)
52
The effect of higher pressures may be significant, especially for petroleum liquids having high viscosity (see Fig. 2-6).
Example 2-4 A petroleum liquid having density of 0.80 g/ml and viscosity of 400 centistokes (cs) at 60°F is subjected to a pressure of 2600 psia. What is its viscosity at this higher pressure? p14.7= (400)(0.80)= 320 CP = 1.65. From Fig. 2-6, there is negligible effect of K , and p2600/pF114,7 p at 2600 psia = (320)(1.65) = 528 cP.
VAPOR-LIQUID
EQUILIBRIUM RATIOS
Petroleum production usually involves handling of vapor and-liquid phases. It is usually assumed that the vapor is in equilibrium with the coexisting liquid phase, although this does not have to be correct in all cases. The concept of equilibrium, however, makes calculations much easier. By definition: yi = K i x i
(2-13)
where yi = mole fraction of compound “i ” in the liquid, and K i = vapor-liquid equilibrium ratio of compound “i ”. For hydrocarbon mixtures at low pressures and reasonable temperatures, the vapor and liquid behaviors approach ideality:
K i =p i / P
(2-14)
where p i = vapor pressure of component b b i ” at the system temperature, and P = total pressure of the system, in the same units as used for pi. Although eq. 2-14 should not be used for most temperatures and pressures encountered in petroleum production, it can be very useful as a limiting asymptote in checking K i P values for accuracy or consistency. Over wide ranges of temperatures, pressures, and compositions of numerous compounds, it has been found experimentally that Ki is a function of the compounds and their concentrations in addition to the temperature and pressure. For most petroleum problems, however, it is usually sufficiently accurate to assume that K i is a function of only temperature and pressure. DePriester (1953) developed two nomographs which are of general use in the industry (Figs. 2-7 and 2-8). The limiting conditions of the two-phase region are the bubble point (100% liquid) and the dew point (100% vapor). Calculations for these two limits are illustrated below.
w
(n
N-HEPTANE
2
N-NONANE
N - OCTANE
N-HEXANE
N -PENTAN€
6
ISOPENTANE
ISOBUTANE N - BUTANE
METHANE 1
. . . ,I .... 1
~
rnz
0-
DISTRIBUTION COEFFICIENTS IN LIGHT HYDROCARBON SYSTEMS
GENERALIZED CORRELATION LOW TEMPERATURE RANGE
K=
YX
f
E
2
'cf
e
Fig. 2-7. Generalized correlation for equilibrium ratio K in low- temperature range. (After DePriester, 1953, fig. 31. p. 41.)
54
I
.
DISTRIBUTION COEFFICIENTS IN LIGHT HYDROCARBON SYSTEMS GENERALIZED CORRELATION HIGH TEMPERATURE RANGE
K=
y/x
Fig. 2-8. Generalized correlation for equilibrium ratio K in high-temperature range. (After DePriester, 1953, fig. 32, p. 49.)
55
Example 2-5
What is the bubble point temperature of the following mixture at 250 psia? The mixture composition equals the liquid composition at the bubble point (eq. 2-13). This is an iterative-type calculation. Assume a temperature, obtain K values from Figs. 2-7 and 2-8, use eq. 2-13 to calculate C K i x i .When the latter is equal to 1.00 exactly, the individual K i x i is equal to yi for each component. Compound
x,
250 psia 100°F
120°F
K, Methane Ethane Propane Isobutane
0.03 0.12 0.25 0.60
-
K,x,
10.2 2.35 0.82 0.275
-
1 .00
K, 11.2 2.80 1.02 0.35
0.958
K,x,
1.137
Linear interpolation suggests trying a temperature of 105'F, at which the K values are 10.7, 2.52, 0.90, and 0.295, respectively. At 105'F the CKixi = 1.025. Therefore, the bubble point is between 100'F and 105"F, and will be about 103°F. Any closer check is not warranted. It is noted that if C K i x i < 1.0, the mixture is below its bubble point at that temperature and pressure. If X i x i > 1.0, the mixture is above its bubble point. The above example shows that the light component plays a predominant role in determining the bubble point. Example 2-6
What is the dew point temperature of the mixture in Example 2-5 at 250 psia? The composition of the mixture equals the vapor composition at the dew point (eq. 2-13). Compound
Y,
250 psia 200°F
Ki Methane Ethane Propane Isobutane
0.03 0.12 0.25 0.60
-
1 .00
180'F .Vi
/Ki
13.5 4.4
1.95 0.80
0.9077
Ki 13.0 4.0 1.70 0.70
.Vi
/Ki
1.037
Linear interpolation suggests trying a temperature of 186"F, at which the K values are 13.2, 4.05, 1.80, and 0.71, respectively. At 186°F and 250 psia, C y i / K i = 1.016.
56
Therefore, the dew point is between 186°F and 200°F and is about 187°F. Any closer check is not warranted. It is noted that if C y , / K , < 1.0, the mixture is above its dew point at that temperature and pressure. When C y , / K , = 1.0 exactly, the individual y i / K , is equal to x i for each component. This example shows that the heavy component of the vapor plays a predominant role in determining the dew point. Calculations where the feed mixture is in a two-phase region in-between the bubble point and the dew point are discussed in Chapter 3 titled "Separation of Oil and Gas".
INTERRELATIONSHIP AMONG VARIOUS PHYSICAL PROPERTIES OF CRUDE OILS
Interrelationship among volumetric average boiling point, gravity, molecular weights and critical temperature is presented in Fig. 2-9. For example, the volumetric average boiling point of a distillate material, having an ASTM slope of gasoline and gas oil and API gravity of 31.5", is 533°F (dashed line in Fig. 2-9; see Nelson, 1949, p. 146). The molecular weight is equal to 195 and critical temperature is 845°F (Fig. 2-9, dashed line; see,Nelson, 1949, pp. 145 and 146). As a second example, if a mixed-base stock has a characterization factor of 11.9 and API gravity of 35" atmospheric molal average boiling point = 580°F (1O4O0R), molecular weight = 250, and critical temperature = 900°F (1360"R) (from Fig. 2-9; see Nelson, 1949, pp. 143 and 146). Thus, if any two properties are known, the other properties can be determined.
SAMPLE PROBLEMS AND QUESTIONS
(1) What is the viscosity of a blend of (a) 65% of oil having viscosity of 550 SSU at 130°F and (b) 35% of oil with a viscosity of 250 SSU at 130"F? (2) A 70-OAPI gasoline has a volumetric average boiling point of 300°F. What is its viscosity at 60 psia and 400"F? (3) Determine the total heat of vapor at 450°F and 600 psia. The API gravity is equal to 80" and characterization factor K is 10.5. Given: ASTM slope = 3"/1%. (4) A mixture of hydrocarbons has a characterization number of 11.0 and a molecular weight of 200. Determine the latent heat of vaporization at atmospheric pressure and also at 450°F. The ASTM distillation slope is equal to 6"/1%. (5) What is the specific weight in lb/ft3 of a 33-OAPI mixture of hydrocarbons at 60°F and at 300°F? (6) One hundred ft3 of ethane at 60°F and atmospheric pressure are compressed to a pressure of 750 psig and temperature of 89°F. What volume would ethane occupy at these conditions? (7) Ten pounds of 60-"API gasoline at 50°F are heated, vaporized, and superheated to 600°F at an absolute pressure of 300 psia. How much heat is required?
I
553
E. 6 W
a 4
9m
2
I $ 0
a 100
200
300
400
500
600
700
000
CRUDE OILS REDUCEDCRUDES GAS OIL AND WBES LIGHT LUBES CASO AND GAS OIL DISTILLATES AVIATION GASO-KERO SOLVENTS
Fig. 2-9. Interrelationship among molecular weight, pseudocritical temperature, characterization factor, gravity (OAPI), average volumetric boiling point, molal average boiling point, and slope of ASTM boiling curve (degrees/$). (After Hougen and Watson, 1954; courtesy of John Wiley and Sons, Inc.)
Given: critical pressure = 350 psia; average molecular weight = 140; critical temperature = 600°F;and molal average boiling point = 300°F. (8) Is mold average boiling point higher or lower than the volumetric average boiling point? Why? (In solving problems, see Nelson, 1949, 1958.).
58 REFERENCES American Petroleum Institute, 1966. Technical Dafa Book Pefroleum Refining. A.P.I., New York, N.Y. Amin, M.B. and Maddox, R.N., 1980. Estimate viscosity vs. temperature. Hydrocarbon Process., 59: 131-135. Bicher, L.B. and Katz, D.L., 1944. Viscosity of natural gases. Trans. AZME, 155: 246-252. Carr, N.L., Parent, J.D. and Peck, R.E., 1955. Viscosity of gases and gas mixtures at high pressures. Chem. Eng. Prog. Symp. Ser., 51 (16): 91. DePriester, C.L., 1953. Light hydrocarbon vapor-liquid distribution coefficients. Chem. Eng. Prog. Symp. Ser., 49(7): 1-43. Hougen, O.A. and Watson, K.M., 1954. Chemical Process Principles. Vol 1. Wiley, 2nd Ed., New York, N.Y., 436 pp. Lockhart, F.J. and Lenoir, J.M., 1961. Liquid viscosities at high pressures. Per. Refiner, 40(3): 209-210. Natural Gasoline Supply Men’s Association, 1957. Engineering Dara Book. Natural Gasoline Association of America, Tulsa, Okla., 7th Ed., 174 pp. Natural Gas Processors Suppliers Association, 1981. Engineering Data Book. N.G.P.S.A., Tulsa, Okla. Nelson, W.L., 1946a. Viscosity of crude oils. Oil Gas J.,Jan.5: 70. Nelson, W.L., 1946b. Viscosity at pipe line temperature. Oil Gas J.,Jan. 12: 87. Nelson, W.L., 1949. Petroleum Refinery Engineering. McGraw-Hill, New York, N.Y., 3rd Ed., 830 pp. Nelson, W.L., 1954. How to handle viscous crude oils. Oil Gas J.,Nov. 15: 269. Nelson, W.L., 1958. Petroleum Refinev Engineering. McGraw-Hill, New York, N.Y., 4th Ed., 960 pp. Ritter, R.B., Lenoir, J.M. and Schweppe, J.L., 1958. Find specific gravities by nomograph. Per. Refiner, 37(11): 225-232.
59
Chapter 3
SEPARATION OF OIL AND GAS F.J. LOCKHART, GEORGE V. CHILINGARIAN and SANJAY KUMAR INTRODUCTION
The name “gas and oil separator” is one of a variety of terms used for pressure vessels which separate multiphase well fluids into gaseous and liquid streams. Other names found in the literature include: stage separator, knockout drum, trap, vapor-liquid separator, flash drum, flash chamber, dry drum, scrubber, and settler. Separators are used in many locations other than at wellheads, such as natural gasoline plants, compressor suctions and discharges, liquid traps in gas transmission lines, dehydration plants, and gas sweetening plants. Such drums are designed to separate a gas from a liquid and, in some cases, to separate three phases, i.e., a gas, a liquid hydrocarbon, and a liquid aqueous phase. At times the removal of slugs of liquid from a gas is of such importance that the separator may be sized for its liquid holding capacity. In general oilfield practices, separators are used to separate oil, gas and water and to remove material such as entrained solid impurities from the crude oil produced from the wells. A simplified diagram of a spherical three-phase (oil-gas-water) separator is presented in Fig. 3-1. GAS OUTLET
CENTRIFUGAL-TYPE
t
COALESCING-TYPE MIST
Fig. 3-1. A simplified diagram of a spherical three-phase (oil-gas-water) separator. (Modified after Smith, 1962, p. 11-18, fig. 11-20),
60
A properly designed wellstream separator must perform the following functions (Ikoku, 1980): (1) Accomplish a primary-phase separation of the liquid from the gaseous hydrocarbons. (2) Refine the primary separation by removing most of the (a) entrained liquid mist from the gas and (b) entrained gas from the liquid. (3) Discharge the separated gas and liquid streams and ensure that no reentrainment of one into the other takes place. Equilibrium flash calculations
In some cases of separator design, the amounts of vapor and liquid and their significant physical properties are known. There are a number of design problems in surface operations, however, where this information is not known and must be calculated. It is essential to know the amount and composition of the total feed to the separator in such cases. Then, assigning a temperature and pressure to the separator, the amounts and compositions of the vapor and liquid streams leaving the separator can be calculated, assuming that equilibrium is attained between these exit streams. The basic concept involves formation of a mixture of vapor and liquid that is kept in intimate contact for a long enough time to enable the entire vapor phase to attain equilibrium with the entire liquid phase. The equilibrium is attained in the piping and equipment just upstream of the separator. The separator itself serves only as a wide spot in the line to help in separating the phases. For the so-called “isothermal” flash, the temperature at the discharge of the mixing zone and within the separator is specified along with the pressure initially, and is attained by proper design and operation of a heat exchanger. If there is no heat exchanger, then an “adiabatic” flash calculation is made, wherein the total feed enthalpy must equal the total product enthalpy at the lower separator pressure. Here the separator temperature is unknown, so there is a double trial-and-error, one nested within the other. Usually the separator pressure is set and a temperature is assumed. The calculation is made exactly as for the isothermal flash. Then the enthalpies of the exit streams are calculated and their sum compared with that of the inlet mixture. At the correct temperature, the exit enthalpies are equal to the inlet enthalpy. Basic equilibrium relations for complex mixtures
In complex mixtures, feed F entering the flash separator consists of F,, F,, F,, ..., F, moles of different components. At a given temperature and a fixed pressure, V moles of gas are produced (V, + V, + V, + ... K). A liquid residue L consists of L,, L,, L,, ... , L, moles of the several components present in F. Thus,
+
See Appendix 3.1.
61
the.mole fraction of each component may be expressed as F,/F, V , / V , L , / L , etc. In addition, F = L + V and F, =.L, + V,. On applying Henry's law:
K/
Kn ( L n / L )
(3-1)
where V , / V = y and L , / L = x ; thus: V,=KnLn(V/L)
Substituting (F, -
(3-2)
V,) for L,:
V,=Kn(F,- V,)(V/L)
(3-3)
Solving for V,:
V, = F n / [ ( L / K n v ) + 11
(3-4)
Thus: V = Fl/[ ( L / K , V ) + 11
+ F 2 / [ L / K , V ) + 11 + ...
and
TYPES OF SEPARATORS
Inasmuch as most separators are designed for the removal of liquid drops from the gas by the action of gravity, most of the discussion here deals with this general type of separation. Another basic type of separator, however, uses the action of centrifugal force to remove liquid drops from gas. These centrifugal separators use multiple cyclones in parallel. Cyclones are relatively small and of standard size which can be mass produced. These separators function best when the gas flows at constant rate and pressure. At lower rates, the separation suffers, and at higher rates, the pressure drop becomes excessive. Gravity separators may be classified according to the shape of the vessel: (I) Cylindrical (a) Vertical (b) Horizontal (single-tube or double-tube types) (2) Spherical Each one of these shapes has its own advantages, and there is no overwhelming favorite among them. The vertical separator occupies less ground area and is claimed to have the ability to handle large quantities of sand and to be easier to
TABLE 3-1 Comparison of different separator types
Advantages
Vertical
Horizontal
Spherical
(1) Easier to clean and can handle large
(1) Can handle much higher GOR wellstreams because the design permits much higher gas velocities (2) Cheaper than the vertical separator
(1) Very inexpensive
quantities of sand (2) Saves space (occupies lesser ground area) Provides better surge control Liquid level control is not critical Less tendency for reevaporization of liquid into the gas phase due to the relatively greater vertical distance between liquid level and the gas outlet
Disadvantages (1) It takes a longer-diameter separator for a given gas capacity as compared to a horizontal separator More expensive to fabricate Difficult and more expensive to ship (transport)
Easier and cheaper to ship and assemble Requires less piping €or field connections Reduces turbulence and reduces foaming (thus, it can handle foaming crudes) Several separators may be stacked, minimizing space requirements Greater space requirements generally
Liquid level control more critical Surge space is somewhat limited
(2) Good for low or intermediate GORs (3) Very compact and easy to ship and install (4) Better clean-out
(1) Very limited liquid settling section and rather difficult to use for threephase separation (2) Liquid level control is very critical (3) Very limited surge space
Much harder to clean (hence a bad choice in any sand-producing area) Ideal use
Low to intermediate GORs and where relatively large slugs of liquid are expected
High GOR crudes, foaming chdes, or for liquid-liquid separation. Good for a diverse range of situations.
Intermediate or low GOR; preferably two-phase separation.
63
clean. The horizontal separator can handle foaming crude oils better and is claimed to be more economical for handling large gas volumes, The spherical separator is easier to install and is more compact and adaptable for portable use (see Table 3-1). INTERNAL PARTS OF A SEPARATOR
A gas-liquid separator may consist simply of an empty vessel, which causes the fluid velocity in the entering pipe to be reduced by enlarging the cross-sectional area of flow. Usually, however, the separator includes internal parts to promote separation of the phases. In both cases the separator may be visualized as consisting of: (1) Primary separation section (entrance) for separating the bulk of the liquid from the gas. It is desirable to remove the liquid slugs and large droplets of liquid quickly from the gas stream, and to remove gas from the liquid. Decreasing the kinetic energy (i.e., the velocity) of the fluids is often accomplished with the use of a tangential inlet to impart a centrifugal motion to the entering fluids.
ov
I OV
OL (1) (2) (3) (4) (5)
I n l e t f o r two-phase feed Outlet f o r leaving vapor Outlet f o r leaving l i q u i d Primary separation s e c t i o n Secondary separation s e c t i o n Liquid separation s e c t i o n Mist extractor Vortex breaker
I
OL (a)
Fig. 3-2. Internal parts of a gravity separator. a. Vertical separator. b. Horizontal separator.
64
(2) Secondary separation section for removing smaller particles of liquid by gravity settling depends to a large extent on the decreased gas velocity and reducing the turbulence of the gas. (3) Liquid separation section (or the liquid accumulation section) for removing gas bubbles which may be occluded with the liquid, and for sufficient storage of the liquid to handle slugs of liquid anticipated in routine operation. This section thus provides the surge capacity. (4) Mist extraction section for removing from the gas the entrained drops of liquid, which did not separate in the secondary separation section. Mist eliminators may be used to decrease the amount of entrained liquid in the gas and/or to reduce the diameter of the vessel. There are two basic types of mist eliminators commonly used: the vane type and mesh pads. The vane type is mounted in such a manner that the gas flows horizontally through a multiple number of closely-spaced vertical baffles. Entrained liquid particles impinge on the baffle surfaces and are forced into liquid drainage pockets which are out of the gas flow path. Separated liquid drains down out of the vanes by gravity. This type of mist eliminator operates at small pressure drops (e.g., less than 0.2-in. water) and is efficient in removing droplets of around 40 p m in size and larger. The large open areas make the vane type advantageous in systems where solid particles may be present.
C
Fig. 3-3. Basic types of mist eliminators. I. Vane type (courtesy of Peerless Manufacturing Co., Dallas, Tex.): A = mist extractor installation in top of column with trays (top outlet): B = side outlet configuration in vertical column; C = plan view of installation A , showing vane bank arrangement; D = horizontal vessel configuration showing angled position of separator elements; and E = photograph of mist extractor bundle. 11. Mesh pads - Fleximesh@mist eliminator (Copyrighted 1984 by Koch Engineering Company, Inc.)
65
When vapor and entrained liquid droplets pass through o FLEXIMESH mist eliminator the vapor moves freely through the mesh pad but the liquid droplets due to their greater inertia cannot make the necessarysharp turns A Sa result they are thrown into contact with the wire surfaces ond bileflyheld there As more droplets enter the pad and collect on the wires they grow in size run down the wire to the bottom surface of the mesh reporotor and fail from the unit Overhead vapor is now free of entrained liquid
Available FLEX I MESH* Mist Eliminator Styles FIIIFT'
70
50 20 0 27 0 40 40
Fig. 3-3 continued.
SurloceArec
Density
117 112 80 320 432 64
64
115 110 163 86 140 65 65 48 450 610 125 150
I I
I I I
I I I
I I I /
MZIM3
%Voids
377 360 535 282 459 213 213 157 1476 2000 410 492
97 6 97 7 94 0 98 2 98 4 98 5 98 6 990
96 0 94 6 97 0 97 0
66
The mesh pads provide a large surface area of many knitted and crimped wires for collection and coalescence of liquid mist. With the pads usually installed in a horizontal position, the vapor flows upward and the liquid downward. Such mist eliminators are capable of removing very fine droplets (e.g., 4-6 pm) and have good removal efficiency over a wide range of throughput. Pressure drops can be less than 1-in. water. Caution is given to the possibility of plugging caused by the deposition of solid materials such as silt, sand, and paraffins. ( 5 ) Vortex breaker to prevent the liquid from sucking any gas into the liquid exit pipe. Typically, the liquid exit pipe is centered at the bottom of the vessel. A simple vortex breaker can consist of a solid circular plate larger than the exit pipe, supported by three legs about 4-6 in. above the bottom of the vessel. (6) Adequate control devices such as the liquid dump (discharge) valves, gas pressure valves, and safety relief valves. Figure 3-2 shows schematically the internal parts of (a) a vertical gravity separator and (b) a horizontal gravity separator. Additional internal parts may be present in some separators: for example, an inlet baffle plate to help reduce the kinetic energy of the entering stream and a “boot” attached to the bottom of the horizontal vessel where an aqueous liquid phase can be withdrawn. Figure 3-3 shows the two basic types of mist extractors. The gas phase flows horizontally through the vane type and vertically through the mesh pad.
FACTORS INFLUENCING SEPARATION
There are several factors which affect the performance of a vapor-liquid separator. For a feed of given composition, the temperature and pressure are significant. With increasing temperature and/or decreasing pressure, the flow volume of vapor increases and the volume of liquid decreases. In addition, there is a decrease in the densities of both vapor and liquid phases. Other physical properties also change with changing temperature and pressure, but the density is the more significant physical property involved. Some systems may contain small amounts of surfactants, which cause formation of foam (in vapor-liquid systems) and emulsions (in liquid-liquid systems). Formation of foam and emulsion adversely affects the performance of separators, especially when they are stable. Inasmuch as the presence of surfactants is usually not known in advance, separators are usually designed on the assumption that they are not present. Most separators are designed for the removal of liquid droplets from gas by the action of gravity. This separation depends on the physical properties of gas and liquid, and the diameter of the particles. Excellent discussions of the scientific concepts are available, but they only identify the general form of the mathematical relationships. Most laboratory studies were concentrated on the settling of (1) solid particles rather than liquid, (2) a single particle rather than a swarm of particles, and (3) particles having a particular diameter rather than a mixture of sizes.
61
In designing separators, the size distribution of the droplets is not known. Even if it was given, however, it is not known how to select a proper “average” diameter to attain the desired separation. Thus, the most important factors influencing the design of a separator are: (1) flow rate of the gas, (2) flow rate of the liquid, (3) density of the gas, and (4) density of the liquid.
SEPARATOR DESIGN
Gas capacity
Gas capacity of a gas-oil separator is usually calculated from the semi-theoretical relationship proposed by Souders and Brown (1934):
or
and A
= Q/60 u
(3-9)
where: u = superficial gas velocity based on the cross-sectional area of gas flow, ft/sec; G = superficial gas mass velocity, based on the cross-sectional area of gas flow, lb/ft2hr; Q = g a s flow rate, actual cubic feet per minute at the flowing temperature and pressure, ACFM (Actual Cubic Feet per Minute); A = crosssectional area of gas flow, ft2; yG = specific weight of gas, lb/ft3; yL = specific weight of liquid, Ib/ft3; K = an empirical factor representing past experience found to give satisfactory operations, with typical values of 0.25 for vertical separators and 0.50 for horizontal separators. Liquid capacity
Liquid capacity of a separator is dependent upon the retention time (the holding time) of the liquid within the vessel. Liquid holding time is provided to: (1) remove slugs of liquid from the flowing stream in order to protect downstream vapor-handling equipment; (2) keep downstream liquid-handling equipment operating satisfactorily should there be a temporary feed stoppage or overload; and (3) separate the occluded gas particles from the liquid phase. This last justification can become quite important when the liquid is very viscous. Answers to the following questions suggest the rules for holding time for each specific case:
68
(1) In the case of removing slugs of liquid from vapor: (a) What is the estimated maximum liquid rate or the size of slug? (b) How much time would be required to attain the liquid removal rate? (2) For a temporary feed stoppage: (a) What would be the effect on downstream equipment if the drum loses liquid? (b) How much time would it take for the operator to correct for the loss? (3) For a temporary feed overload: (a) What would be the effect on the downstream vapor-handling equipment if the drum overflows? (b) How much time would it take for the operator to correct for the temporary overload? Past experience suggests the following holding times for liquids: (1) Gas-oil separators-2-4 minutes. (2) Gas-oil-water separators-with oil viscosity of less than 100 cP, 3-10 min, and for oil viscosity of greater than 100 cP, 10-20 min. These are typical values which may be changed for specific cases. Vessel design
The following must be considered in designing separator vessels: (1) The volumes of the dished heads are negligible as compared with the volume of the cylinder. (2) Unless specifically stated, the length/diameter ratio ( L / D ) is considered to be acceptable when it is between about 3/1 and 8/1. There is not a great change in costs over this range and other factors such as foundations, plant layout, and symmetry are significant. (3) For a vertical separator, the gas flows through the entire cross-section of the upper part of the vessel. The feed enters the separator just above the vapor-liquid interface, which should be at least 2 ft from the bottom and at least 4 f t from the top of the vessel. The interface does not have to be at the center of the vessel. (4) For a horizontal separator, the interface does not have to be at the centerline of the vessel. In some cases, a smaller-diameter vessel may be obtained by making the interface location off-center and a design variable. The feed enters at the end of the separator just above the vapor-liquid interface, which should be at least 10 in. from the bottom and at least 16 in. from the top of the vessel. Table 3-11 (see also Fig. 3-4) gives some geometric properties of circles which have been derived empirically. The exact relations are more difficult to use. The empirical values in Table 3-11 are of sufficient accuracy and may be used directly. Example 3-I Calculate the diameter and height of a vertical separator for the following conditions: Gas rate = 300 ACFM Gas density = 3.90 lb/ft3 Liquid rate = 22 GPM Liquid density = 40.0 lb/ft3
69
TABLE 3-11 Some geometric properties of circles a ~
Given A / A , to calculate h / D (see Fig. 3-4) Between A / A , of 0 and 0.20 (maximum error = 3.2%): h / D = 2 . 4 8 1 ( A / A T ) - 1 2 . 2 9 (A/A,)2+31.133
(A/AT)3
Between A / A , of 0.20 and 0.80 (maximum error = 1.8%): h/D
= 0.8123
( A/A,) +0.0924
Between A / A , of 0.80 and 1.0 (maximum error = 0.8%): ( 1 - h / D ) =2.481 ( 1 - A/A,)-12.29
(1- A / A ~ ) ~ + 3 1 . 1 3(31 - A / A T ) ~
Given h / D to calculate A / A , Between h / D of 0 and 0.25 (maximum error =1.35%): A/A,=0.21
(h/D)+3.52 (h/D)2-4.93 ( h / D ) 3
Between h / D
= 0.25
A / A , =1.231 ( h / D )
and 0.75 (maximum error = 1.7%): =
0.1138
Between h / D of 0.75 and 1.0 (maximum error = 0.1%): ( l - A / A T ) =0.21 ( l - h / D ) + 3 . 5 2 ( l - h / D ) 2 - 4 . 9 3
a
(1-h/D)3
Exact relations may be found in Perry and Chilton (1973).
L
L
i
Fig. 3-4. Diagram for Table 3-11,
Liquid residence time = 4.0 min Liquid level in the separator is expressed as a fraction of the vessel length between limits of 0.20 and 0.70. (This is not considered to be a rigid specification.) Using eq. 3-7: u = 0.25[(40.0 - 3.90)/3.90)]0.5= 0.76 ft/sec Vapor rate = ACF G min )= T)( 300 = 5.00 ACF/sec
( T)(
(
&)
70
5 .oo
Vapor area = o.76 - 6.58 ft2 Vessel diameter = [4(6.58)/~]'.~= 2.89 ft = 34.7 in. Using a 36-in. diameter vessel, the liquid height required for a 4-min residence time would be: r
1
Vessel length, ft
h r/D
L/D
8 10
0.21 0.17
2.67 3.33
The designer will probably select a diameter of 3.0 ft for the vessel and one of the above two lengths. Example 3-2 Repeat Example 3-1 for a horizontal separator. Using eq. 3-7: u = o.50( 40.0 - 3.90)O.' = 1.52 ft/sec 3.90 Vapor rate = 300/60 = 5.00 ft3/sec Vapor area = 5.00/1.52 = 3.29 ft2 (1) Initially the vapor-liquid interface is considered to be the centerline of the vessel. This may be adjusted later. Vessel area = (2)(3.29) = 6.58 a:t 4 X 6.58 Vessel diameter = = 2.89 ft or 34.7 in. . I I Thus, a 36411. diameter-vessel can be used. Length of vessel needed for 4 min residence time would be:
(
)
Ths indicates that the liquid holdup requirement is not as significant as the vapor handling requirement. (2) Reduce the vessel diameter by lowering the vapor-liquid interface so as to maintain the same required vapor area of 3.29 ft2. On assuming a diameter of 30 in.: A, = 77/4(%)' = 4.909 ft2 A / A , = 3.29/4.909 = 0.67
71
From Table 3-11, h/D 1.59 ft = 19.08 in.
= 0.8123
(0.67) + 0.0924 = 0.6366; h
Vessel length to get 4 min retention time=
= (0.6366)(2.50) =
I(-)(1 -) 1 0.33
= 7.26
4.909
ft
Thus, a vessel 2.5 ft in diameter and 8 ft long would meet the specifications. SEPARATOR DESIGN USING ACTUAL MANUFACTURERS FIELD TEST DATA
Inasmuch as the Souders-Brown equation is basically empirical in nature, a better design can usually be made using the actual manufacturers’ field test data (see Appendix 3.11 and Ikoku, 1980). These correlations account for the additional gas capacity that can be obtained by increasing the height of a vertical separator or the length of a horizontal separator. The Souders-Brown equation does not account for any height-length differences. The correlation charts apply to one-quarter liquid full, one-third-full, and half-full situations. The gas capacity can be increased by decreasing the liquid-filled fraction, e.g., from one-half full to one-third full. Usually the design standard is the one half-full liquid condition. The liquid capacity Q (in bbl/day) can be determined as follows:
Q=
1440(V,)
(3-10)
where Vp = liquid settling volume in bbl and t = retention time in min. The liquid settling volumes for different separator types and sizes are presented in the Appendix 3.11. STAGE SEPARATION
Usually, the single-stage separation is not desirable. By separating the gaseous and liquid hydrocarbons into vapor and liquid phases in two or more equilibrium flashes at consecutively lower pressures, a more stable stocktank liquid can be obtained. In addition, liquid recovery is enhanced. Stage separation can be defined as a “process in which produced crude is separated into liquid and vapor phases by two or more equilibrium-flash vaporizations at successively lower pressures”. Equilibrium or flash vaporization differs from differential vaporization in that the vapor is not removed as it is formed, but is kept in intimate physical contact with the remaining liquid until heating is completed. The storage tank is usually considered as one stage of separation. Differential liberation of gas is the ideal liquid separation system to maximize the liquid recovery. In this process, the pressure is decreased in infinitesimally small steps and the gas liberated at each stage is removed. Inasmuch as this would need an infinite number of separators connected in series, it is obviously uneconomical.
72
WELL-
FLUID
FIRST
LI'
STAOE
TWO
- STAGE
I +'
SLCOMD STA0E
2nd
THREE
- STAGE
J FOUR
- STAGE
Fig. 3-5. Schematic diagrams of two-stage, three-stage, and four-stage separations.
In actual field practice, three stages are usually optimal. Economic study involves determination of fixed and operating costs for each additional separator and comparing them to the incremental oil production that results from the addition of t h s stage to the separation system. Inasmuch as the production varies with time during the lifetime of a producing field, the optimum also changes. Thus, a detailed economic analysis is usually not very useful or justifiable in most cases. Examples of two-stage, three-stage, and four-stage separations are presented in Fig. 3-5. The two-stage separation is most applicable for low-gravity oils, low gas/oil ratios, and low flowing pressures. On the other hand, the three-stage separation is most applicable for intermediate-gravity oils, intermediate to high gas/oil ratios, and intermediate wellhead flowing pressures. Finally, the four-stage, separation is most applicable for high-gravity oils, high gas/oil ratios, and high flowing pressures. Four-stage separation is also used where high-pressure gas is needed for market or for pressure maintenance.
73
The simplified analysis of a stage separation system involves the determination of the operating pressure of each one of the separators connected in series by the following relationship (Campbell, 1976): (3-11) where R = pressure ratio; p 1 = pressure in stage 1 (high-pressure end), psia; p, = stocktank pressure, psia; and n = number of stages minus 1. This implies assumption of equal pressure ratios between the stages, which has been found to be the optimum operating condition for maximizing liquid recovery. The pressure at any stage in between can then be determined using the following equation: Pr’
Pr- 1 R
where pr = pressure at stage r , psia; and R
(3-12) = pressure
ratio.
Determination of optimum pressure for first stage when second stage is atmospheric
To determine the optimum pressure for the first stage when the second stage is atmospheric, the following procedure can be used. (1) Vary the pressure on the high-stage separator. (2) Follow the gas/oil ratios on both stages until they are stabilized. (3) Determine the gasoline content of the gas from each stage, by a method such as gas chromatography. (4) Plot the gas/oil ratio of each separator, and the cumulative for both separators, versus the pressure of the high-stage separator (Fig. 3-6.a). ( 5 ) Plot the gasoline content of the gas from each separator versus the pressure of the high-stage separator (Fig. 3-6.b). The gasoline content for each separator may decrease with increasing pressure of the high stage. For the low stage, this is due to the increase in gas evolved overshadowing the increase in the gasoline vaporized. On the other hand, for the high stage, it is due to the decrease in gasoline vaporized, with increasing pressure. (6) From the graphs of Fig. 3-6.a and 3-6.b, compute and plot the gallons of gasoline lost per barrel of crude oil versus the pressure of the high-stage separator (Fig. 3-6.c). The optimum pressure is at the minimum of the latter curve, providing the primary objective of the separation is to remove as much gasoline as possible from the gas. Determination of optimum pressures for three-stage separation (Whinery-Campbell technique)
The pressure of the highest stage may be fixed by the requirements of a high-pressure sales line or pressure maintenance. In this case, therefore, the forego-
14
Y
0 IL
a
.---Curnulotire
P
PRESSURE OF HIGH.STAGE SEPARATOR. PSI
Fig. 3-6. Schematic diagram of variation, with increasing pressure of high-stage separator, in a. gas/oil ratio; b. gasoline content of separator gas; and c. total gasoline lost (to the gas phase) in two-stage separation.
ing type of evaluation can be used to choose the optimum pressure for the second stage, while the first stage is operating at the desired value. If the pressure of the highest stage is not fixed by such requirements, it may be treated as another variable in the evaluation. Whinery and Campbell (1958, p. 53) developed a method for determining the optimum second-stage pressure in three-stage separation. Their method is simple, accurate (mean error of * 5 % ) , and eliminates the need for flash vaporization calculations. An empirical analysis yielded two equations: p 2 = A ( p,)o.686+ c,
(3-13)
where C, = ( A - 0.057)/0.0233 for wellstreams having a gravity greater than one, referred to air, and:
p 2 = A ( p,)o.765+ c2
(3-14)
75
0
Fig. 3-7. Relationship between constant A and pseudo-specific gravity of feed ( T = 80OF). (After Whinery and Campbell, 1958, p. 54, fig. 2; courtesy of the S.P.E. of A.I.M.E.)
where C, = ( A + 0.028)/0.012 for wellstreams having a gravity less than one, referred to air. In the above equations, p 2 = second-stage pressure in psia, p 1 = first-stage pressure in psia, and A = dimensionless constant which is a function of stocktank pressure ( p , ) and the system composition. Whinery and Campbell found that composition could be expressed in terms of the wellstream gravity and the percentages of methane, ethane, and propane ( C , + C , C,, all in mole%), as shown in Figs. 3-7 and 3-8. Constants A , C,, and C, can be obtained from these graphs (Figs. 3-7 and 3-8). Variation of p 2 with p 1 for a fixed system composition was presented by Whinery and Campbell (1958, p. 53) and is shown here in Fig. 3-9. The maximum of each curve shows the value of p 2 at which the stocktank recovery (gal)/(MMcf') * residue gas is a maximum. It should be remembered, however, that the Whinery-Campbell technique is an excellent tool for the field engineer where time is of prime importance, but is not a replacement for a computer.
+
* M-1000 cases = 60'F.
and MM =1000OOO; Mcf -1000 standard cu ft; standard legal temperature in most
16
. C, DIMENSIONLESS SHIFTING CONSTANl
Fig. 3-8. Relationship between shifting constant C and constant A . (After Whinery and Campbell, 1958, p. 54, fig. 3; courtesy of the S.P.E. of A.I.M.E.)
139
,
138. SECOND STAGE PRESSURE, PSlA
Fig. 3-9. Relationship between second-stage pressure (psia) and stocktank recovery in gal/MMcf of residue gas. (After Whinery and Campbell, 1958, p. 53, fig. 1; analysis 6, T = SOOF; courtesy of the S.P.E. of A.I.M.E.)
71
If it is not desirable to use the experimental methods outlined above, the problem may be solved by calculations with equilibrium ratios. Such calculations, however, are tedious unless a computer is available.
METHODS OF SUCCESSIVE APPROXIMATIONS
In computing the compositions of the liquid and the vapor when neither composition is known, but the equilibrium constants are given, the following procedure can be used. (1) Assume a value for V / L (or L / V ) to get (V/L)ass,l. (2) Preferably using eqs. 3-5 and 3-6, compute the number of moles of each component in the liquid and the number of moles of each component in the vapor. (3) Add each group to get the calculated V . (4) Divide the calculated V by the calculated L to get the ( V/L)calc,l. (5) If the ( V/L)cdc,lis equal to ( V/L)ass,l, the assumption was correct within the desired accuracy. ( 6 ) If ( V/L)calc,l is not equal to ( V/L)ass,l, assume another value for V / L where (V/L)ass,2= (V/L)cdc,l + [(V/L)ca,c.l - (V/L)ass.lI; that is, make (V/L)ass,2 as far beyond (V/L)caIc.l, as (J‘/LlCdc.1 is beyond (V/L)ass.l* (7) One could also plot [( V / L ) a s s-, ( V / L ) c d c ,versus ] ( V/L)ass,;but the above rule has worked out better in the experience of the writers in this type of problems, because the relationship is far from being linear. In computing the temperature and composition of vapor when the composition of liquid is known, the procedure described below can be followed. (1) Assume a temperature and record equilibrium ratios. (2) Compute y1 = K1xl, y2 = K 2 x 2 , . . , and Cy. (3) Assume another temperature which will make E y approach 1. The higher the temperature, the greater is the value of K . For a fixed x , therefore, y will increase with increasing temperature. (4) Plot Cy versus temperature, and establish the point of intersection at Cy = 1. Isothermal flash for two phases The assumption of steady-state flow through the separator gives the mass balances for each individual component as follows:
r;] = y
+ L,
(3-15)
where I;] = moles of component “ 1 ” in the feed, = moles of component “i” in the vapor, and Li= moles of component “ i ” in the liquid. On an over-all basis:
r;;=v,+L,
(3-16)
The second assumption of equilibrium between vapor and liquid in the mixture gives: Y, = K , x ,
(3-17)
where y, = mole fraction of component “ i ” in the vapor, x i= mole fraction of component “i” in the liquid, and K i = vapor-liquid equilibrium ratio of component “ i ” . Equation 3-17 may be represented as follows:
Substituting eq. 3-15 into eq. 3-18 and rearranging gives the following relationship: (3-19)
If the substitution is made for
v,rearrangement gives: (3-20)
Equation 3-18 or 3-19 or any of the possible rearrangements of eqs. 3-15 and 3-18 is may be used in a conventional iterative method. A value of L, is assumed, obtained from eq. 3-16 and the individual L, is determined using eq. 3-20. The L, agrees with it within closely defined assumed L, is correct if the calculated limits. After satisfactory agreement, each is calculated from eq. 3-15, and mole fractions for vapor and liquid can be determined. Although the solution to eq. 3-19 or 3-20 is straightforward, complications may arise for some problems due to slow convergence strategy for manual as well as computer calculations to attain required accuracy in a reasonable time interval. Two strategies for achieving this are described here: (1) the Lockhart-McHenry (1958) method, which is currently used by many engineers, and (2) the Lockhart method (pers. comm., 1983) presented here for the first time.
xy=l
LOCKHART-MCHENRY METHOD OF FLASH-EQUILIBRIUM CALCULATIONS
Lockhart and McHenry (1958) proposed a method which reduces the usual time and luck required for making flash-equilibrium calculations for multicomponent mixtures. It reduces a multicomponent mixture to a hypothetical binary system. When using the conventional method of calculation, the convergence may be slow (calculated V / F versus assumed V / F ) , which makes it difficult to bracket the
TABLE 3-111 Lockhart-McHenry method, sample problem 3 (from Lockhart and McHenry, 1958; with permission of Hyakxarbon Processing, by Gulf Publishing Co., Houston, Tex.) Component Methane Ethane u Light
Propane Isobutane nButane Isopentane n Pentane Hexanes u Heavy Total B
K:
+
K1; +
V* -+ V * / F -+
Fn
Kn
FnK n at V / F = 0
4/Kn
F,/(Kn+l)
at V / F = I
at V / F = 0.5
163.9 369.5 533.4
3.80 1.76
622.5 651.0 __ 1273.5
43.1 210.0 253.1
34.2 134.0 __ 168.2
450.0
0.79
74.1 239.9 56.6 59.3 71.4 951.3
0.44
355.3 32.6 84.0 11.1 9.5 5.4
570.0 168.4 686.0 290.2 370.7 951.0 3036.3 3289.4
251.5 51.5 177.7 47.4 51.1 66.4
1484.7
0.35 0.195 0.160 0.075
~
497.9 1771.4 V/F= 0 2.39 (eq.3.23) 0.523 (eq.3-23) 1118 685 433 (eq.3-22) 0.292
~
V / F =1 2.11 (eq.3-24) 0.314 (eq.3-24) , 778 - 858 - 80 (eq.3-22) -0.0539
645.6 __ 813.8
V / F = 0.5 2.17 (eq.3-25) 0.475 (eq.3-25) 1016 814 202 (eq.3-22) 0.136
80
correct answer. On the other hand, on using the Lockhart-McHenry method, the convergence to the correct answer is very sharp: it is equivalent to the intersection of two lines at an angle of never less than 45". For a binary mixture of components A and B (the more volatile and less volatile, respectively), V can be solved by the following equation:
Lockhart and McHenry also offered the following equations for the multicomponent mixtures:
v* = Fl/( 1 - K,*)- F h / ( K? - 1)
(3-22)
for
ASSUMED
V/F
Fig. 3-10. Calculated V * / F versus assumed V / F . The convergence to the correct answer is very sharp, and is equivalent to the intersection of two lines at an angle of never less than 45'. (After Lockhart and McHenry, 1958, fig. 3; with permission of Hydrocurbon Processing, Gulf Publishing Co., Houston, Tex.)
81
The multicomponent feed F is resolved into moles of “light” component F, and moles of “heavy” component Fh (components with K , of 1 or less are classified as heavy and those with K,, of 1 or more as light). KT = equilibrium ratio for the light component in the hypothetical binary mixture; K z = equilibrium ratio for the heavy component; V* = moles of vapor calculated by eq. 3-22 for the flash-equilibrium of the hypothetical binary mixture; and u = subtotals of the light component and the heavy component. KT = and Kg are evaluated on using eqs. 3-23 to 3-27, involving the subtotals for the light and heavy components. Example 3-3 As shown in Table 3-111 (Lockhart and McHenry, 1958), calculations are made at assumed V / F values of 0, 1, and 0.5, where 0 and 1 correspond to the bubble point and dew point calculations. In Fig. 3-10, the intersection of the V* curve and 45O-line is shown. This approximate answer is usually correct to within *0.02 in V / F . Usually, the true answer can be determined in one more trial and never in more than two trials. The Lockhart-McHenry method also lends itself very well to computer programming. THE LOCKHART METHOD
In the Lockhart method, the first calculation is made at an assumed value of L , / F = 0.5 in eq. 3-20, which becomes: (3-28)
,: L, as compared with the assumed L, gives the direction of the correct L,. The C For example, if at an assumed value of L , / F of 0.5 the calculated Zym1 L , / F < 0.50, then the correct value of L , / F will be less than the calculated value of Z:- L , / F . If the’calculated c m 1 L , / Fequals 0.5 exactly, then this is the correct value of L,/F. On the other hand, if the calculated Z:-l L , / F exceeds 0.5, the correct value of L , / F will be greater than the calculated value of Z ,: L , / F . The second calculation should always be made to insure that there are two phases at the temperature and pressure of the flash equilibrium. If the CymlL , / F < 0.5, then the second calculation should be made at an assumed value of L , / F = 0, which is the dew-point calculation: (3-29) i-1
For some liquid to be present, this ratio (eq.3-29) must be greater than unity.
82 TABLE 3-IV Calculation of average
E
for the mixture
Condition
Equation no.
At L, = V ,
3-28
At L , / F = 0 (Dew-point)
3-29
At L , / F = 1.0 (Bubble-point)
3-30
At any value of V,
3-19
At any value of L,
3-20
Average
ZI;
If the Ci“,,L,/F > 0.5, then the second calculation must be made at an assumed value of L,/F = 1.0, which is the bubble-point calculation:
(3-30) 1-1
In order for some vapor to be present, this ratio (eq. 3-30) must be greater than unity. The third calculation at intermediate values of t ; / F (or L , / F ) may be made using either eq. 3-19 or 3-20, as desired. The proposed convergence technique uses the conventional calculations, but has one additional step at the end of each iteration to calculate the “average” K for the entire feed mixture. At the correct values of L , and t;, the average f equals unity. In a legitimate flash calculation, K > 1.0 at the bubble-point and K < 1.0 at the dew-point. The average K is calculated by using equations given in Table 3-IV. versus the For manual calculations, it is fast and easy to plot a graph of assumed L , (or t;). Sketching-in the curve through these points determines the next assumption to be made. The curve is revised if necessary, after each new point is calculated until satisfactory closure is attained. When using computers, it is convenient to have a convergence scheme and iterative technique built into the program. The hyperbolic convergence routine proposed by Hohmann and Lockhart (1972) is recommended.
83
Example 3-4
Calculate the lb-moles of vapor and liquid for the following conditions: Compound
Moles E
K.=
Methane Ethane Propane Isobutane nButane
5 5 20 35 35 100
19.0 4.0 1.5 0.70 0.55
v/x
At L, = V;, eq. 3-28 gives C:= L, = 52.42 moles. Therefore, correct L, > 52.42 moles and K = 0.908. One can check now whether or not vapor is present. Bubble-point calculation (eq. 3-30) gives X:,”,,F;K,= 189. Thus, vapor is present; K = 1.89. Correct value of L, lies between 52.42 and 100 moles. Try L, = 75.0 moles in eq. 3-20, which gives C:,,L, = 74.11 moles and K = 1.048. Correct value of L,lies between 52.42 and 74.11 moles. Try L, = 69.0 moles in eq. 3-20, which gives X ,-: L, = 68.95 moles and K= 1.002. Correct value of L,lies between 52.42 and 68.95 moles. Try L, = 68.0 moles in eq. 3-20, which gives Cy- Li = 68.08 moles and K = 0.996. Correct value of L, lies between 68.08 and 68.95 moles. This may be repeated until the desired closure is attained. The composition and amount of vapor is obtained from eq. 3-15.
, ,
THREE-PHASE FLASH EQUILIBRIUM
Three-phase systems (a vapor phase, a liquid hydrocarbon phase, and a liquid aqueous phase) are frequently encountered in petroleum operations. Formerly, it was assumed that the water vapor reduced the hydrocarbon partial pressure in the vapor according to Dalton’s law (Nelson, 1958). This assumption implied that the solubility of water in liquid hydrocarbons and solubility of liquid hydrocarbons in water are negligible. This approximation worked satisfactorily at low pressures and/or medium or high temperatures, but became increasingly less accurate at higher pressures and lower temperatures. Three-phase flash calculation techniques have been presented by Osborne (1964), Deam and Maddox (1969), Erbar (1973), and Peng and Robinson (1976). Two sets of equilibrium data are required: a set describing equilibrium between the vapor phase and each liquid phase. Equilibrium relationships between vapor and liquid hydrocarbon phases are available, whereas those between vapor and liquid aqueous phases are mostly present in company files and, hopefully, will be published. Furthermore, most of the published data deals with pure water, whereas the actual aqueous phases contain various amounts of different salts.
84
SAMPLE PROBLEMS AND QUESTIONS
(1) Draw a sectional view showing the interior construction of a typical oil-gas separator. ( 2 ) In two-stage separation, what effect does the increasing of pressure of high-stage separator have on the gasoline content obtained from (a) a low-stage separator, and (b) a high-stage separator. Explain. (3) Outline steps in determining the optimum pressure of the high-stage separa tor in two-stage separation. (4) Describe the physical principles involved in the design and operation of float-controlled separators. (5) Define stage separation. (6) A lube oil is treated with liquid propane in the proportions of 8.00 moles of propane to 2.00 moles of oil. The mixture is flashed at 90 psig and 120°F. Under these conditions the equilibrium ratio ( K = y / x ) for propane is ? (find it) and for the lube is 0.00. For 400.0 moles of the original mixture, calculate the numbers of moles and mole% of lube oil and propane in the equilibrium liquid phase. (7) (a) A liquid was subjected to an equilibrium-flash vaporization at 90 psia. At equilibrium, the analysis of the vapors on a mole% basis was as follows: propane (C,), 30; isobutane (iC4), 35; n-butane (nC,), 25; isopentane (iC5), 10. Estimate the temperature and the composition of the equilibrium liquid. (b) Estimate the maximum temperature at which the equilibrium vapors could be condensed practically completely at 90 psia. (8) The liquid (mixture of heptane and octane) is heated in a closed container at 212°F. The composition of the vapor produced is 70%by volume of heptane and 30% by volume of octane. The vapor pressures at 212°F are 18.42 psia for the heptane and 8.32 psia for the octane. The molecular weights are 100.2 for the heptane and 114.2 for the octane. The specific gravities of liquids at 212°F are 0.684 for the heptane and 0.704 for the octane. Compute the composition of liquid in volume per cent. [xl = (1 - K 2 ) / ( K 1- K 2 ) . ] (9) In determining the optimum pressure for the first stage of gas-oil separation when the second stage is atmospheric, certain graphs may be drawn. With the pressure of the high-stage separator as the horizontal axis, give illustrative curves of: (a) gas/oil ratio for low-stage and high-stage; also cumulative; (b) gasoline content of separator gas for low-stage and high-stage; and (c) total gasoline lost in two-stage separation. (10) For a wellstream having the following composition, determine the optimum second-stage pressure for a three-stage separation, if p1 = 600 psia. Component
Mole Fraction
Mol. Wt.
0.30 0.30 0.10 0.15 0.05 0.05 0.05
16.01 30.07 44.09 58.12 72.15 86.17 115.22
85
(Reference: Whinery and Campbell, 1958, p. 54.) (11) A 12-ft high vertical separator, with a 24-in. inside diameter, is operating at a pressure of 500 psia at 60°F. Compressibility factor = 0.909, oil gravity = 30"API, gas specific gravity (with respect to air = 1) = 0.65, and separation coefficient = 0.167. Determine: (1) gas specific weight in lb/ft3 at operating conditions; (2) oil capacity in bbl/day (retention time = 1 min); and (3) gas capacity in MMscf/day. (Reference: Craft et al., 1962, pp. 453-482.)
APPENDIX 3.1-RAOULT'S, DALTON'S AND HENRY'S LAWS
One can use Raoult's law in calculations involving normal paraffin hydrocarbons if the pressure is below 60 psi. According to Raoult's law, at any particular constant temperature, the partial pressure of one component of a mixture is equal to the mole fraction of that component multiplied by its vapor pressure in the pure state at the temperature of the liquid: p!
= PfXn
(3.1-1)
where p! = partial pressure of a component in the liquid phase, p," = vapor pressure of the component in the pure state, and x, = mole fraction of a component in the liquid phase. If a mixture is below its bubble-point temperature, the total pressure = p i + p i + p ; + . . . +p,.1 Dalton's law states that the partial pressure of an individual component in a gaseous mixture is equal to the product of the total pressure, n, and the mole fraction of that individual component, y,,:
P,' = VY,,
(3.1-2)
where p,' = partial pressure of a component in the vapor phase, 7~ = total pressure of the system, and y, = mole fraction of the component in the vapor phase. Raoult's law and Dalton's law can be combined, because at equilibrium the partial pressure of a component in the vapor is equal to the partial vapor pressure of the component in the liquid or p! = p:; consequently: y,, = P:xn/n (Henry's law)
( 3 .I-3)
This equation expresses the equilibrium between the vapor and the liquid of an ideal solution at any temperature and pressure. Example 3.1-1 shows calculations involving Henry's law.
86
The P;/m ratio is not a constant, however, and is altered by the total pressure and to some extent by the kind of materials associated with it in the mixture. Consequently, the y / x ratio must be determined experimentally (equilibrium ratio
K =y / x ) . Example 3. I-I A liquid consists of 42.5% butane and 57.5% pentane by volume at 60°F.If the liquid is heated to 180°F at 100 psia, what will be the composition of the vapor that is produced? (Assume that Raoult’s law holds.) Vapor pressure of C4H,, at 180°F is 152 psi; vapor pressure of CSHI2at 180°F is 56 psi; and sp.gr. at 60°F is 0.585 and 0.631 for C4H,, and C5H12, respectively. Solution: Component
C,H,, CSH,,
Vol. in liquid at 60°F (cc)
Sp.gr. at 60°F
42.5 57.5
0.585 0.631
Wt. in liquid
Mol. wt.
Grammoles in liquid
Mole S in liquid
58.1 72.1
0.428 0.504
46.0 54.0
(g)
24.9 36.3
,
0.9321oo.o
Vapor pressure (psi)
Raoult’s law partial pressure (psia) .
Mole S in vapor
152 56
69.9 30.3 100.2
69.R 30.2 100.0
If this problem were solved on the basis of 100 ft3, one would multiply 42.5 and 57.5 by 62.4 lb/ft3; and if on the basis of 100 gal, by 8.33 lb/gal. In each case, one should continue with the steps shown in the above table. The reader is also referred to an excellent treatment of the subject by Nelson (1958,pp. 434-464).
APPENDIX 3-II-ILLUSTRATION, ACCESSORIES, GAS CAPACITIES, SETTLING VOLUMES, AND SPECIFICATIONS FOR: (1) VERTICAL LOW-PRESSURE SEPARATORS; (2) VERTICAL HIGH-PRESSURE SEPARATORS; (3) HORIZONTAL LOW-PRESSURE SEPARATORS; (4) HORIZONTAL HIGH-PRESSURE SEPARATORS; A N D (5) SPHERICAL SEPARATORS (COURTESY OF HTI-SUPERIOR, INC., A BERRY INDUSTRIES COMPANY).
VERTICAL LOW PRESSURE SEPARATORS
GAS OUTLE
MIST EXTRACTOR
I\
5-
STANDARD ACCESSORIES CONTROL
lF'1i
I
FlrTE OUTLET
Standard accessories furnished with two-phase (oilgas) separators. 1 . Low pressure, diaphragm operated dump valve. 1 . Float operated level control 1 . ASME safetv relief valve 1 . Pressure gage with isolating valve 1 .Stainless steel wire mesh mist extractor 1 . Inlet diverter 1 . Drain connection 1 .Outside ladder on 10' high separators and higher 1 . Quieting baffle over settling section Companion flanges bolted on gas inlet and outlet (threaded or slipon) 1 .Tubular gage glass with safety cocks 8 draln valve
OPTIONAL ACCESSORIES DRA,N
Thermometer Safety Head Heating Coil Additional Connections Skid Mounting Three Phase Operation
88
VERTICAL LOW PRESSURE SEPARATOR INFORMATION GAS CAPAC:ITIES
wx20 W'XlS
WXl0
4W'xlS 49x10
36"xlO WXT.W,
M"Xl0 WXS
24"xT-9
24"xS
SPECIFICATIONS STANDARD AND ASME CODE CONSTRUCTION Model No.
Size Dia.x HI.
V.245 V-247 V.3010
24" x 5' 2 4 ' x 7v2, 30" x 1 0 36" x 5' 36" x 71h' 3 6 ' x 10' 48" x 1 0 48" x 15'
V-3610 V.4810 V-4815 V-M)lO V43015
W'X10 60" x 15' 80' x 2 0
Nominal W.P. psi
125 125 125 125 125 125 125 125 125 125 125
-
inlet 8 Gas Outlet Conn.
Oil Outlet Conn.
2" Thd 2'Thd 3" Thd 4'Thd 4"Thd 4" Thd 6' Fig 6" F.g 6' Fig 6 ' Flg 6" Flg
2'Thd 2'Thd 3'Thd 2'Thd 3'Thd 4" Thd 4" Thd 4" Thd 4" Thd 4" Thd 4" Th,d
Standard Valves Oil Gas Veive Valve
2'
2'
2"
2"
2'
2'
2"
2"
2, 2'
2' 2'
3" 4" 4" 4"
2" 2" 3" 3' 3"
4"
Normal volume carried in vessel is to center line of fiodt opening. Gas capacities shown are for maximum working pressure.
Settling Gas Volume Capacity bbl MMSCFD 0.65 1.01 2.06 1.61 2.43 3.04 5.67 7.86 9.23 12.65 15.51
1.9 3.1 5.7 4.3 7.1 8.3 14.6 17.3 23.1 27.0 32.9
Oil Shippin( Capacity Weight bbllday Ib
860 1290 2700 1960 2940 3920 8980 10460 10900 16400 21800
950 1150 2000 2wM 2350 2700 3400 4500 5200 6400 7600
89
VERTICAL HIGH PRESSURE SEPARATORS
GAS WTLET
-p&fJ-.SAFETY U' 'ALVE
la.
I
PRESSURE GAGE
STANDARD ACCESSORIES
THERMOMETER
FLOAT PROTECTOR
GAGE GLASS
LIQUID LEVEL CONTROL
Standard accessories furnished with twc-phase (oii.gas) separators, 1 . High pressure, screwed, angle type, diaphragm motor valve 1 . Pneumatic level control 1 . ASME safety relief valve 1 . Pressure gauge with isolating valve 1 .Control gas regulator set with fittings 1 . Stainless steel wire mesh mist extractor 1 . Inlet diverter 1 . Drain connection 1 . Outside ladder on 1 0 high separators and higher 1 . Quieting baffle over settling section Companion flanges bolted on gas inlet and outlet (threaded or slipon) 1 - Reflex gage glass with steel cocks 8. draln valve Additional accessories furnished with threephase (oil-gas-water) separators. 1 . High pressure, screwed, angle type, diaphragm motor valve 1 .Pneumatlc level control 1 . Transparent gage glass with ateel safety cocks & drain valve
OPTIONAL ACCESSORIES Safety Head Heating Coil Additional Connections Skid Mounting
90
CAPACITIES OF VERTICAL HIGH PRESSURE SEPARATORS
I
GAS CAPACITIES
200
300
400 XD
6W
8W
1MD
SEPARATOR OPERATING PRESSURE, PSI0
SETTLING VOLUMES SIZE O.D. x Ht. 16' x 16' x 16' x 10' x 20" x 20' x 24' x 24" x 24" x 30' x
30" X 30" x 36" x 38" x 36" x
5' 7H' 10 4'
7%' 10 5'
7%' 10 5' 1%'
10 7%' 10 1s
VOLUME
SIZE
bbl.
O.D. x Ht.
0.27 0.41 0.51 0.44 0.65 0.82
42" x 7%' 42" h 1 0 42" x 15' 48" x Vh' 48" x 1 0 48" x 1 s 54" x 7 H ' 54" x 1 0 54" x 15' 60'x 7'h' 60' x I 0 60" x 15' 60" x 2 0
0.66 0.97 1.21 1.13 1.64 2.02 2.47 3.02 4.13
B a d on loo0 psl W.P. Smparcltor
VOLUME bbl.
3.53 4.29 5.80 4.81 5.80 7.79 6.33 7.80 10.12 8.08 9.63 12.73 15.31
15W
1ooo
91
VERTICAL HIGH PRESSURE SEPARATORS SPECIFICATIONS Model No.
I
VS.1010.2 V52052 VS.207.2 , vs.2010-2 VS-2452 VS-247.2 VS-2410-2 VS.305-2 VS-307.2 VS-3010-2 VS-367-2 vs-3610.2 , vs-36152 V5427.2 VS-4210-2 VS.4215-2 VS-487-2 vs-4810.2 VS.4015.2 V5547-2 VS-5410-2 VS-54152 VsBo7.2 vs6010-2 VS60152 VS60202
-t
'451055 VS-107.5 VS-1010-5 VS-2055 v5207.5 VS-20105 VS.245-5 V5247-5 VS-2410-5 vs305-5 VS-307-5 VS-3010.5 VS-387-5 VS-36105 VS-30155 VS-427-5 V5-42105 VS-42155 v5487.5 V5481Q5 V548155
Size O.D. x Ht.
16~x10' 2O'XS ZO"x71h' 2O"XlO 24"xs 24"x7%' 24"XlO 3O"XV 3O"x7v2' 30"xlO 36"x74/2'
Nominal W.P. PSI
230
230
230
230
W'XlO
W'XlS 42"x74/2 42"xlO ~. 42"x15 46'X7M' 48"xlO 46'x15 54"x74/2' 54"xlO 54"x15 0O"x7'h'! 00"xlO 00"x15 8O'XZo' 16"x5' lWx71h' 16'Xlo' Zo"X5 2O'x7M" 2O'XlO 24"x5 24"x79h' 2r)"XlO W'XS 30" x 74/21 W'X10' W'x7M' W'XlO W'x15' 42"x7%' 42.~10' 42"x15 4WX7M' 48"XlO' W'X15'
230
230
230 230
500
500 500
500 500
500 500
Inlet LL Gas Outlet Conn.
Std. Oil Valve
Gas Capaclty MMCFD
2" Thd
1" 1" 1" 1" 1" 1" 1" 1" 1" 1" 1" 1" 1" 2"
2.1 1.7 2.9 3.4 2.5 4.1 4.0 3.9 0.5 7.0 9.4 11.0 13.7 12.7
3" FIa 3" FIO 3" Flg 3" FIg 3" Flg 3" FIg 4" Flg 4" Flg 4, Flg 4" Flg 4' Flg 4" FIg 6' Flg 6' FIa 6" FI; 6' Fla 6' F l i 6' F l i 6' FIa 6' FIQ 6' Flg 6 ' Flg 6 ' FIg 6' FIg 6,Flg
2" Thd 2 ' Thd 2" Thd 3" FIg 3" FIg 3" FIg 3" Flg 3" FIg 3" FI@ 4" FIg 4" Fla 4" FI; 4" FIa 4" FIO 4" Flp 6" FIg 6' FIg 6 FIg 6'FIg 6'FIg 8" FIg
2"
14.0
2,
18.4 16.7 19.5 24.5 21.1 24.6 30.7 26.1 30.5 38.1 43.4
2" 2" 2" 2" 2"
2, 2" 2" 2" 2" 1" 1" 1" 1" 1" 1" 1" 1" 1" 1"
1" 1" 1" 1" 1" 2"
1.0 2.7 3.1 2.8 4.3 5.0 3.7 0.1 7.1 5.7 9.3 10.9 13.3 15.5 19.4 18.4 21.4
2, 2,
28.8
2" 2" 2'
24.3 28.4 35.4
Llquld Capacity bbl/day
750 590 890 1180 850 1270 1700 1330
Shlpplng Weight Ib.
1150
1000
1200 1400 1200 1450 1700 110 XXK) 1750 2880 1900 2900 2300 3880 2700 58003500 3900 3100 52M) 3860 _-__ .... 7800 4800 5150 3700 8870 4800 lo300
esoo
esoo4800 8850 5700 13Ooo 7800 8 0 5 0 5 8 0 0 10700 7100 16100 8800 21400 12300
380
1000
540 720 580 870 1150 820 1230 1040 1280 lee0 2520 2700
1150 1300 1300 1500 1700 2100
2500
2900 2700 k3M __..
3800 4700
38005300 8500
5400 3750 5Ooo 7500 5 O o o 5
8800
ssqo
5200 8200 8200 8 0 0 7540 loso0
92
VERTICAL HIGH PRESSURE SEPARATORS SPECIFICATIONS Model NO.
I I
VS.54155 VS-607-5 VS-6010.5 vs-60155 vs-w20-5 VS-1656 VS-167-6 VS-1610-6 VS.2056 VS-207.6 VS-2010.6 VS.2456 VS.247-6 VS.2410.6 VS.3056 vs-307-6 VS-3010-6 VS.367.6 VS-3610-6 VS-38156 VS-427-6 VS-4210-6 VS.4215-6 VS.487-6 VS4810-6 VS-48156 VS-547-6 VS-5410-6 VS-5415-6 VS.807-6 VS-6010-6 VS.60156 V56020-6 VS-16510 VS.167-10 VS.1610-10 VS-20510 VS-207.10 vs-201 010 vs-24510 VS-247-10 VS.2410-10 VS-30510 VS.307-10 v53010-10 V5367.10 vs3610-10 vs-381510
Size O.D. x l i t .
54"xlS 6O'x7W' 60"x 1 0 60"xlS 6O"x20 W'X5 16"x71/2+ 16"xlO ZO"x5' ZO"x71h' 2O"XlO 24"x5 24"x71hh' 24"xlO 30"x5 30"X7M' W'x10 36"X7%' 36"xlO 36"x15 42"x71/2' 42"xlO 42"x15' 48"X71/z' 48"xlO W'x15 5493X71h9 54"xlO 54"x15' 6OSx7H' 60"xlO W'XlS W'x20' 16'x5' l6"x71hs 16"XlO' 20"xs 20"x71/2
Nominal W.P. psi
Inlet a Gas Outlet Conn.
Std Oil Valve
6" 6" 6' 6'
2'
500
600
600
800
600
600
600
600
600
600
1000
1000
20"XlO'
24"x5' 24" x7 V2' 24"XlO' W'XS W'x7H' 30"xlO W'x7H' 38"XlO' 38"XlS
loo0
loo0
loo0
Fig Flg Flg Flg 6 ,Flg 2" Thd 2 , Thd 2'Thd 3" Flg 3 , Flg 3 ! Flg 3" Flg 3 ' Flg 3 , Flg 4" Fig 4" FIa 4" F l i 4" Flg 4" Flg 4" Flg 6" Flg 6" Flg 6" Flg 6 ' FIg 6" Flg 6" Flg 6" FIg 6" FIg 6" Flg 6" Flg 8" Flg 6" Flg 6" Flg 2" Thd 2" Thd 2 ' Thd 3" Flg 3" FlQ 3" Flg 3" FIg 3" Flg 3" Flg 4" FIg 4" Flg 4" Flg 4" Flg 4" Flg 4" Flg
2, 2' 2'
2, 1" 1" 1"
1" 1"
1" 1" 1" 1" 1" 1" 1"
1" 1"
1"
22' 2'
2" 2" 2, 2, 2"
2, 2' 2"
2"
2, 1" 1" 1"
1" 1" 1" 1" 1"
1" 1"
1" 1"
1" 1" 1"
Gas Capacity MMCFD
44.6 38.1 44.5 55.6 63.3 1.8 3.0 3.5 2.8 4.6 5.4 3.6 6.3 7.3 6.0 9.6 11.5 14.7 17.2 21.5 20.4 23.8 30.0 27.1 31.7 39.6 34.0 39.6 49.6 42.3 49.4 61.7 70.3 2.4 3.9 4.5 3.7 6.1 7.1 5.3 8.6 10.4 8.2 13.8 15.9 20.7 24.1 30.1
Liquid Capaclty bbl/day
Shlppli Weigh1 Ib.
12400 7700 10300 15400
12500 9500 11500 15800'
20600
zoo00
360 540 720 560 840 1120 760 1140 1520 1200 1800 24W 2700 3600 5400 3750 5000 7500 5000 6600 9900 6200 6300 12400 7700 10300 15400 20600 340 500 670 530 790 1050 760 1140 1520 1180 1760 2350
1100 1250 1400 1400 1600 1800 2200 2600 3000 2800 3400 4000 4900 5500
2680 3570 5360
6800 5600 6700
8900 8400 8200 11800
eooo loo00 13900 10300 12500 17000 21500 1100 1200 1500 1800 1900 2200 2500 2650
3300 3200 3650 4200 5400 6400 8700
93
VERTICAL HIGH PRESSURE SEPARATORS Model No.
Size O.D. x Ht.
I VS-307.12
30"x7'/2' 3O"xlO 36"x71/2' 36"xlO' 36"xlS 42"x71/2' 42"xlO' 42"xlS 4B"X71/2' 48"xlO 46"x15' 54"x7'/2' 54"xlO 54"xlS 60"x71/2' 60"xlO' 60"x15' 60"X20 16"XS 16"X7 1/2 ' 16"XlO' 20'xS 20"x7Vz" 20"XlO 24"x5'
VS.3010-12 VS.367.12 VS.3610-12 VS.3615-12 VS-427.12 VS.4210-12 VS-4215.12 VS.467.12 VS-4810.12 VS-4815.12 vs.547.12 VS.5410-12 VS5415-12 VS-607-12 VS.6010-12 VS.6015-12 VS-6020.1 VS.165.14 VS-167.14 VS.1610.14 VS-205.14 V5207.14 VS.2010-14 VS.24514
Nominal W.P. psi
Inlet 8 Gas Outlet Conn.
1200
4" 4" 4" 4" 4"
1200
1200
I200
1200
Flo Flg Flg Flg Flg 6 ,Flg 6" Flg 6" Flg 6" Fig 6" Flg 6" Fig 6 ,Flg 6 ' Flg 6" Flg 6" FIa 6" Fla 6" Fla
Std. Oil Valve
1" 1" 1"
1" 1"
2, 2" 2"
2, 2" 2'
2, 2' 2" 2" 2'
2s
Gas Capacity MMCFD
Liquid Capacity bblldav
15.3 17.9 23.1 26.9 33.6 31.0 36.1 45.2 40.5 47.3 59.1 51.4 59.9 74.9 ._ . 62.3 72 7 90.9 104.0 -. 2.9
1760 2350 2610 3470 5220 3510 4680 7020 4650 6200 9300 5650 7600 11700 7250 9700 14550 19400 320 460 640 450 670 900 740 1100
Shippini Weight Ib.
3950 4500 6000 7300
9900 8300 9900 13100 11OOO 13500 18400 14400 17500 23500 18OOO 21500 29000 36000 1500 1800 2100 2100 2600 3100 2800 3200
9 2, I
1440
1440
1440
2" Thd 2 ' Thd 2" Thd 3 , FIa 3 ' Fla 3" Fla 3" Flg 3" Flg
1" 1" 1" 1" 1" 1" 1" 1"
4.8
5.4 4.1 6.7 7.6 6.7 11.2
94
HORIZONTAL LOW PRESSURE SEPARATORS
STANDARD ACCESSORIES Standard accessories furnished with two.phase (oil.gas) separators. 1 . Low pressure, diaphragm operated dump valve. 1 . Float operated level control 1 . ASME safety relief valve 1 . Pressure gage with isolating valve 1 .Stainless steel wire mesh mist extractor 1 . Inlet diverter 1 . Drain connection 1 - Quieting baffle Companion flanges bolted on gas inlet and outlef (threaded or slipon) 1 -Tubular gage glass with safety cocks & drain valve
OPTIONAL ACCESSORIES Thermometer Safety Head Heating Coil Additional Connections Skid Mounting Three Phase Operation
95
H0 R E 0NTAL LOW PR ESSURE SEPARATOR INFORMAT10 N iu
GAS CAPACITIES
60 x20
60'r15 60 '"10
4W'xlS 48"x10
36x15
36 " 1 0
30x10 30'XT.6'
30' x
s
24 X10 24 x 7 . 6
24.35
I PO
! ! ! ! ! ! l ! ! ! ! ! ! ! ! . ' ! I III.IIII 1 1 1 1 1 1 I 15
20
30
1
Model No.
Size Dia. x Lgth.
H.247 H-2410 H-305 H-307 H-3010 H-3810 H-3815 H-4810 H-4615 H.8010 H.6015 H.6020
24"x7%' 24"xlO' 3O"x5' 3O'x7H' 3O"xlO 38"xlO 36'x15 48"xlO 46'x15 6O"xlO lO"x15 60"x20'
40
! L ' ! ' ! ! ! ! ! ! ! ! ! ! ! W '
I I I I I 1111111II
1
50 60 7 0 8 0 9 0 1W 125 ~
1
~~
SEPARATOR OPERATING PRESSURE PSlG
Nominal W.P. psi 125 125 125 125 125 125 125 125 125 125 125 125
inlet a Gas Outlet Conn. ~2"Thd 3"Thd 3"Thd 3"Thd 4"Thd 4"Thd 4" Thd 8" Flg 6' FIg 6" Fig 8" Fig 6" Fig
Oil Outlet Conn. 2'Thd 3"Thd 3'Thd 3" Thd 4" Thd 4"Thd 4" Thd 4"Thd 4"Thd 4"Thd 4"Thd 4" Thd
Standard Valves Oil Gas Valve Valve
-
-
2" 2' 2'
2" 2"
3" 3" 3"
2, 2'' 2' 2" 2' 2' 3' 3" 3"
4"
4"
2" 2"
2" 2' 2'
Gas Capacity MMSCFD
5.3 8.0 7.4 8.4 9.4 13.6 16.5 18.5 22.4 38.0 48.0 53.3
Oil Capacity Bbllday
2900 3850 3ooo 4500 Boo0
Shipping Weight Ib. 1200 1800 1200 1800 2100
88002800 3800 3500
13200 lsB00 23700 24500
36800 4oooo
4600 8200 8100 loo00
'
'
96
HORIZONTAL HIGH PRESSURE SEPARATORS SAFETY VALVE
THERMOMETER INLET
RESSURE GAGE
MIST EXTRACTOR NLET DIVERTER
LIQUID
LEVEL CONTROL
DUMP VALVE
STANDARD ACCESSORIES Standard accessories furnished with two-phase (oil.gas) separators 1 . High pressure screwed, angle type, diaphragm motor valve 1 . Pneumatic level control 1 . ASME safety relief valve 1 . Pressure gage with isolating valve 1 . Control gas regulator set with fittings 1 . Stainless steel wire mesh mist extractor 1 . Inlet diverter 1 Drain connection 1 . Outside ladder on 10' high separators and higher 1 . Quieting baffle Companion flanges bolted on gas inlet and outlet (threaded or slipon) 1 Reflex gage glass with steel cocks 8 drain valve Additional accessories furnished with three-phase (oil-gaswater) separators 1 . High pressure, screwed, angle type, diaphragm motor valve 1 . Pneumatic level control 1 Transparent gage glass with steel safety cocks & drain valve
OPTIONAL ACCESSORIES Thermometer Safety Head Heating Coil Additional Connections Skid Mounting
LIQUID OUTLET
StlOlVtlVd3S 3tlflSS3tld H U H WlNOZltlOH A 0 S31113VdV3 L6
98
HORIZONTAL HIGH PRESSURE SEPARATORS
I I I
I '
Model No.
Slre
HS-1210-2 HS-1852 HS-187-2 HS-1610-2 HS-205-2 HS-207-2 HS-2010-2 HS-2452 HS-247-2 HS-2410-2 HS-24152 HS-3052 HS-307-2 HS-30102 HS-30152 HS-367-2 HS-3810-2 HS-36152 HS362&2 HS-427-2 HS-42102 HS-4215.2 H.s-4220-2 HS-487-2 HS4810-2 HS-48152 HS-4820-2 HS-547-2 HS-5410-2 HS-54152 HS-542&2 HS807-2 HS8010-2 HS80152 HS802G2 HS125-5 nsi27.5 HS-12165 HS-1655 HS-187-5 HS-1810-5
12%"x10' 1WX5' lWx7H' 16'Xlo' 2O"x5' X)"x7%' 20"Xlo' 24"x5' 24"x7*r' 24"xlO 24"x15' W'x5' W'x7H' W'X10' W'x15' 38"x7%' W'X10' 36"xW 36"X20' 42"x7%' 42"xlO' 42.~15' 42"xX)' W'x7H' 48"XlO' W'x15'
nszw HS-207-5 HS-2010-5 HS-265 HS-247-5 ns.2410-5 H82415.5
O.D.
X
Lgth
Nominal W.P. psi
Inlet a Gas Outlet Conn.
2 , Thd 2 , Thd 2 , Thd
2,
230
2" Thd 230
3" Thd
2, 2,
230
230
230
230
230
48"W 54"x7M9 54"XlO 54"xlB W'X20' BO"x7H' 60"XlO W'XlS 60"x20" 12~x5' 12W"x7H' 12rk"xlO Wx5' lV'x7M' l6'XlO' 20"X5 20"X7H' 20"XlW 24"x5' 24"xIH' 24"xlO' 24"r15'
230
230
500 500
500 500
3" Thd 3" Thd 4" Thd 4" Thd 4" Thd 4" Thd 4,Thd 4" Thd C' Thd 4" Thd 6" FIg 8" FIg 8" Fla 8" Fig 8" Fig 8" Flg 8" Fla 8" FIg 8" Flg 8" FIQ 8" Fla 8" FIg 8" Fla 8" FIE 8" FIa 8" Fig 8" Fig 8" Fig 8" Flg 8" Fig 2" Thd 2" Thd 2" Thd 2 fhd 2,Thd 2" Thd 3" FIg 3" Flg 3" FIg 4" FIg 4" FIg 4" FIg 4" Fio
Std. Oil Valve
2' 2"
2"
2, 2" 2" 2"
2" 2 2" 2" 2" 2"
2" 2"
2 1 2, 2 2" 2' 2" 2" 2" 2" 2"
2, 2" 2" 2"
2' 2" 2" 1"
2, 2" 2" 2" 2" 2"
2, 2" 2" 2" 2" 2"
Gas Capacity MMCFD'
2.1 2.7 3.1 3.5 4.4 4.9 5.5 6.3 7.0 7.9 9.6 9.8 11.1 12.4 15.0 16.1 18.1 21.8 25.0 21.7 24.5 29.5 33.0 28.5 33.3 38.9 44.5 36.0 40.5 49.0 56.9 44.8 50.2 60.7 70.5 2.5 2.8 3.2 4.0 4.5 5.1 8.4 7.2 8.1 9.2 10.3 11.8 14.0
Liquid Capacity Bbllday
Shipping Weight Ibs.
1000 840 1260 1880 1330 2000 2850 1900 2850
1000 1000 1100 1200 1200 1300 1400 1300 1450 3800 1550 5700 1700 3ooo 1500 4500 1800 8ooo 2200 9 o o o 2 8 0 0 8 m 2400 8700 2800 13ooo 3800 17000 4800
8800
3300 3900
11700 17800 23500 11500 15400 23000
5200 8800 4200 5100 7000
eooo
32000 14800 19400 29200
5500
8800 Baoo 11000 8800
38800 18Ooo 24wo
85M)
11200 14000 800 1000 1100 1400 1500 1800 1800 2050 2400 2100 2800 3100
38ooo 48OOO
500 750 1000
800 1200 1800 1300 1850 2800 1850 2750 3700
5
5
M
m
99
HORIZONTAL HIGH PRESSURE SEPARATORS SPECIFICATIONS
I I I
Model
Size
No.
O.D.x Lgth
HS-4210.5 HS-4215-5 HS.42205 H5487.5 H548105 HS-4815-5 HS48205 H5547.5 HS-5410-5 HS-54155 HS-54205 HS507-5 HS.6010-5 HSS0155 HS60205 HS-1256 HS-127-6 H51210.8 H51856 H51675 HS-1810.8 H5-205-8 HS207-6 HS2010.8 H5245.6 H5247-8 HS2410-6 H52415.6 HS-3058 HS307-8 HS-30108 H530156 HS367-6 H53610.8 H53615-6 HS-36208 HS-427-6 HS4210.8 H54215.6 HS-42206
42"XlO 42"x15' 42"x20 48Vx7M' 48'xlO 48"x15' 48"x20' 54"X7'/2 54"xlO 54"x15' 54"x20' 6O"x71h' 6O"xlO 60"x15' 6O"x2O 12%"x5' 12%"x7H' 12%"x10' 18"XS 16"x71h' 16'xlO' ZO"x5' 20"x7%' 2O"Xlo' 24"xS 24"x71h' 24"XlO 24"x15' 30"x5' 30"X7H"
Nominal W.P. psi
500
500
500
800
800
600
600
600
W'XlO
30"x15' Wx7H' 38"XlO' W'x15' 36"x20 42"x7H8 42"x 10' 42"x15' 42"xZO'
600
600
Inlet 8 Gas Outlet Conn.
Oil Valve
8" FIO
2"
6" FIR 8" FIO 8' FIa 8" FIO 8" Fla 8' Flg 8" Fla 8" F l i 8" FIR 8" FlO 8' Flg 8, FIg 8" FIg 8" Flg 2 , Thd 2" Thd 2" Thd 2 , Thd 2 , Thd 2" Thd 3" FIg 3 ' FIg 3" Flg 4" FIg 4" FIg 4" Flg 4" Flg 4" Flg 4" Flg 4" FIo 4" FIO 6' Fin 6" FIg 8" FIg 8" FIg 6" FIg 6" Flg 8" FIg 6" FIG
2"
2, 2" 2, 2, 2" 2, 2" 2"
2' 2" 2" 2"
21" 1"
1" 1" 1" 1"
1" 1" 1" 1" 1" 1" 1" 1" 1" 1" 2' 2" 2"
2, 2" 2"
2" 2"
2"
Gas Capaclty MMCFD'
35.1 42.5 49.4 41.3 48.5 58.2 85.3 52.0 58.5 70.8 82.3 85.0 73.0
Llquld Capaclty Bbllday
11200 16800 22400 11100 14800 22200 %MI 14ooo
lssoo
28000 37200 17400 23200
88.0
34800
102.0 2.8 3.1 3.5 4.5 5.1 5.7 7.0 7.8 8.8 9.5 10.8 12.0 14.5 14.9 18.7 18.8 22.7 25.1 26.2 34.1 39.8 34.7 39.0 47.2 54.9
48400
500 750 1000 800 1200 1800 1260 1800 2500 1700 2550 3400 5100 2700
4ooo 5400 8100
eo00
Shipping Welght Ibs.
7800 loo00
12200
8OOO 8800 13500 17200 9700 11800 17000 21200 12800 15100 20100 25100 lo00 1100 1200 1500 1600 1600 1700 2150
2800 2350 2700 3200 3700 2700 3700
4800 6800 5100 8200
8OOO 12Ooo
8500
8400
lo800 6soO
leooo
11200 lee00
22400
8300 loB00 12900
100
HORIZONTAL HIGH PRESSURE SEPARATORS SPECIFICATIONS Model No.
Size O.D. x Lgth
HS-207.10 HS-2010.10 HS.245-10 HS.247.10 HS-2410-10 HS-2415-10 HS-305-10 HS.307.10 HS-3010-10 HS-3015.10 HS.367-10 HS-3610-10 HS4615.10 HS.3620-10 HS-427.10 HS-421010 HS-4215-10 HS.422010 HS.467-10 HS.4810-10 HS.4815-10 HS-4820-10 HS-547-10 HS5410-10 HS-541510 HS-5420.10 HS-807-10 HS.6010-10 HS-801510 HS-6020-10
20"x71hh' 2O"xlO 24"xS 24'%71/z' 24"xlo' 24"x15' 303x5' 30"X71h' 30"xlO' 3@'~15' 36"x71/z * 38"xlO' 36"X15' 36k20 42"X77h' 42"XlO 42"X15' 42"~20' 48"x71h' 48"xlO' 48"xlS 48"~20' 54"x7M' 54"XlO 54"~15 54"xZO 6O'x7H' 60"XlO 6O"XlS 60"xZO
Nominal W.P.
'
lo00
lo00
lo00
loo0
lo00
lo00
lo00
Inlet 8 Gas Outlel Conn.
3" 3" 4" 4" 4" 4" 4" 4" 4" 4"
FIO Flg FIg FIg Flg Flg FIg Flg FID Flg 6" Flg 0 ' Flg 0 ' FIg 6" FIg 6" Flg 6" FIg 8" Flg 6'' Flg 8" FIg 8" Flg 8" Flg 8" Flg 8" FIg 8" FIg 8" Flg 8" Flg 8" Flg 8" FIg 8" FIg 8' FIg
Std.
Gas
Valve
MMCFD
1" 1" 1" 1" 1" 1" 1" 1" 1" 2" 2'
2, 2"
2, 2"
2, 2' 2"
2, 2" 2' 2"
2, 2"
2, 2, 2'
2"
.2# 2"
10.3 11.6 13.3 15.0 16.8
20.3 20.5 23.1
Liquid Capacity
1770 2380 1700 2550 3400 5100 2800
3800
26.0
5200
31.4 35.0 39.4 47.4 55.4 48.7 52.5 63.5 73.6 62.5 70.3 85.0 88.8 78.2 88.0 108.4 123.6 97.8 110.0 113.0 154.0
7900 8ooo 8ooo 1 m lso00 8ooo 10700
leooo
21400 10700 14300 21400 28800 13400 17900 28800 35800 18800 22400
Shipping
2300
2800' 2200
3OOO 3800 5400
3200 4300 5500 7800 8100 7500 10200 1 m
8200
es00 13400 18900 logo0
12700 17500 22100 13400 lso00 21200 2&(00
18700 les00
3380026400 32800
44800
101
HORIZONTAL HIGH PRESSURE SEPARATORS Model
Size
NO.
0.0. x Lgth
HS-121012 HS-165.12 1 HS-167-12 HS-1610.12 HS.20512 HS.207-12 HS.2010-12 HS-24512 HS-247.12 HS-2410.12 HS.2415.12 HS-30512 HS-307-12 HS-3010-12 HS.3015-12 HS.367.12 HS.3610-12 HS.3615-12 HS.3620-12 HS-427.12 HS.4210-12 HS.421512 HS.4220-12 HS.487-12 HS.481012 HS-4815.12
'
l2W"xlO 16"x5' 16"x71/2' 16"XlO' 2O"x5 20"x71/2' 2O"xlO 24"x5' 24"x7 7/2 ' 24"X 10' 2 4 " 15' ~ 3O"x5' 3O"x7V' 3O"xlO' 30"X15' 36"x7'/2' 36"XlO 36"X15' 36"x20 42"x7M' 42"XlO 42"~15' 42"X2O 48"X7'h' 48"XlO 48"X15'
Nominal W.P. psi
1200
1200
1200
1200
1200
1200
1200
Inlet 8 Gas Outlet Conn
2" Thd 2" Thd 2' Thd 2" Thd 3" Flg 3 ' Flg 3" Flg 4" Flg 4" Flg 4" Fig 4" Fig 4" Flg 4" Flg 4" Flg 4" Flg 6" Flg 6" Flg 6" Fla 6" Fl; 6" Flg 6" Flg 6" Flg 6 ' Fig 8" Fig 8" Flg 8" Fig
Std. Oil Valve
1" 1"
1" 1"
1" 1" 1" 1" 1" 1" 1" 1"
1" 1"
2" 2" 2" 2" 2" 2" 2"
2"
2' 2" 2"
2%
Gas Capacity MMCFD
5.0 6.3 7.1 6.0 9.9 11.1 12.5 14.7 16.5 18.6 22.5 23.2 26.1 29.4 35.5 40.0 44.3 53.6 62.3 53.0 60.0 72.0 83.7 69.3 79.9 94.2
Liquid Capacity Bbllday
Shipping Weight Ibs.
900 710 1070 1430 1120 1660 2250 1670 2500 3300 5000 2600 4000 5300 7900 6000
1450 1500
'8000 12000 16000 6000 10700 16000 21400 10500 14000 21000
1800 2150 1650 2400 2900 2600 3100 3900 5400 3600 4600 5700 7600 6500 7900 10600 13300 8700 14300 18000 11700 13700 18500
-
102
HORIZONTAL HIGH PRESSURE SEPARATORS Model
Size
NO.
O.D. x Lgth
HS.2415.15 HS-305-15 HS-307.15 HS.3010.15 HS.3015-15 HS.367.15 HS.3610.15 HS.3615.15 HS-3620.15 HS-427.15 HS-4210-15 H5.42 15.1 5 HS-4220-15 HS.487.15 HS.4610.15 HS.4815.15 HS4620.15 HS-125.20 HS.127.20 HS-1210-20 HS.165-20 HS-167-20 HS-1610.20 HS.205-20 HS-207.20 HS.2010.20 HS.245.20 HS.247.20 HS-2410.20 HS.2415-20
24"x15' 3O"XS 3O'x7%' 3O"xiO' 30"x15' 36" x7 '/2 36"xlO' 36"x15' 36"x20' 42"x71/2' 42"xlO 42'x15' 42"x20' 48',x77/2 ' 48"xlO' 46"x15' 48"x20 12W'x5' 123/4*'x7'/2 ' 123/r"x108 16"x5 16" x7 V 2 ' 16"x'10' 20"x5' 20'1x71/z' 20"XlO' 24"x5' 24"x71/2' 24"x 10' 24"x15'
Nominal
Inlet B
Std.
W.P. psi
Gas Outlet Conn.
Oil
4 ,Flg 1500
1500
1500
1500
2000
2000
2000
2000
4" 4" 4" 4" 6' 6' 6' 6" 6" 6" 6" 6"
Flg Flg Flg Flg Flg Flg Flg Fig Flg Flg Flg Flg 8" Flg 8" Flg 8" Flg 6" Flg 2" Flg 2" Flg 2" Flg 2" Flg 2" Flg 2" Flg 3" Flo 3" Flg 3" Flg 4" Flg 4" Flg 4" Flg 4" Flg
Valve
1" 1" 1" 1" 2"
2'
2' 2' 2' 2"
2" 2" 2" 2" 2" 2" 2" 1" 1"
1" 1"
1" 1" 1" 1" 1" 1" 1" 1"
1"
Gas Capacity MMCFD
Liquid Capacity Bbllday
Shipping Weight
26.2 27.1 30.5 34.3 41.5 43.9 49.4 59.7 69.4 60.0 67.0 81.o 94.5 76.0 86.0 106.0 123.5
5000 2600 3900 5200 7900 5700 7600 11400 15200 7700 10300 15400 20600 10000 13500 20000 27000 400 600 800 590 880 1180 1000 1500 2000 1570 2260 3040 4500
5900 4100 5200 6300 8500 7300 6800 11800 14800 9600 11600 15900 20000 13300 16100 21700 27500 2000 2200 2400 2200 2600 3000 3000 3600 4300 3600 4700 5800 8000
5.0 5.7 6.4 7.4 8.3 9.3 12.6 14.1 15.9 19.0 21.4 23.4 30.0
Ibs.
103
HORIZONTAL HIGH PRESSURE SEPARATORS SPECIFICATIONS ~
~
Model No. HS-305.20 HS.307.20 HS.3010.20 HS-3015-20 HS.367.20 HS.3610-20 HS.3615.20 LHS-3620-20
size O.D. x Lgth 30"xS 30'X7'h' 3O"xlO 30'x15' 3SS'x7Vz' 36"XlO 36"xlS 36'x20
Nominal W.P. psi
inlet 8 Gas Outlet Conn.
Std. Oil Valve
Gas Capacity
2000
4" 4" 4" 4"
2000
6" Flg 6 ' Flg
1" 1" 1" 1" 1" 1" 1" 1"
Fig Fig Fig Fig
6 ' Flg 6" Flg
MMCFD
Liquid Capacity Bbllday
Shipping Weight Ibs.
31.0 34.8 39.2 47.3 44.6 50.2 56.4 66.3
2460 3720 4960 7440 5350 7100 10700 14200
5800 7400 8700 11600 9600 12200 16400 20800
SETTLING VOLUMES Slze O.D.
Settling Volume, bbl. X
LQth.
12%" x 5' 12%" x 7 M V 12%" x 1 0 1s" x 5' 16' x 7%' 16' x 1 0 20" x 5' 20' x 7M' 20' x 10 24" x 5' 24" x 7Vz' 24" x 1 0 24" x 15' 30" x 5' 30, x 7'h' 30' x 10' 3 0 ' x 15' 3 6 ' X 7V2' 36" x 1 0 36" x 15' 36' x 20 42" x 7%' 42" x 1 0 42' x 15' 42' x 20' 48" x 7%' 48" x 1 0 46' x 15' 48" x 20' 54" X 7l/2' 54" x 1 0 54" x 15' 54" x 20' 60, x 7 % 60' x 1 0 60' x 15' 60' x 20'
M Full
0.38 0.55 0.72 0.61
0.88 1.14
0.98 1.39 1.80 1.45 2.04 2.63 3.61 2.43 3.40 4.37 6.30 4.99 6.38 9.17 11.96 6.93 6.83 12.62 16.41 9.28 11.77 16.74 21.71 12.02 15.17 21.49 27.61 15.05 18.93 26.68 34.44
W Full 0.22 0.32 0.42 0.35 0.50 0.66 0.55 0.79 1.03 0.83 1.18 1.52 2.21 1.39 1.96 2.52 3.65 2.67 3.68 5.30 6.92 3.98 5.09 7.30 9.51 5.32 6.77 9.67 12.57 6.87 8.71 12.40 16.08 6.60 10.86 15.38 19.90
1/4
Full 0.15 0.21 0.28 0.24 0.34 0.44 ,038 0.54 0.70 0.55 0.76 1.01 1.47 0.91 1.29 1.67 2.42 1.90 2.45 3.54 4.63 2.61 3.35 4.63 6.32 3.51 4.49 6.43 8.38 4.49 5.73 6.20 10.68 5.66 7.17 10.21 13.24
-
104
SPHERICAL SEPARATORS
HTI Hydrotek's Spherical Separator utilizes a tangentiai inlet to increase flow velocity as the wellstream enters the separator. The fluid is directed against the inside of the sphere for initial separation in the annular space. Oil is forced against the shell as the gas moves to the center. The gas makes two complete directional reversals before entering the mist extractor. The liquid drains to the liquid accumulation area. A large ratio of liquid surface area to liquid volume assures rapid release of solution gas.
FEATURES No unused or "Dead" space. A baffle arrangement separates the liquid and gas sections to provide a quiet liquid surface for quick gas release and proper liquid level control operation. Weii planned orientation of fittings and connections facilitates fast hook-up. Separators available for three-phase operation (gas, oil and water) and for handling extremely foamy oil.
105
SPHERICAL SEPARATORS
STANDARD ACCESSORIES
OPTIONAL ACCESSORIES
Thermometer Standard accessories furnished with two-phase (oil-gas) separators. Safety Head 1 . High pressure, screwed, angle type, diaphragm motor valve Heating Coil 1 - Pneumatic ievei control Additional Connections 1 .ASME safety relief valve Skid Mounting 1 . Pressure gauge with isolating valve 1 . Control gas regulator set with fittings 1 . Stainless steel wire mesh mist extractor 1 . Tangential inlet diverter 1 . Drain connection 1 . Quieting baffle over settling section Companion flanges bolted on gas inlet and outlet (threaded or slipon) on 250 P S I . & higher 1 . Reflex gage glass with steel cocks & drain valve (tubular gage glass on 125 P.S.I. separators) (Mechanical gas valve same size as oil valve furnished on 125 PSI separators) Additional accessories furnished with three-phase (oil-gas-water) separators. 1 . High pressure, screwed, angle type, diaphragm motor valve 1 . Pneumatic level control 1 . Transparent gage glass with steel safety cocks & drain valve 1 . Thermometer with S.S.
106
cn a 0
3a e W
cn
a0 z W
I
e
cn
& cn W
I
k
0
2a
0
cn
W
K 3 v) v)
W
a
n
3
9
a
(3
107
SPHERICAL SEPARATORS SPECIFICATIONS Model No.
Nominal Diameter
W.P. psi
88-4212 ss-4812 ss.5412 55.4225 SS.4625 SS.6025 55.2450 55.3050 SS.3650 88-4250
42' 48"
125
__
SSa850
558050
S52480 ss-3oBo _... SS-3660 SS.4260 SS-4660 SS-6060 SS.24100 SS-30100 5s-36100 SS-42100 SS.46100 SS.66100 55.24120 85-30120 55.36120 5542120 SS-48120 ss.60120 SS-24144 SS-30144 85-36144 88-42144 88.46144 ss-80144 88.24200 ss.30200 SS-36200 ss.42200 SS24300 S530300 S536300 SS.42300
__
iniel 8 Gas Outlet
54'
42" 483 60, 24" 30' 36" 42' 48' 60'
250
24' 30' ..
600
36' 42' 48" 60, 24" 30" 36' 42, 48" 60' 24" 30" 36" 421 48" 60" 24, 303 36' 423 48" 60" 24" 30' 36'. 42' 24" 30"' 36' 42'
500
Std. Liquid Valve
Gas Cap. MMCFO
Oil Cap. bbidday
Approx. Weight, Lb
2" 3' 4" 2" 2' 2" 1" 1" 1" 2, 2' 2, 1"
6.2 9.0 12.0 6.4 12.0 22.0 3.6 6.9 10.0 12.0 18.0 31.0 4.0 7.6 11.0 14.0 20.0 33.0 5.1 10.0 15.0 18.0 25.0 45.0 5.7 11.0 16.0 20.0 28.0 49.0 6.5 13.0 19.0 22.0 31.0 54.0 7.4 15.0 21.0 26.0 6.3 16.0 23.0 26.0
1900 2500 4500 1900 2500 5100 300 600 1100 1900 2500 5100 300 600 1100 1900 2500 5100 300 - 600 1100 1900 2500 5100 300 600 1100 1900 2500 5100 300 600 1100 1900 2500 5100 300 600 1100 1800 300 600 1100 ls00
1000 1300 1700 1100 1400 3400 1000 1200 1500 2300 3100 3400 1200 1300 1600 2400 3200 3600 1300 1400 1600 2600 3700 4300 1400 1500 1900 2900 3800 4700 1500 1800 2400 3300 4100 5400 1600 2200 2600 3900 2600 3200 3800 4900
1"
1" 2, 2" 2' 1" 1"
1 000
1"
2" 2" 2, 1"
1200
1"
1" 2" 2, 2,
1440
1"
1" 1"
2' 2" 2, 2 m
1"
1" 1" 2, 1" 1" 1"
3000
Nominal capacities based on separator
2 q
V2
tuil of liquid.
All separators, 250 psi and above are ASME code constructed and stamped.
.
108 REFERENCES API, 1960. Specification of Oil and Gas Separators. (Tentative API Study 12J). API, Division of Production, 1st ed., Dallas, Tex., p. 4. Campbell, J.M., 1976. Gas Conditioning and Processing. Campbell Petroleum Series, Norman, Okla. Chilingar, G.V. and Beeson, C.M., 1969. Surface Operations in Petroleum Production. Am. Elsevier, New York, N.Y., 397 pp. Craft, B.C., Holden, W.R. and Graves Jr., E.D., 1962.Well Design, Drilling and Production. Prentice-Hall, Englewood Cliffs, N.J., 571 pp. Deam, J.R. and Maddox, R.N., 1969. How to figure three-phase flash. Hydrocarbon Process., July: 163-164. Erbar, J.H., 1973. Three-phase equilibrium calculations. N.G.P.A. 52nd Annu. Conv., pp. 62-70. Hohmann, E.C. and Lockhart, F.J., 1972. Remember the hyperbola. Chemrech, Oct.: 614-619. Ikoku, Chi U., 1980. Natural Gar Engineering, A Systems Approach. PennWell, Tulsa, Okla., 788 pp. Lockhart, F.J. and McHenry, R.J., 1958.Figure flash equilibrium easier, quicker this way. Pet. Refiner, 37 (3): 209-212. Nelson, W.L., 1958. Petroleum Refinery Engineering. McGraw-Hill, New York, N.Y., 4th ed., pp. 440-443. Osborne, A., 1964. How to calculate three-phase flash equilibrium. Chem. Eng., 21: 97-100. Peng, D. and Robinson, D.R., 1976. Two- and three-phase equilibrium calculations for systems containing water. Can. J. Chem. Eng., 5412): 595-599. Perry, R.H. and Chilton, C.H. (Editors), 1973.Chemical Engineers’ Handbook. McGraw-Hill, New York, N.Y., 5th ed., pp. 2-6. Sivalls, C.R., 1977. Fundamentals of Oil and Gas Separation. Proc. Gas Conditioning Conf., Univ. Oklahoma, 17 pp. Smith, H.V., 1962.Oil and gas separation. In: T.C. Frick (Editor), Petroleum Production Handbook, Vol. I, Mathematics and Production Equipment, pp. 1-40. Souders, M. and Brown, G.G., 1934.Design of fractionating columns, I: Entrainment and capacity. Ind Eng. Chem., 26: 98-103. Uren, L.C., 1953. Petroleum Production Engineering: Oil Field Exploitation. McGraw-Hill, New York, N.Y., pp. 558-565. Whiney, K.F.and Campbell, J.M., 1958.A method of determining optimum second stage pressure in three stage separation. J. Pet. Tech., lO(4): 53-54. Wilkins, R.B., 1949. Stage separation of crude oil. Oil Gas J., 48(26): 62. Worley, M.S. and Laurence, L.L., 1957. Oil and gas separation is a science. J. Per. Tech., 9(4): 11-16.
109
Chapter 4
OIL FIELD EMULSIONS AND THEIR ELECTRICAL RESOLUTION L.C. WATERMAN, R.L. PETTEFER and G.V. CHILINGARIAN INTRODUCTION
In 1850 there was a thriving industry in West Virginia and Pennsylvania based on the evaporation of natural brines for the production of salt. An occasional troublesome contaminant of the brine was crude oil, which would seep from the earth with the brine or would accompany brine produced from wells. T h s oil was skimmed off the surface of the brine pools and discarded. The more enterprising of the salt producers used gas that was produced from brine wells for firing the evaporating pots. Small bottles of the oil were sold at “medicine shows”. The garish labels bore the picture of a fierce Indian and attested to the universal curative qualities of the “rock oil from the bowels of the earth”. Eventually a bottle of the oil was sent to Yale University for analysis by distillation. Based on this report, a syndicate promoted an oil well. Drilling of the Drake Well was accomplished by a group of salt-water well drillers from West Virginia, U.S.A. Thus, from the beginning of the petroleum industry, salt water and oil have been associated, though the roles of contaminant and product have been reversed. Brines reside in crude oil principally because salt water generally underlies the crude oil in the geological formation from which it is produced. With careful completion and production methods, it is possible to produce wells initially with no salt water. As the producing life of a field is extended, however, increasing proportions of salt water are produced with the oil. Salt water encroachment normally starts at the edge of the fields and progresses until the production is predominantly water rather than oil. The oil field waters whch are produced with crude oils vary widely in composition and the amount of salts, which they carry in solution, but their salinity is generally greater than that of sea water. Overall, the total productions of oil and water from oil wells are approximately equal. Emulsification of the water and oil, by intimate mixing, may occur in the formations themselves, or in mechanical equipment such as pumps, chokes, gas separators, and piping. These emulsions may comprise varying proportions of oil and water, from 0 to loo%, but pipelines will not purchase the oil until its water content is reduced to the 0.5-2% range, varying with the specifications prevalent for the geographical area or dictated by the purchaser. These specifications for the maximum water contents are dictated not by the amounts which the pipeline operator would prefer to have, i.e., no water at all, but rather by the difficulties of
110
reducing commercially the water contents in specific crude oils to lower values by even the best dehydration methods. The emulsified water exists predominantly in the form of dispersed particles (water-in-oil type of emulsion), which will vary in size from large drops down to small ones of about 1 pm (0.00004 in.) in diameter. Character of the water and of the oil (gravity, surface tension, chemical constituents, etc.) and production methods will determine the size distribution and stability of emulsion particles. Chlorides, sulfates, and carbonates of sodium, magnesium, and calcium are generally present in the water in decreasing order. The mineralogy and petrology of geologic formations (e.g., clayey sands, limestones, and dolomites) from which the oil is produced and diagenetic and catagenetic processes influence the composition and concentration of brines. Concentration of salts in the brine varies widely from field to field. As the ratio of water to oil increases to predominantly water, there is a tendency for inverse emulsions (oil-in-water) to be formed, particularly if the pumping equipment is worn and wire-drawing of the mixture occurs. The O/W emulsions have the viscosity of water and may be milky or coffee-colored. Indeed, milk is an oil-in-water emulsion. Particle size is of the order mentioned above. Treatment of inverse emulsions is done by chemical methods, but prevention of emulsification by good mechanical maintenance is often helpful. These emulsions may be stabilized by organic acid salts of monovalent metals.
THEORY OF EMULSIONS
A discussion of the theory of emulsion formation involves a study at molecular level of the individual groups and their environment. The formation of a drop and its interface between oil and water requires energy. It is a rule of nature that all energetic systems tend to seek the lowest level of free energy. Thus, an elevated object tends to fall, a pressure tends to relieve itself, and a hot object cools to the temperature of its environment. The energy of drop formation causes the drops to be spherical, because this shape represents the least area and free energy for a given volume. Drop formation by agitation is beautifully demonstrated by Dr. Edgerton’s famous flash photograph “Coronet” of a drop of milk falling into a saucer of milk (Fig. 4-1). The energy also tends to cause the drops to coalesce and to settle to bulk water. Impurities in the system, however, will interfere with coalescence and reduce the free energy. In the case of a molecule internal to a liquid system, it is attracted equally in all directions toward all of its neighboring molecules, which is a characteristic of liquids (see Fig. 4-2). If the liquid is divided in a horizontal plane and separated into two parts, the molecules at the new surfaces are attracted by equal forces in the horizontal plane and to the adjoining molecules beneath in the lower portion and, similarly, upward for the upper section. The separation of the molecules at the plane of cleavage was accomplished by a force necessary to overcome the intermolecular
111
Fig. 4-1. “Coronet”, formation of drops.
attraction in the vertical direction. The product of the vertical force times the short distance through which the forces are significant is the work done in forming the surfaces. This is the “free energy” of the system. It is proportional to the area of the surfaces formed. Numerically, the “free energy” per unit area is equal to the “surface tension”, which is, at least, a convenient mathematical concept. The work done in increasing a surface area is that which is necessary to move a molecule to
+.t
c
Molecule internal to the liquid
-0-I f
Molecule at the interlace
Fig. 4-2. Forces of attraction between molecules.
112
the unbalanced forces at the surface from the interior of the liquid where the forces are balanced. The foregoing is true whether the material at the new surface is in contact with a gas, such as air, or another immiscible liquid, as is the case with oil and water. In the latter case, “interfacial tension” is the expression used. The surface tension of water at 20°C is 73 dynes/cm, whereas that of most organic liquids is about 28 dynes/cm. Interfacial tension between crude oil and water varies from about 20 dynes/cm to 32 dynes/cm. When oil and water are vigorously mixed, both types of emulsion are formed, but primarily the minor phase tends to become dispersed. Both types of emulsion tend to resolve and the surviving type (O/W or W/O) depends largely on the nature of the stabilizer and the phase ratio. Further reference to “emufsions” in this chapter implies water-in-oil type emulsions, which is the predominant type in crude oil production. The formation of an emulsion involves the creation of enormous areas of interface with attendant free energy that is supplied by agitation in pumps, friction in lines, or pressure drop through valves. For example, in half a gallon of oil, a 1% emulsion consists of about one cubic inch of water. If the water is divided into drops 0.0001 in. in diameter, there will be about 2 trillion of them. The total area of interface formed would be 400 ft2, almost equal to the surface of two 9 X 12 ft rugs on both sides. This area can gather a considerable amount of stabilizer or dust. The free energy tends to subside and will do so in pure systems of distilled water and a pure hydrocarbon, such as hexane: the drops coalesce to form free water. In impure systems, however, another form of energy degradation may precede and prevent coalescence. Polar molecules (i.e., molecules which have external electrical fields) or groups of molecules in the oil, that are least similar to the most prevalent oil molecular species, will be subject to somewhat lower intermolecular forces. Being less attracted to the internal body of the oil and by virtue of its polar nature, some material will be adsorbed to the oil-water interface, water itself being highly polar. Such materials include asphalt, asphaltenes, resins, oxygenated sulfur and nitrogen compounds, porphyrins, waxes, metallo-organic salts, organic acids, and sediments. Of these, asphaltenes are most prominent. If present, impurities from the water side may be adsorbed to the interface. Interfacial adsorption of these surface-active materials results in a reduction of free energy. Such surface-active materials are called stabilizers or emulsifying agents. Stabilizers constitute the third essential component of stable emulsions of oil and water. They give rise to a physical barrier that prevents water drops from getting close enough for the intermolecular forces (water-to-water attraction) to be of sufficient strength to bring about coalescence. Stabilization of the interface begins at the instant a drop is formed. The process is called “aging”. It may proceed rapidly or extend over a period of days. Emulsion that has been held in storage may have been “aged”, thereby increasing the dehydration problem. When concentrated, the extraneous molecules (similar in kind) have a mutual attraction, which results in an elastic, sometimes tough and viscous, film around the drop. When molecules are present that are attracted and enveloped by the water on one end and the oil on the other, such as is the case with the soaps (e.g., calcium naphthenate), the interface becomes
113
indistinct, the free surface energy may become very low, and the emulsion will be very resistant to treatment.
DEHYDRATION
Dehydration is concerned with reduction, removal, rupture, or counteraction of the stabilizing films, coalescence of the droplets, and gravitational separation of the oil and water phases in a relatively brief residence time, e.g., 20 min. According to Stokes’ law, velocity of settling of a water drop is proportional to the cross-sectional area, difference between the gravity of oil and that of water, and the viscosity of the oil. Thus, the most favorable combination for separation exists when the oil has high API gravity and low viscosity, and water consists of large, unstabilized drops of salty water. By good production practice, the encroachment of water may be delayed, whereas the degree of emulsification may be mitigated by good equipment maintenance. Consequently, to some extent, produced water may be unemulsified (or “free water”) and will separate rapidly unassisted. Nonetheless, a large proportion of oil that is produced must be treated. The means of treatment are: (1) heat, (2) chemical destabilization, (3) electrical coalescence, and (4) gravitational settling. Heat
Heat increases the solvency of the bulk oil for the stabilizer and the rate of diffusion of the stabilizer into the bulk oil; it decreases the viscosity, thickness, and cohesion of the film. Heat also reduces viscosity of the oil. Heating the oil to promote dehydration has been used from the earliest days of oil industry. Like many tools that have served well in terms of effectiveness, however, its use has decreased in favor of more efficient means or their combinations. Evaporation losses from heat result not only in a loss of oil volume, but a reduction in price because of a decrease in the API gravity. Decrease in the API gravity of one degree corresponds to a volume loss of about 2.5%. In some instances, even a slight loss of API gravity may lower the price into the next lower bracket. Furthermore, fuel gas, that was formerly wasted, is now a valuable product. Thus, the penalty for inefficient heating is multiplied and treatment by heat is fast becoming obsolete or an accessory to other methods. The most effective and efficient source of heat is that of the producing formation. By treating oil on the flowline as close to the well as other considerations permit, loss of connate heat is avoided. Furthermore, “aging” of the emulsion and emulsification by surface equipment is held to a minimum. Such “flowline treatment”, therefore, is highly desirable. Frequently no additional heat is required for such treatment and maximum efficiency is gained at minimum cost.
114
Fig. 4-3. Rupture of emulsion film by chemical destabilizer.
Chemical additives
The emulsion can be further modified by addition of chemical destabilizers. These surface-active agents adsorb to the water-oil interface, rupturing the skin and/or displacing the stabilizer back into the oil. Figure 4-3 shows the effect of additions of destabilizer to an oil-water interfacial film. The motivation derives from a still further lowering of the interfacial tension of the drop. The added molecules, however, are of such dimension, in such limited concentration, and of such arrangement at the interface that the film quality and thickness are drastically changed. Time and turbulence aid diffusion of the treating chemical through the oil to the interface; therefore, it is usually added to the oil at the wellhead. The synthetic interfacial material, having caused the natural skin to recede from part or all of the interface, comprises a thin film susceptible to rupture by the attractive intermolecular forces of water-to-water at very close distances. Inasmuch as these forces vary inversely as the seventh power of the distance between molecules, the remaining problem is bringing the drops into very close contact quickly and without such severe turbulence that could cause the drops to redisperse. William S . Barnickel discovered and pioneered the use of treating chemicals about sixty-eight years ago. Thousands of his “Tret-0-Lite”* demulsifier compounds have been used wherever oil is produced. Selection of a compound is ordinarily accomplished by actual test on a sample of emulsion. The mechanism of the chemical process is not explicable by any simple theory.
* Registered Trademark, Petrolite Corporation.
115
0
( C )
Fig. 4-4. Effect of electric field on water drops in oil.
Electrical treatment
The application of electric field is a powerful tool for causing dispersed drops to rapidly collide with one another and for overcoming the resistance of stabilizing films. The collision and coalescence of drops is accomplished by an induced dipole attraction between them. After coalescence, separation of the phases is due to gravity. The interaction of a field and induced dipoles is illustrated in the drawings and photomicrographs of Fig. 4-4. Figure 4-4.a shows a suspended uncharged particle, whereas Fig. 4-4.b exhibits the displacement of charges on a single drop (induced dipole) by an applied field. Full-line drawing in Fig. 4-4.c represents the effect of an applied field upon adjacent particles separated by a distance equal to a few radii. The broken-line drawing and photo are explained later. In Fig. 4-4.b, a single water drop is shown in an alternating electric field. There is a displacement of electric charges induced on it. The field and the induced charges shown in this figure reverse 120 times per second, but the relationship holds at any
116
instant. The electrodes establish the electric field to whch the drop is responsive. The left end is electrically attracted to the left by the field and the right end is pulled to the right. The drop is an induced dipole. Inasmuch as the forces are equal and opposite in direction, the drop remains in place and merely elongates. If the field is disconnected, the drop resumes its spherical shape and there is no residual change as a result of the electric field application. While the drop is distorted and under the influence of the field, however, the highly-polar stabilizing film will be responsive, and coalescence can more readily occur when the drops come in contact. The attractive coalescing force, F, between the aligned drops of equal size is equal to: 6KE2a6 F= d4
(4-1)
where K is the dielectric constant of the oil, E is the potential gradient, a is the drop radius, and d is the distance between centers. Accordingly, the force increases very rapidly as drop size increases and as the distance between particles decreases. Action is almost instantaneous. There is a limitation of the size to which the drops may coalesce for a given field strength and a maximum voltage that may be applied to a given system. This is due to a tendency for electrically charged drops to disperse. The condition necessary for stability in undisturbed drops is given by the following relationship:
where E, is the critical dispersing gradient at the surface of the drop, C is a proportionality constant, T is the interfacial tension, and a is the drop radius. For any effective coalescing gradient in a given system, there is a corresponding maximum size beyond which drops discharge small droplets. Broken-line drawing and photo in Fig. 4-4.c show a large drop dispersing toward a smaller stable drop. Conductance between drops reduces attraction. Optimum gradients and electrode configuration have been determined by 70 years of worldwide practical application, since Dr. F.G. Cottrell invented electrical dehydration. The principles have been extended to other applications of which electrical desalting of crude oil is the most important. Desalting is the removal of contaminants (salts and sediments) that reside in crude oil after normal dehydration has reduced the BSW to pipeline specification. In the process, salt particles in the residual BSW are dissolved and dispersed by added fresh water (about 5%), after which brine droplets are electrically coalesced and separated by gravity. The final contaminant content of the crude oil may be a few parts per million. Though desalting equipment is a normal part of all refineries, salt specifications are being imposed on oil moved by pipelines and tankers in many
Fig. 4-5. Pictorial assembly of Petreco” electric dehydrator. ‘I
118
parts of the world. “Petreco”* desalters, first used about 1935, have been applied to streams of a few hundred barrels per day to up to 300,000 bpd (bbl/D) in a single unit, meeting specifications as low as about one part per million chloride content. ELECTRIC DEHYDRATORS
Through the years, design of dehydrators has progressed steadily from the early models treating a few hundred barrels per day and delivering oil with a few per cent of BSW up to the current dehydrator models delivering from 1000 to 182,000 bbl/D containing less than 0.5% BSW. Design features of dehydrators of all sizes are essentially similar. These features are shown in Fig. 4-5. Size ranges from 6 ft in diameter by 12 ft long to 14 ft in diameter by 140 ft long. Normally, operating voltages are 480 V at switchboard and
Fig. 4-6. Equipment (10x21 ft Petreco dehydrator: first floor) for handling production at Lake Maracaibo. Venezuela.
* Registered Trademark, Petrolite Corporation.
119
16,500 V at electrode. Optionally, both electrodes may be electrified with opposing polarity, thereby impressing the sum of the two voltages upon the field. All electrical equipment is enclosed and protected against overload, and meets all safety requirements. Electric power consumption varies according to the conductivity of the crude oil. Ordinarily, the least conductive crude oils have high API gravity for whxh the load is about 0.5 kW for each 1000 bbl/D capacity. For heavy crude oils the load may be 2 kW per 1000 bbl/D capacity. An important feature of the design is to combine the dual function of electrical coalescence with optimum settling in a single vessel. The internal piping and electrode arrangement is directed to this end. Residence time of the crude oil in the dehydrator is about 20 min. Cost of dehydrators varies inversely with the size and the API gravity of the crude oil. Large units treating high-API-gravity crude oils are most economical in cost. Electrical dehydrators are adaptable to special situations. Figure 4-6 shows an installation in Lake Maracaibo, Venezuela. The operation of a large number of wells is controlled at these “stations”. The electrical dehydrators operate at a well temperature ( = 130°F), to conform with the restriction that no fire is permitted. “Tret-0-Lite” demulsifier, in the amount of about 6 ppm is added to the oil prior to electrical treatment. The 10 x 12 ft “Petreco” dehydrator reduces the water content of 12,000 bbl of oil per day from 50%to 0.3%.
Fig. 4-7. Typical “Petreco” electric dehydrator.
120 GAS OUT
GAS [RELIEF
INLET
FIRE TUBE
WATER OUT
DRAIN
WATER OUT
Fig. 4-8. Diagram of “chemelectric” treater.
Fig. 4-9.Facilities for Gulf Coast, U.S.A., offshore production.
AA DRAIN
121
Figure 4-7 shows a typical 10 X 41 ft “Petreco” dehydrator installed in the Linda Field, Indonesia. This dehydrator is treating 35,000 bbl of 17” API crude oil per day. AUTOMATED DEHYDRATION
The Petreco “chemelectric”* dehydrator is a combination unit for oil-gas separation, heating, chemical destabilization, and electrical dehydration of oil as it is produced from the wells. The “chemelectric” dehydrator is ideally adapted for lease automation and flowline installation. Thus, the oil is treated with a minimum of handling and “aging”, which is an ideal arrangement. Aside from the convenience, a substantial saving in investment, piping, real estate, pumping, heat, chemicals, vapor losses, and labor has resulted from coordination of the several treating elements to produce the best end result for the least cost. The combination of treating elements makes it possible to proportion them for
Fig. 4-10. “ Chemelectric” treater for Colombia, South America, production.
* Registered Trademark, Petrolite Corporation.
122
Fig. 4-11. LACT operation treaters in Oklahoma, U.S.A.
best operation, such as increasing content of chemicals and reducing heat, or to increase the throughput capacity by adjustment of both. This unit is especially useful in dehydrating the very stable emulsions produced by secondary recovery operations, such as flooding by water, steam, or fire. Figure 4-8shows diagrammatically the operation of a “chemelectric” unit. Figure 4-9 is an aerial view of an automated installation for handling offshore production, in the Gulf Coast. Oil is received from offshore in the high-pressure gas separators at lower left. Then it flows through the two “chemelectric” treaters at right center and, finally, through meters on platform at center to transfer to storage tank at upper left. In Fig. 4-10,a 10 x 25 ft “chemelectric” treater is being installed in Colombia, South America. The unit will degass and dehydrate 10,000 bbl/D of 20” API crude oil containing up to 20%water and 3 MMscf of gas per day. Figure 4-11 shows 6 x 15 ft and 10 x 20 f t “chemelectric” treaters installed on a LACT lease at County Line, Oklahoma. The highly simplified enclosed electrical gear of the “chemelectric” treater is presented in Fig. 4-12. The conduit between the 16,500V transformer and the vessel contains the high-voltage cable.
123
Fig. 4-12. Electrical equipment on “chemelectric” treater.
It must be remembered that the gravity, viscosity, and asphaltic content of the crude oil together with the gas pressure and the methods of production will influence both the equipment cost and the cost of operation. The latter may vary from a few tenths of a cent per barrel to a few cents per barrel. Whatever the situation, the combination of heat to modify the oil, chemicals to modify the emulsion, and electricity to consummate the operation will provide versatility, reliability, and economy.
SAMPLE QUESTIONS
(1) Draw a schematic vertical cross section through an electrical dehydrator. (2) Describe the two steps involved in dehydration by electrical method. (3) List different types of electric dehydrators. (4) Illustrate method of arranging electric dehydrator equipment.
124 ACKNOWLEDGEMENTS
The help extended by D.L. Kraft, G.C.Hardwicke, and F.D. Watson is greatly appreciated by the authors. REFERENCES Blair, C.M., 1960.Interfacial films affecting the stability of petroleum emulsions. Chem. Ind., May 14: 538-544. Greenlee, R.W., 1960. Factors in the Stability of Petroleum Emulsions. Prepr. Pap. Div. Pet. Chem., Am. Chem. Soc., Sept. 11-16. Monson, L.T. and Stenzel, R.W., 1946.The technology of resolving emulsions. In: J. Alexander (Editor), Colloid Chemistty, Vol. VI. Rheinhold, New York, N.Y., pp. 535-552. 1932.A new theory of emulsions. J. Phys. Chem., 36: 3087-3107. Roberts, C.H.M., Shea, G.B., 1937. Practices and Methoah of Preventing and Treating Crude Oil Emulsions. Bur. Mines Bull., 417: 106 pp. Swigart, T.E., 1961. Histoty of Petroleum Engineering. Am. Pet. Inst., Div. Prod., Dallas, Tex., pp. 925-931. Waterman, L.C., 1965. Electrical coalescers, theory and practice. Chem. Eng. Progr., 61(10): 51-57.
125
Chapter 5
CHEMICAL RESOLUTION OF PETROLEUM EMULSIONS DONALD U. BESSLER and GEORGE V. CHILINGARIAN
INTRODUCTION
More than 80% of the crude produced in the world comes to the surface with various amounts of free and emulsified water. As oil fields grow older and are subjected to secondary recovery waterflooding, the amount of water increases. Older oil fields that have been waterflooded for many years may have 95-98% water produced along with the crude oil. Petroleum emulsions of the oil-in-water and water-in-oil type plus any free water present cause serious problems to the producer, transporter, and refiner of petroleum. Reduced transportation costs, increased throughput of pipelines, and reduction in corrosion, scale formation, and bacterial growth can be achieved by removing the water in the field. With the increasing government regulations on effluent water and the cost of crude oil, the use of oil-in-water demulsifiers and water clarification equipment has become an important aspect of the processing of crude oil in the field. Chemical resolution of petroleum emulsions of both types is an established routine procedure in the production and handling of crude oil throughout the world. It is effective, easily practiced, and inexpensive, when used with settling tanks, heater treaters, and chemelectric treaters.
NATURE OF EMULSIONS
Definition of emulsions
An emulsion can be defined as a mixture of two mutually immiscible liquids, one of which is dispersed as droplets in the other and is stabilized by an emulsifying agent. The dispersed droplets are known as the internal phase. The liquid surrounding the dispersed droplets is the external or continuous phase. The emulsifying agent separates the dispersed droplets from the continuous phase. In the oil field, oil and water are encountered as the two phases. They generally form a water-in-oil (W/O) emulsion, although as the water cut increases and secondary recovery methods are employed, the “reverse” or oil-in-water (O/W) emulsions are increasing.
126
Basically, there are three components in a water-in-oil emulsion: (a) Water-the dispersed or internal phase. (b) Oil- the continuous or external phase. (c) Emulsifying agent-stabilizes the dispersion. Besides these three components, certain conditions must also be met before an emulsion could form. Two conditions necessary to form a stable emulsion are: (a) The two liquids must be immiscible. (b) There must be sufficient agitation to disperse the water as droplets in the oil. Pure water and oil which contains no emulsifying agents will never form an emulsion no matter how much agitation is applied. Inasmuch as these two liquids “dislike” each other intensely, if confined in the same container, they will quickly find a state of existence which gives the least contact or the smallest surface area. A drop of water in a body of oil will take the shape which gives the least surface area, i.e., that of a sphere. Role of the emulsifier
Emulsifiers are surface-active materials found in crude oil or added as in the case of sulfonate floods. The inclusion of solids tends to stabilize these emulsions to an even greater degree. These include asphaltenes (a general term applied to a large variety of chemical compositions containing sulfur, nitrogen, and oxygen), resins, creosols, phenols, organic acids, metallic salts, silts, clays, and many others. The emulsifier either tends towards insolubility in either liquid phase or has an attraction for both phases, but always concentrates at the interface. The molecules of emulsifier are mutually repulsed. The three principal actions of the emulsifier are: (1) reducing surface tension, (2) forming a physical barrier, and (3) suspending water droplets. Stability of emulsions
The stability of emulsions is dependent on the various factors described below: (1) Drop size. The size of the dispersed water drops is a measure of stability. The type and severity of agitation generally deter*nes the drop size. The more shearing action that is applied to the oil-water mixture, the more the water will be divided into smaller drops and the more stable the emulsion becomes. Stable emulsions have been found to contain all sizes of droplets, but the percentage of larger droplets is very small. (See Figs. 5-1 and 5-2.) ( 2 ) Type of emulsifier. The type of emulsifying agent will drastically affect the stability of an emulsion. There is a considerable difference in the power of different agents. Their activity is generally related to two general functions- speed of migration to the interface, and performance at that site. When the water and oil first mix, the emulsifying agent may be evenly distributed throughout the oil. At this time, the emulsion may be relatively unstable. With time, the agent migrates to the oil-water interface due to its surface-active characteristics. This migration, with
127
time, produces a thicker and tougher film surrounding the droplet, resulting in an emulsion that is more difficult to break than the fresh one. ( 3 ) Water content. The amount of water present in the regular emulsion and available at the time agitation takes place is directly related to the emulsion’s stability. Generally, as the water content increases, the stability of the emulsion decreases. Water content above 50% tends to encourage the formation of water as an external phase; thus, on dilution with water, the emulsion may invert and the water then becomes the continuous phase. (4) Viscosity. The viscosity of the external or oil phase plays a dual role. In an oil having a high viscosity (high resistance to flow), a given amount of agitation will not break up the water phase into droplets as numerous or fine as would be the case with a lower-viscosity oil. On the other hand, the high-viscosity oil is able to maintain larger drops of dispersed water in suspension and smaller dispersed droplets will have an even greater resistance to settling. Thicker crudes also retard the movement of emulsifier particles to the interface. In general, it can be said that higher-viscosity (lower-API gravity) crudes form less stable emulsions in terms of many small water drops, but this is more than offset by the difficulty of resolving what is formed and promoting water separation. (5) Solids. Emulsions can be stabilized by the presence of solids. Many of the same assumptions discussed for the emulsifying agent will apply. In the case of solids, the size of the water droplet will depend on the emulsifying solid. ( 6 ) Efectrical charge. The existence of electrical charges on the water globules can cause the globules to be mutually repelling. At the time the globules are dispersed in the continuous phase, they acquire an electrical charge which is located at the interfacial boundary of the oil and water. The droplets of the dispersed phase are surrounded by a film which possesses a charge, usually negative, the effect of which extends outward into both phases. The repelling action of like charges prevents the droplets from approaching each other close enough to allow coalescence. (7) Surface tension. Surface tension of an emulsion system refers to the surface tension of the continuous phase-dispersed phase boundary surface and is called interfacial tension. The absorption of substances (emulsifying agents) at this boundary surface is important in changing the surface tension between the two phases. Emulsification is brought about by the emulsifying agents, which lower the interfacial tension between the two phases, so that they will remain dispersed. When the concentration of absorbed materials is sufficiently high, it may lead to the formation of a tough membrane, which increases the emulsion stability. At the interface and between the phases of an emulsion, therefore, two forces are at work: (a) Interfacial tension, which tends to cause coalescence or dispersion; and (b) coherence of the film of emulsion of emulsifying agent around the dispersed phase, which tends to resist coalescence. If the coherence of an agent is so great that the surface tension cannot overcome it, the emulsion is stable. (8) Film strength. The presence of some foreign materials in an emulsion
128
Fig. 5-1. Tight emulsion with small dispersed droplets which have considerable resistance to settling (a tight emulsion is usually a stable emulsion). Fig. 5-2. Loose emulsion with larger dispersed droplets (a loose emulsion is less stable than tight emulsion because the large dispersed droplets tend to settle easily).
increases the strength of the film surrounding a drop of water. The smaller droplets are harder to rupture than the larger ones. The film strength of a drop of water, therefore, varies not only with drop size, but with impurities. To break the film, it is necessary to introduce chemical action and/or to apply heat to rupture this film. (9) Density. Another factor that affects emulsion stability and settling time is the relative density of the oil and water. As this difference becomes greater, the action between droplets is increased and more rapid settling is promoted. (10) Aging. As discussed earlier, the naturally occurring emulsifiers in crude oil are initially dispersed throughout the oil, and after mixing with water, tend to go to the interface. It is this action that causes crudes to “age” and become more difficult to treat with time. Inasmuch as the lower-API gravity crudes are thicker, this action is slower as is the aging process. Normally, the higher the API gravity, the more rapid the aging process. CRUDE OIL PRODUCTION
Crude oils differ in characteristics according to their geological age, chemical constitution, and associated impurities. Consequently, crude oil emulsions are stabilized by a variety of materials, depending upon the origin of the crude. It is not necessary to determine the exact character of the emulsifying agent to resolve the emulsion by the use of chemical demulsifiers. The waters associated with crude oils likewise vary widely in characteristics. Some have densities greater than 1.20, whereas others are essentially non-saline.
129
Fig. 5-3. Treating plant with gas separator, gun barrel, and stock tanks.
Ions present usually include Na+, Ca2+, Mg”, C1-, HCO; , SO:-, and sometimes Ba2+. Oil-bearing formations commonly contain in addition to petroleum, varying proportions of water and natural gas. Inasmuch as crude oil usually contains sufficient emulsifying agents to stabilize an emulsion, it is only necessary that the well produces some water as well as oil and that sufficient agitation be available to accomplish the required mixing. The agitation arising from the turbulent flow of the oil and the water through the well casing, tubing, and surface equipment is usually sufficient to give rise to emulsification. Gas v e n t
Fro w line LHeofer fue/line
Chemicol feeder
Gas
Heater
Conductor Settling
Storoge tank
+ank
Fig. 5-4. Diagram of typical flowline treatment system. Many variations of this basic design are possible; all elements shown here may not be required in all installations. Heavy line shows flow of chemicalized liquids. (Courtesy of Dr. Louis T. Monson.)
130
The pipeline and other transportation companies, as common carriers, have long-established specifications for crude oil which preclude the introduction into their lines or systems of any lot of oil that contains more than a predetermined (small) maximum percentage of water and emulsion. The permissible maximum is determined in part by the ease or difficulty with which the emulsions can be commercially resolved, but in general it is less than 3% and in most cases it is less than 1%. High-API gravity crudes often have a 0.1 or 0.2% maximum allowable water content. Consequently, crude oil is required to be “treated”, usually at or near the point of production, to resolve any emulsion present and to remove the water which is separated in the process. (See Figs. 5-3 and 5-4.) A diagram of a typical flowline treatment system is shown in Fig. 5-4.
REDUCED-TEMPERATURE TREATING
In the past, the energy required to heat crude oil was supplied by products produced on the lease or of such a low cost that no significant importance was given to them by many companies. There was also an abundance of crude and the losses incurred were not considered significant. Reducing the heat required is not novel and this has been done for many years. The increased cost of fuel, crude oil prices, and economics of operation have renewed interest in treating the crude oil in the field at lower temperatures. The amount of heat required to treat a given crude will depend on gravity, viscosity, pour point, equipment, time, and chemical used. Heating of crude oil influences the treating process in several ways: (1) It makes the oil thinner and, therefore, water droplets are able to fall faster. (2) It increases the difference in the specific weight between water and oil, causing the water droplets to fall faster. (3) It melts and solubilizes solids such as paraffin, which may be acting to stabilize the emulsion, thereby removing them from the interface. (4) It improves the mobility of demulsifier and water droplets, thus improving the rate of emulsion breaking. From this it can be seen that if the treating temperature is reduced, it may be necessary to increase the time, agitation, and/or chemical concentration. The use of heat in treating contributes several advantages but at a loss of profits due to unnecessary costs. Those costs are reflected in: (1) the amount of fuel used for heating, (2) lower volume of produced oil, (3) lower gravity of produced oil, and (4) equipment failure. In the past, fuel for heating was provided by the produced gas, with little or no value being placed on it. Today, with a market for everything produced, the heater fuel has a definite value. For example, with certain assumptions listed below, this value can be calculated. Assumptions: Fuel gas value = $2.00/Mcf 150 Btu/bbl oil/OF
131
300 Btu/bbl water/"F 30% water-cut emulsion 50% heating efficiency 1100 Btu gas 60°F temperature rise Then the cost of fuel gas would be $6.08 per 100 bbl of treated oil. As oil is heated, vapor pressure increases and light ends evaporate, thus reducing volume and API gravity. The percentage loss is quite variable and subject to such influences as crude gravity and composition, temperature change, treater pressure, heat exchange efficiency, time, and other factors. Field evaluation of three leases producing broadly differing gravity crudes showed the following: With 60°F temperature rise and 25 psig treater pressure, volume of saleable oil was reduced by an average of 2.9%. At a value of $31.00/bbl, this results in a loss of $89.90/100 bbl. At $2.00/Mcf, the value of the increased gas produced at the higher temperature averaged $8.71/100 bbl. This yields a net loss of $81.19/100 bbl as a result of volume loss at the hgher temperature. Loss of light ends results in a decrease in gravity. Inasmuch as much crude is sold on a sliding scale with the posted price decreasing as the gravity decreases, the profit loss may be appreciable. A 2.9% volume loss on a 30" API crude oil exhibits a gravity loss of 1.3" API. At a $0.02 penalty per 1.0" API gravity drop, this amounts to a profit loss of $2.60/100 bbl. Increased heat results in increased scaling and more frequent burnout of fire tubes. While this constitutes a definite operating expense or potential profit loss, it will fluctuate so broadly that no attempt will be made here to affix a dollar value. Costs attributable to heat in the above examples total $89.87/100 bbl plus equipment maintenance costs. Experience shows that with most oilfield emulsions, some if not all of the added heat can be replaced by a chemical. The increase in the amount of chemical necessary to add will fluctuate considerably depending on the crude oil and the temperature range involved. Relating to the previously used example, a 60" heat reduction from 140°F to 80°F required three times as much chemical, which was an increase from two quarts per 100 bbl at 140°F to six quarts per 100 bbl at 80"F, or an increase of 1.0 gal. At a typical cost of $8.00/gal for chemical, the increased cost would be $8.00/100 bbl. Compared to the previously calculated heat cost of $89.87, the net saving by reducing heat would be $81.87/100 bbl. Reducing heat can be a good way to sell more oil and increase profits. As a practical approach to reducing heat, it is recommended that this be done 20°F at a time. This allows for adjustments as needed to prevent possible accumulations of "bad" oil. CHEMICAL RESOLUTION PROCESS
The problem of resolving water-in-oil emulsions has been approached in a number of ways over the years. Today, however, the chemical demulsification
132
process is by far the most widely used in the oil industry. It was pioneered by William S. Barnickel, a pharmaceutical chemist from St. Louis, Minnesota (U.S.A.), who discovered that mixing a minute proportion of a properly selected chemical composition with a petroleum emulsion, under suitable operating conditions, would cause the water to separate. The scientific basis for the resolution phenomenon is not yet well defined. This is not totally unexpected, if one recalls all the possible combinations arising from the variations in crude oil composition, aqueous phase composition, phase/volume ratio of the two liquids in any emulsion, and the occasional presence of such other materials as fine silt and various formation particles. It can easily be shown that tens of thousands of different emulsion systems could be produced from crude oils and oil field waters. Some chemical demulsifiers, such as the early soap reagents, are water-soluble or water-dispersible, whereas others are oil-soluble or oil-dispersible. Some chemicals are both water- and oil-dispersible, whereas others are apparently not soluble or dispersible in any appreciable concentration in either the aqueous or oily media. They may be anionic, cationic, or nonionic depending on the type of emulsion to be resolved. Inorganic reagents are rarely used today in W/O “regular” emulsions, but are still commonly used in O/W “reverse” emulsions. Before one can determine which class of reagent is most effective in a particular application, the formation from which the emulsion was produced, conditions of production, and many other facts must be known. As a consequence, reagent selection is ordinarily accomplished by actual demulsification tests on a representative sample of the emulsion. Because of the variability of petroleum emulsions, the manufacturers supplying reagents for their resolution commonly make many different commercial formulations, a number of which may be in use in a single oil field at any one time. Obsolescence rates are high and a reagent of one type may be succeeded by another having entirely different composition and properties, at a particular location. All these facts have not simplified the problem of evolving a theory which will satisfactorily explain observed field results. The highly specific nature of the chemical reagents used today to resolve petroleum emulsions suggests that the mechanism of the resolution process is quite complex, and that it cannot be explained by any simple theory. Action of demulsifiers
Strangely enough, demulsifiers are very similar in nature to emulsifiers. The action is all at the oil-water interface and, therefore, the faster the demulsifier gets there, the better job it can do. After it reaches the interface, it works by flocculation, coalescence, and solids wetting, as described below. For oil-in-water emulsions, the actions of oil-in-water demulsifiers are very similar. (1) Flocculation. The first action of the demulsifier on the emulsion involves a joining together or flocculation of the small water drops. When magnified, the
133
flocks take on the appearance of bunches of fish eggs. If the emulsifier film surrounding the water drop is very weak, it will break under this flocculation force and coalescence will take place without further chemical action. In most cases, however, the film remains intact and, therefore, additional treatment is required. Good flocculation is characterized by bright oil. (2) Coalescence. The rupturing of the emulsifier film and the uniting of water droplets is defined as coalescence. Once this process of coalescence begins, the water droplets grow large enough to settle out. Good coalescence is characterized by a good water drop. (3) Solids wetting. In most crude oils, solids such as iron sulfide, silt, clay, drilling mud solids, paraffin, etc., complicate the demulsification process. Often such solids are the primary stabilizing material and their removal is all that is necessary to acheve satisfactory treatment. For removal from the interface, these solids can be dispersed in the oil or water-wetted and removed with the water.
OPERATING PROCEDURES
Chemical injection
The purpose of the chemical injection is to introduce a chemical into the wellstream so that it will neutralize the foreign materials acting as emulsifying agents and permit the separation of water. In order to get the most out of chemical injections, the actual location of the injection point must be properly chosen. It should be located at a point far enough upstream from the treating equipment to allow adequate mixing. A good point is at the wellhead (Fig. 5-5) or the location of some other restriction such as a choke or header. There should be sufficient turbulence, mixing, and time for the chemical to exert its influence on the emulsion before entering any other equipment. To effectively use the large number of different demulsifier formulations requires a proficient method for selecting the proper compound for a given emulsion and system. Agitation
Sufficient agitation must be applied to the crude oil after the introduction of chemical. Additional hard agitation may or may not be beneficial. Increase in the amount of gentle agitation, such as in flow lines and settling tanks, is beneficial in promoting coalescence. Reemulsification may occur if severe agitation is given to an emulsion once it has broken into water and oil. This reemulsification may occur in gas separators, pumps, or any other location in the system that produces severe agitation once the emulsion has broken.
134
Heating
Many plants use heat in the treating process because it provides an aid to mixing, coalescing, and settling. Heat aids the treatment in the following three ways: (1) reduces the viscosity of the oil; (2) weakens or ruptures the film between the oil and water drops by expanding the water; and (3) alters the difference in gravity of the fluids and thereby tends to reduce the settling time. In effect, heat accelerates the treating process and is used primarily to reduce the size of the treating vessel. It must be remembered, however, that heat vaporizes the light ends of the oil. Unless some means is taken to conserve these, a reduction in API gravity and volume will result and it may be more beneficial to treat the crude oil at lower temperatures. Settling
Settling is the basic component in all treating procedures. All operations involving the use of heat, chemicals, or mechanical devices are designed to prepare the oil-water mixture for the settling step in the chemelectric heater treater or settling tank. The treating vessel usually provides sufficient time for quiet settling to allow all the water to settle. Time necessary to allow the water to settle is determined by the difference in specific gravity between the water and oil, viscosity of the oil, and by the size and condition of the water drops. While gentle movement will aid in coalescence, more severe turbulence in the settling section will increase the settling time. Mechanical systems Free water knockout Many of the free water knockout systems used in the field have been designed for specific applications. In general, free water knockouts have been used to separate the oil and water produced from a single well or several wells. Larger units can also be found where all the wells go into a free wqter knockout after the manifold. In either case, their function is to remove the excessive volumes of free water ahead of the treating plant. Free water knockout systems are generally used in connection with production having a high water/oil ratio. Separation of gas may also occur in the upper section of the knockout system. Free water is defined as water produced with the oil that will settle out within five minutes while the well fluids are stationary in a settling space within a vessel. Specific application of a free water knockout system and requirements it is expected to fulfill may vary and will have to be determined for each location. In some areas, the amount of BSW leaving the knockout system is unimportant as long as the treating plant is operating effectively. Other fields may have requirements of 20%or even as low as pipeline oil. Chemical injection may be applied before, after,
135
or both ahead and behind of the free water knockout system. Generally, it is ahead of the knockout system in order to assist in the removal of the water. Gas separator Horizontal and vertical gas separators provide tremendous agitation potential. The evolution of the gas in itself creates the turbulence and agitation which may be even greater, depending upon the separator design. Without the efficient removal of gas in these vessels, unwanted agitation will, at times, be created in downstream treating vessels. The use of chemicals, such as silicones, will greatly aid in the removal of gas. Settling tanks The rate of water drop is not too important because the chemical may continue acting over a relatively long time. The water-oil interface will be at the bottom of the settling tank and will not interfere with the saleability of the oil. (See Fig. 5-6.) Gun barrels Speed of water drop is generally not too important because gun barrels usually have a high volume to throughput ratio. The chemical may continue acting over a relatively long time. The interface need not be clean, but if an interface layer does develop, it must stabilize at some acceptable thickness. An interface layer in a gun barrel sometimes helps to treat in that it acts as a filter for solids and unresolved
Fig. 5-5. Flowing well with wellhead connections, which are referred to as a “Christmas tree”.
136
emulsion. Fresh oil containing demulsifier passing up through this interface layer helps treat it out and prevents an excessive buildup. Vertical heater treater Inasmuch as volume to throughput ratio in a vertical heater treater is generally lower than in a gun barrel, speed of chemical action becomes more important. With this higher throughput, it is harder to stabilize an interface layer, so more complete treatment is necessary in a shorter time. Solids control may be important in controlling the interface. Horizontal heater treater Horizontal heater treaters normally have a high throughput, so chemical action must be fast. The large interface area and shallow fluid depth require that the interface be maintained fairly clean. This type of treater can tolerate only very little interface accumulation. The higher the throughput, the less this can be tolerated and, therefore, the more complete the chemical treatment must be. Inasmuch as solids tend to collect at the interface, the chemical must also control any solids which might be present. Electrical dehydration Inasmuch as chemelectric treaters are horizontal vessels, the same general actions are required for them as for horizontal treaters. The chemical must break the emulsion rapidly and completely. The electric field promotes excellent water coalescence, so the chemical need not provide this. The electric field tends to throw down solids in the oil. These solids then accumulate at the interface, build up into the electrical field, and being conductive, short it out. Chemelectric treaters, therefore, require a chemical to effectively water wet any solids which are present.
RESOLUTION OF OIL-IN-WATER EMULSIONS
Oil-in-water emulsions, although not so commonly or widely encountered in oil-producing operations as the water-in-oil type, are receiving increasing attention, particularly because of the growing interest of governmental authorities in pollution prevention. Although such emulsions commonly contain less than 1% oil and frequently contain less than 1000 ppm of oil, they are unacceptable to disposal systems at such levels. Maximum permissible oil content is usually on the order of 25 ppm or even less. The oil producer must, therefore, be prepared to clean his waste water of oil before disposing of it. Thus, it behooves him to become familiar with the characteristics of oil-in-water emulsions. Oil-in-water emulsions are readily miscible with water. They are thereby distinguishable from the more common water-in-oil type crude oil emulsions, which are miscible with oil. The O/W emulsions are usually more fluid than W/O emulsions.
137
Chemical demulsifiers have been developed which are extremely effective in recovering the oil dispersed in such emulsions. This is accomplished without the production of undesirable residues or flocs, which would themselves present disposal problems. Oil-in-water demulsifiers, which ordinarily have different compositions than water-in-oil demulsifiers, are usually effective in proportions much smaller than those required to resolve conventional water-in-oil type petroleum emulsions. This reduces the cost of resolving oil-in-water emulsions and allows the producer to sell the recovered oil. For optimum results, the ratio of demulsifier to water-in-oil emulsion may be of the order of 1: 10,000-1 : 50,000, whereas the ratio of demulsifier to an oil-in-water emulsion is more likely to be of the order of 1:40,000-1 : 200,000. The procedure for demulsifying oil-in-water emulsions is substantially similar to that described for water-in-oil emulsions: (1) The demulsifier is added to the emulsion in the required proportion; (2) the chemicalized emulsion is agitated to promote coalescence of the oil particles; and (3) quiescent settling is thereafter provided in a suitable facility to achieve separation of the oil and water, which may be aided by air-gas flotation. Froth flotation procedures may be helpful in accelerating separation of the oil particles. It is important to note that, in contrast to the experience with water-in-oil emulsions, use of an excess of oil-in-water demulsifier is sometimes very disadvantageous. An appreciable excess of reagent may produce an oil-in-water emulsion which is at least as stable and undesirable as the emulsion originally subjected to treatment. Ordinarily, however, an acceptably wide range of reagent concentrations will be found to produce satisfactory resolution of the emulsion, before “overtreatment” becomes apparent. Heat is rarely useful in resolving oil-in-water emulsions chemically. TROUBLE SHOOTING
When chemical, agitation, heat, time, and, in some cases, electricity are used to produce good treatment, they are in balance with each other. If one of these variables is changed, therefore, another one must normally be changed in order to regain the most economical balance. For example, as treating temperatures are reduced, it is usually necessary to increase the amount of chemical or time. Often treating problems which develop are a result of some malfunction in the mechanical system. Chemical injection
If a system is not functioning properly, the first thing that must be done is to check to make sure that the correct amount of chemical is getting into the system. Even though a chemical injector is running properly, it may not be injecting properly. Filter screens can become plugged, pumps can become gas locked, balls and seats may be worn, and check valves may be leaking.
138
Check the location of the chemical pump. Chemical should be injected at a mixing point and relocation of the pump may solve the problem. Changing conditions of the wellstream may necessitate a change in the chemical used. Before bottle testing is done, however, the various problems possible in the system should be checked. Gas separators
Gas separators can be plagued by solids buildup, especially if the solids content is high. This may be particularly true if the demulsifier is injected prior to the separator, resulting in release of solids from the emulsion. A suitable demulsifier will break the foam and release the gas, but if excessive foam is the problem and the chemical does not readily break the emulsion and release the gas at this point, it is possible for foam to build up and carry emulsion out the gas outlet. Ineffective gas removal in the separator can cause some rolling of fluids in vessels at atmospheric pressure or in pressure vessels where operating pressures are exceeded. Injection of 1-5 ppm of a silicone prior to the gas separator (Fig. 5-7) may be beneficial to proper separator performance. Unless the silicone is an emulsion type, normally a 2%solution is made in diesel or other solvent and then injected into the well fluids.
Fig. 5-6. Beam pumping unit with settling tanks at ambient temperature.
139
Fig. 5-7. Flowline header with chemical injection and vertical gas separator.
Free water knockout
If the demulsifier is sensitive to excessive agitation, it may not have sufficient quiet time to release the water. When the water leg carries free oil, it may be caused by the demulsifier either being too slow or developing a poor interface. An associated problem concerning a water-dump valve is a vortex reaching up into the interface, causing an oily effluent water even if the interface is clear. This can be eliminated by the installation of a vortex breaker, installed over the interior of the outlet. Poorly operating interface level controllers can upset the interface level and cause a loss of water resulting in oil being dumped out. There is also a problem of improperly positioned inlet splash baffles within the vessel and this can cause problems regardless of the chemical. A lack of quiet time caused by an increase in production can also cause excessive water carry-over with the oil. Flow splitters
When flow splitters are provided with coalescing and knockout sections, excessive agitation can be detrimental, resulting in water carry-over. Corrosion of the weir boxes will also result in excessive water carry-over. Any part of the system where a pressure drop across a water outlet line occurs is susceptible to scaling, depending on the mineral content of the water. Such scaling can cause the control valve to either block open or shut.
140
Heat exchangers
Heat exchangers are often the culprit in faulty treating systems. Holes between the incoming wet stream and outgoing dry stream permit commingling of the two and, regardless of what is tried, oil going to stock will be dirty. Injecting water-soluble dye into the incoming stream will result in the rapid appearance of dye in the clean oil line if there is communication. Another method is to get a sample of oil from the top sight glass of the vessel and compare its grindout to a sample obtained from the dump valve. Should the oil from the sight glass be cleaner, it is quite possible that there is a hole in the heat exchanger. The same result would show up if there were a leak in the top plate of the treater which separates the gas section, or a hole in the upper part of the incoming line. Gun barrels
Leaks in internal conductors can contaminate the treated oil, whereas external conductors will not present this problem. Spreader variations can also cause problems. A cone-type spreader,,similar to an inverted funnel, causes very irregular spreading and may not provide a suitable coalescing water wash. Improper distribution can also cause turbulence and rolling. Almost all gun barrels are at atmospheric pressure, which can supplement the effects of rolling. Heating elements, if placed in the water, will provide satisfactory heating action. If by chance the heating elements are also in the oil phase, convection currents or rolling will prevent treatment and result in unsaleable oil. Corroded steam coils can cause turbulence and steam cutting of the oil. Heater treaters
If the desired treating temperature cannot be maintained: (1) Check the thermometers and thermostat. Make sure of the proper combustion mixture. (2) When the treater is constantly firing, determine the amount of oil-water throughput and the fire box rating. Normally, 150 Btu’s are required for raising 1 bbl of oil 1°F and 300 Btu’s are required if it is water. By calculation it can be determined if the heater treater is overloaded. (3) If the treater is overloaded due to an excessive amount of water, the use of a free water knockout and/or demulsifier that exhibits a faster water drop may allow the treater to operate properly. (4) If the treater is not overloaded, an inspection for fouling of the fire tube and sand on the bottom of the treater is required. If the pilot light will not stay lit: (1) Check to insure the pilot gas is dry. (2) Check the air-gas mixture for a blue flame.
141
If the interface is continually increasing in height: (1) Check for proper operation of the water discharge valve. (2) Check the lines and valves for salt or scale deposits. (3) Check bottom of treater for a blockage that will reduce the flow of water. If the interface is decreasing in height: (1) If the oil level is above the oil outlet, check for a loss of pressure on the treater and restrictions in the oil discharge line. (2) If the oil level is below the outlet, check the temperature and chemical injection. Other possible problems that will cause improper treatment are: (1) Foaming of the oil caused by the release of entrapped gas. This can be corrected by preheating the oil or the use of silicones. (2) Channeling in the filter section which will require repacking. (3) Unlevel or corroded spreader plates. (4) Solids or scale buildup within the treater. ( 5 ) Holes in internal piping. When everything else has been checked and stable levels are being maintained, but treating is still not adequate: (1) Check for foaming oil being carried out the vent line. This will usually indicate that emulsion in the foam is bypassing the settling section through the equalizer line directly to the top of the oil in the settling section. This may require a preheated coil. Severe foaming may be noticed in oil carryover in the vent line even before poor treatment is indicated. (2) Check filter section for channeling. Repacking may be necessary. (3) A buildup of solids or scale may drastically reduce the residence time in the treater. Solids buildup can usually be detected by feeling the difference in skin temperature around the bottom. Inasmuch as solids do not transmit heat to the skin of the vessels as readily as liquids, if cool spots are observed, a buildup is indicated. Solids buildup may also cause channeling from under the spreader plates. Unlevel or eaten-out spreader plates will also cause channeling. (4) Make sure the thermometers are not faulty and that the sight glasses are not plugged up. Either of these could lead to faulty judgements. Chemelectric treater
Problems in a chemelectric treater are usually indicated by low voltage or the pilot light blinking, dimming,. or going out. This is a signal transmitted from the electrode section, which is caused by the charged (lower) electrode shorting to ground. The upper electrode is grounded to the vessel and any conductive material between the electrodes can cause a short circuit. The charged electrode can also short to the vessel or to the oil-water interface. Any accumulation of water, BSW, iron sulfides, or similar material can result in a short circuit. The most common cause of shorting is an interface buildup.
142
The electrical supply to the treater should be checked and examined for blown fuses, burned out light bulbs, or tripped switch gears. If electricity is the source of trouble and a shorting situation is indicated, the water level should be lowered gradually. If the condition was caused by an interface buildup, then lowering the water level would lower the interface and treater conditions should return to normal, i.e., voltage will increase, amperage will decrease, and the light will become bright and steady. Increasing temperature or slugging the system with a proper chemical should clear up the buildup in the treater, but more than likely, some basic change in treating will be required to prevent frequent recurrence of this situation. If this procedure does not disclose the source of the trouble, further investigation will involve closer inspection of the transformer and testing of electrical circuit to pinpoint shorting conditions caused at entrance bushings, electrode, the insulated hangers, or the safety float switch. Produced fluids
Significant changes in emulsion characteristics occur rather infrequently and then usually slowly over a long period of time. Slow changes in treating, however, may also be the result of some slowly developing problem downhole or in the treating plant. Rapid changes in the emulsion may occur following the introduction of a new well, well workover, etc. Such changes may be temporary or permanent. Occasionally, an acid job may leave a quantity of acid trapped in the formation. ,Such trapped materials may break out weeks or even months later and cause temporary treating upsets. Frequently, a well will start cutting a tighter emulsion because of pump wear and may require additional chemical. A change in the water/oil ratio may also change the required quantity of demulsifier.
WASTE OIL TREATING SYSTEMS
There are no universal waste or slop oil treating plants, because there are no universal slop oils or volumes of waste oil processed. The following comments, suggestions and diagram (Fig. 5-8) are designed to aid in the selection or modification of a waste oil treatment plant. Treatment tanks
The size of the tank will depend on the volume of waste oil to be processed and the time required for complete treatment. A good starting point is to allow 24 hr for each batch treated. There should be sample valves at regular intervals on the side of the tank in order to check the treatment. Most tanks have these sample cocks at 2-ft intervals.
143
n O0
' 1#1oBBL -M 1
oo
con 'INSULAN
WASTL
oa ACCOMULAT" CHEMICAL
TREATMENT
Fig. 5-8. Waste oil treating facility.
If heat is to be used, the tank should be insulated. This will result in more rapid heating, less heat loss; and more economical fuel costs. When solids are present, there should be some method for removing them from the tank. Solids should not be allowed to build up in the bottom of a tank, because they will only make the next batch of waste oil harder to treat. Cone-bottom tanks are highly recommended and preferred in handling solidsladen waste oil. One common cone-bottom tank used in slop oil treatment has a 1000-bbl capacity., If a cone-bottom tank is not used, some method should be provided for removing the solids from the bottom of the tank. A swingline to remove the oil from the top can be used, if the separated oil-solids-water are sensitive to agitation or if the interface easily redisperses into the clean treated oil. Heating
Heat aids demulsification and reduces the amount of processing time that is requiied. The most desirable and safest method is to install steam coils from the bottom to the middle of the tank. A steam generator or source of steam that is capable of heating the oil to the desired temperature in 4-6 hr should be provided. Heat lances, heat exchangers, gas-fired U-tubes, or circulation through an external heater could also be used, but they are less desirable.
144
Even if the planned treatment temperature is 120-140°F, the system should be designed so that temperatures of 180-190°F can be achieved should they ever become necessary. A well-insulated tank will hold temperatures for a long time and will require less heat input. Agitation
No matter how good the chemical is, it would not work if it does not make contact with the emulsion. Tanks can be mixed using air, steam, recirculation, or commercial tank mixers. The effectiveness of these various methods depends on tank size and temperature. Generally, more mixing gives better results. The preferred method when installing new tanks or modifying older tanks is the use of commercial tank mixers. The number of mixers required depends on the size of the tank. Suppliers of commercial mixers can supply data and recommend the size and number of mixers required. If a tank is rolled with air or gas, it should be done for a period of about 3 hr at high temperature. The size of the tank will determine the time required for mixing. The 1000-bbl tanks are generally mixed for 3 hr and larger tanks are mixed for longer times. The type of mixing is not important. What is important is that the chemical is thoroughly mixed with the emulsion. Chemical addition
Chemical (demulsifier, acid, or caustic) poured in the top of a tank with 5-gal buckets does not promote good treatment. A chemical pump injecting chemical into a flowline, preferably into the suction of a pump, helps mixing and demulsification. Chemical should be added, at least, as the tank is filled. A chemical pump aids dispersion of the chemical when batch treating, but is mandatory for continuous systems. When the chemical is batch treated into a tank after it is filled, the type and amount of agitation becomes more critical than if the chemical were continuously injected into the line that was used to fill the tank. The chemical should also be added slowly as the tank is being mixed in order to achieve a better distribution of the chemical. Settling time
The amount of settling time required for a given slop oil depends on the gravity of the crude oil and the treating temperature. Lower API gravity oil and/or lower treating temperatures require longer settling times. Conversely, for a given oil, the settling time can be reduced by increasing the temperature. The settling time required to obtain treated oil can also be reduced by increasing the amount of demulsifier that is being used or changing to a different demulsifier. The normal settling time for most slop oils is 8-24 hr after the mixing has been stopped. In some cases, the settling time may be as much as 48 hr.
145
Solids
When a noticeable amount of solids is present, it may be necessary to use a wetting agent and/or caustic in addition to the demulsifier. The solids will be on the bottom of the tank if their specific gravity is greater than water and at the oil-water interface if their specific gravity is less than water. If the treated oil is to be processed through the same system that generated the slop oil, the solids must be removed in the waste-oil treating facilities and not be permitted to re-enter the system. The waste-oil treating facilities are the best place to remove the solids from the oil-treating system. STOKES LAW
The separation of two immissible liquids is governed by Stokes’ law (see Appendix 5.1), which gives the rate of fall of a small sphere through a viscous fluid. Stokes’ law states that when this sphere is under the influence of gravity it attains a constant velocity, which is given by the following equation:
where: g = 980 cm/sec2 (gravitational acceleration); pd = density of dispersed phase in g/cc; pc = density of continuous phase in g/cc; pc = viscosity in poises of continuous phase at settling temperature; r = radius of dispersed phase droplets in cm; and u = rate of fall of dispersed phase in cm/s (or rate of rise if negative). Examination of eq. 5-1 shows that the following three factors influence the rate of fall of the water droplets in a water-in-oil system. (1) As the viscosity of the continuous phase (oil) increases, the rate of fall decreases. Temperature will also have an effect on the viscosity. As the temperature increases, the viscosity decreases. (2) As the difference in density of the dispersed phase (water) and the continuous phase (oil) becomes greater, the rate of fall of the water drop increases. (3) The radius of the dispersed phase (water) has the greatest influence, because it is not only squared but can be increased considerably by coalescence. As the size of the water drop increases, the rate of fall increases. Initially, in petroleum emulsions the size of dispersed phase droplets is in the 3-10 pm range. In addition to Stokes’ law, there are four other factors discussed below that will aid in coalescence, thus allowing more rapid treatment: (1) Stokes’ law applies to static systems, whereas oil field treaters contain horizontal and/or vertical movement. It is this movement that encourages coalescence of water droplets yielding greater settling rates. (2) The use of a water leg in treaters not only removes free water, but also aids in coalescence and increasing size of water drops.
146
(3) As the temperature is lowered, the viscosity increases and, thus, velocity of water droplets decreases. As these water droplets fall, however, more effective sweeping and coalescence may occur yielding larger drop size. (4) The use of increased amount of chemical or using a different chemical may give rise to more interaction between the emulsion droplets and, thus, improve coalescence. SAMPLE QUESTIONS
(1) Describe the action of an emulsifying agent. (2) Describe Bancroft’s rule. (3) What relationship exists between the viscosity of an emulsion and the volume of the inner phase? (4) What is the function of a conductor (boot)? (5) Describe the heat factor in treating the heavier crude oil emulsions. ( 6 ) What effect do fine solids have on emulsions? (7) Derive Stokes’ law equation. (8) Explain how the law derived in (7) may be used to help in breaking emulsions. (9) Name a few chemicals that can break emulsions. (10) Plot a schematic diagram of temperature versus density for oil and water. Compare. (11) If viscosity of emulsion (O/W) at 70°F is 3200 cP, estimate its value at 170°F. APPENDIX 5.1-DERIVATION
OF STOKES LAW EQUATION
For Reynolds number, NR, below about 0.4, the drag coefficient, C,, for a sphere is equal to 24/NR. Thus, for a laminar or viscous flow the drag force in lb is equal to: (5.1-1)
where p is mass per unit volume in slugs/cu ft, u is velocity in ft/s, d is diameter of sphere in ft, A is the largest projected area in sq ft, and p is viscosity in lb-s/sq f t (or slug/ft-s). Inasmuch as the buoyant force, B, and drag force, D, are acting in upward direction, whereas the weight, W , in lb is acting down:
B+D= W
1 poise = 2.089X10-3Ib-s/sq ft.
( 5 .I-2)
147
or (5.1-3) where y is specific weight in lb/cu ft ( y = p g ) ; g (gravitational acceleration) = 32.174 ft/s2. Thus: ( 5 .I-4)
and
(5 .I-5)
REFERENCES Abraham, H., 1960.Asphalt and Allied Substances. Vol. 1. Van Nostrand, New York, N.Y., 6th ed., p. 52. American Petroleum Institute, 1975. Primer of Oil and Gar Production. Dallas, Tex. Bansbach, P.L. and Bessler, D.U., 1975.Cold Treating of Oil Field Southwestern Petroleum Short Course. Texas Tech. University. Barnickel, W.S., 1914. U.S. Patent No. 1,093,098. Bessler, D.U., 1980. Treating Emulsions from Enhanced Oil Recovery Projects. Chemical Marketing and Economics, A.C.S. Meet., 179,Houston, Tex., March 26,1980,pp. 201-208. Chilingar, G.V. and Beeson, C.M., 1969. Surface Operations in Petroleum Production. Elsevier, New York, N.Y., 397 pp. Clayton, W., 1954. The Theory of Emulsions and Their Technical Treatment. Chemical Publishing Co., New York, N.Y., 5th ed., 699 pp. DeGroote, M., 1926-1964.U.S. Patents Nos. 1,590,617through 3,148,154(total 546). Dodd, H.V., 1923.The resolution of petroleum emulsions. Chem. Met. Eng., 28: 249-253. Dow, D.B. and Reistle Jr., C.E., 1925.The physical chemistry of oil-field emulsions. U.S. Bur. Mines, Rep. Invest., 2692: 14 pp. Gunvitsch, L., 1935. The Scientific Principles of Petroleum Technology. Van Nostrand, New York, N.Y., 572 pp. Monson, L.T. and Stenzel, R.W., 1946. The technology of resolving petroleum emulsions. In: J. Alexander (Editor), Colloid Chemistry, Vol. 6. Reinhold, New York, N.Y., pp. 535-552. Nellensteyn, F.J., 1938.The colloidal structure of bitumens. In: A.E. Dunstan (Editor), The Science of Petroleum. Vols. 1-4. Oxford Univ. Press, Oxford, pp. 2760-2763. Petroleum Extension Service, 1962.Treating Oil Field Emulsions. 2d ed., rev. University of Texas, Austin, Tex., 86 pp. Traxler, R.N. and Coombs, C.E., 1938.Structure in asphalt. Znd. Eng. Chem., 30:440-443. Williams, A.R., 1953.A wash tank design. World Oil, 137: 203-213, 278-284.
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
149
Chapter 6
VAPOR RECOVERY VAREC
',G.V. CHILINGARIAN and S . KUMAR
INTRODUCTION
The first step towards energy conservation is probably taken in the oil and gas producing field itself in the form of vapor loss prevention. A properly designed and installed vapor recovery system is very effective in preventing all kinds of vapor loss. In addition, it reduces fire and corrosion hazards, pollution, and. associated problems. At the very outset, it is important to establish some clear-cut concepts about vapor recovery. Some operators consider a vapor recovery system to be a storage tank with a gas-tight roof; equipped with pressure apd vacuum relief valves. This kind of system is not a recovery system, but merely a primitive means of preventing some vapor loss under certain conditions, such as when no pumping operations are taking place and when no temperature change in the vapor space is occurring. A true vapor recovery system collects the vapors from the storage tanks at all times under all conditions. EVAPORATION LOSS
Evaporation loss is common when a volatile product is stored in any conventional oil tank (cone or dome-shaped roof structure). There are essentially three types of evaporation losses: breathing loss, filling loss, and boiling loss. Breathing loss
Breathing loss refers to the daily evaporation loss due to the normal cycle of atmospheric temperature change. During the daytime, the vapor space absorbs heat from the sun causing the air-vapor mixture to expand, and the temperature of the liquid surface gradually rises, increasing evaporation of the liquid. This causes some air and vapor to be vented to the atmosphere. During the late afternoon, the venting gradually decreases. At night, heat is radiated from the vapor space through the tank shell and roof. This results in condensation, creating a partial vacuum which draws in air through Permission was granted to reproduce portions of Varec, Inc. Handbook and Catalog.
150
the vent. Warm air rising in the center of the tank creates convection currents until air and vapor become stabilized. In the early morning, when the vapor space becomes uniform in temperature and hydrocarbon content, conditions are nearly static. In late morning, the heating and venting cycle begins anew. Filling loss
As liquid is withdrawn from a tank, the vapor space is filled by vapors evaporating from the surface of the remaining liquid. When the tank is refilled, a volume of air-vapor mixture is displaced, equal to that of the incoming liquid. The excess vented to the atmosphere is the filling loss. Boiling loss
Boiling occurs when a liquid is heated to its boiling point and the process of vaporization takes place. This may occur along the shell on the sunny side of the tank. Boiling causes excessive evaporation losses, but fortunately is not commonly encountered in lease tank batteries. EVAPORATION CONTROL
Conservation by evaporation control may be defined as any control of product evaporation, which results in the reduction or complete elimination of stored product loss to the atmosphere (Varec Division, Emerson Electric Co., 1970). It is quite difficult to quantitatively estimate evaporation losses and vent valve requirements, because of the numerous variables that govern these relationships, i.e., vapor space volume, vapor pressure of product, vapor pressure change with temperature change, rate of temperature change, etc. Data has been gathered, however, by various researchers, which enables an engineer to make close estimates. According to Varec (1976), methods of conservation vary in accordance with the degree of evaporation loss control desired as determined by consideration of economics which may be affected, plant safety, and the requirements of law. On low-pressure fixed roof tanks, conservation vent valves reduce evaporation losses by limiting the amount of air admitted to, or vapors released from, the vapor space. If vent valve settings and the tank are capable of maintaining pressures equal to or in excess of vapor pressure, vaporization will take place until the partial pressure of product vapors in the vapor space is equal to product vapor pressure at the prevailing liquid surface temperature. Any change of conditions, such as an increase or decrease of vapor space volume, and a corresponding increase or decrease of pressure in the vapor space, change of temperature in the vapor space, or change of product vapor pressure due to temperature change may result in loss of equilibrium. Thus, unless the tank is capable of withstanding the resulting increase or decrease in pressure, vapors must be released or air admitted to prevent structural damage.
151
Fig. 6-1.System of manifolded tanks. (Courtesy of Varec Division, Emerson Electric Co., 1976.)
A system of manifolded tanks, such as shown in Fig. 6-1, is commonly referred to as a vapor balancing system. This is a somewhat more efficient method of conservation, especially where operations can be arranged so that when one tank in the system is being filled, at the same time and at the same rate another one is being emptied. Theoretically, assuming no temperature change in the product or its vapors, vapors are transferred from one tank to another without loss. Actually, appreciable vapor losses through the vents will occur due to: (1) unequal pumping rates; (2) vapor space temperature changes; and (3) the probable increased rate of product vaporization in the tank being filled, due to agitation and product temperature increase from heat added by line friction and other outside sources. Any one of the aforementioned methods of conservation is effective in reducing evaporation losses. A third method, which completely eliminates the loss of vapors, is known as a.vapor recovery system. A typical tank installation is shown in Fig. 6-2.Basically, this is a closed system, wherein the vapor space pressure of individual tanks (or a group of manifolded tanks) is very closely regulated at some pressure well within the range of vent valve pressure and vacuum settings. In addition to the necessary control equipment, a complete system requires a means of pumping off excess vapors (which may be processed or stored as a supply for repressuring) and a source of makeup gas for repressuring. Excellent results have been reported from all installations where large volumes of high vapor-pressure product are handled (Varec, 1976).
152
PRESSURE IEXROS
BREATKR W V E M ARXSTER
WITH F
__
TNUtnGE
ITYFiCbL EbCH TANK 1
OF C W R E S S O R
WRY GAS
\
i l M E UP IUTOMATIC DRIP TRAP
Fig. 6-2. Typical vapor recovery system. (Courtesy of Varec Division, Emerson Electric Co., 1976.)
Several tank manufacturers produce conservation type storage tanks, which effectively reduce evaporation losses. Although there is an increasing trend toward this type of tank, particularly in new installations of large-capacity, fixed-roof tanks operating at pressures below 15 psig are the most common type of storage tanks. In this category, the cone-roof tank, operating at essentially atmospheric pressure, is the minimum accepted standard in present day practice (Varec, 1976). FUNDAMENTALS OF THE VAPOR RECOVERY SYSTEM
Vapor recovery equipment is installed on field production and storage tanks and on storage tanks containing crude oil and refined products at the refinery. In general, a vapor recovery system serves five purposes: (1) Economy-conserves valuable light components and, therefore, there is more income from sales. (2) Conserves oil (or product) gravity by reducing evaporation losses. (3) Reduces fire and explosion hazards by preventing air from entering and mixing with the vapor during out-pumping operations. (4) Reduces internal tank corrosion by preventing the addition of oxygen to vapor from humid air, which otherwise would be drawn into the tank.
153
( 5 ) Controls air pollution by preventing loss of vapor to the atmosphere. This is an important function in view of the E.P.A. regulations relating to smog and pollution control in oil field operations. The level of exposure of oil field personnel to toxic vapors, which are invariably associated with oil and gas production, is also reduced. When vapor recovery equipment is installed on field production tanks, the amount of gas which can be recovered per barrel of produced crude varies widely, depending upon several factors. Most important are gravity, temperature of the crude oil, and the drop in pressure which takes place following low-stage separation. Where a pressure drop of about 20 psi or higher occurs between the low-stage separator and the flow tanks, the volume of vapors recovered in California (U.S.A.) fields, for example, range from less than 20 to more than 60 cu ft/bbl of produced crude oil. In case of appreciable pressure drop, there is a high content of entrained liquids in the vapors, making them even more valuable (Varec, 1976). If elevated separators are operated at near atmospheric pressure, the great bulk of recoverable vapors will be removed from the crude oil before it enters the tanks. Because of agitation and especially if heat is used to achieve an oil-water separation, it may be worthwhile to recover the vapors even in this case. The air content in vapors is reduced by “repressuring” the tanks with dry gas from the absorption plant, for example, whenever required. For example, when a tank is being pumped out and the release of vapors from the oil within is not sufficient‘to fill the vapor space at the rate of pumping, instead of taking in air to fill the vapor space, dry gas from the absorption plant is used. Reduction of the air content in the tank reduces the fire hazard. The gravity of the oil is preserved as oxidation is reduced, thus lowering the temperature of the vapors and the oil surface. The presence of inert, dry gas slows down corrosion. The above discussion was directed towards economics of a vapor recovery system. The cost of the system is balanced against the profits accrued through conservation of the light fractions and the increased value of the product due to the preservation of product gravity. Indirect savings include reduced corrosion and fire hazards. It is assumed that an engineer would recommend installation of a recovery system only after carrying out a detailed financial analysis. Generally speaking, vapor recovery systems are viable in all cases where gas production is significant. The factors that affect the amount of vapor that can be recovered are: (1) Drop in pressure that occurs between the last stage of the separators and the tank. (2) Crude oil gravity, which is a function of various chemical and physical properties such as vapor pressure and content of light fractions. (3) Temperature of the crude oil. Heating for the purposes of demulsification is included here.
154
EQUIPMENT REQUIRED
The main functions of a vapor recovery system are: (1) removing pure vapor from the tanks during product inpumping operations and during thermal expansion of the vapors inside the tanks, as a result of atmospheric temperature increase or heat from the sun, etc; and (2) adding pure vapor to the tanks during product outpumping operations and during thermal contraction of vapors inside the tanks due to temperature reduction. This can be accomplished by various means from the standpoint of controlling the vapors to and from the tank portion of the system. A vapor recovery system should be automatic in operation, self-protective, and should function without maintenance for long periods of time, even in highly corrosive service. The equipment required can be briefly summarized as follows: (1) Devices to make the tanks gas-tight and spark-proof. May need self-closing, spark-proof gauges and thief hatches. (2) Pressure and vacuum relief vents, either in conjunction with the balance-line header or as an integral part of the hatches. (Refer to the API Guide for Tank Venting presented in the Appendix of this chapter.) (3) Vacuum and pressure regulators. (a) Regulator on vacuum line, to remove vapors. (b) Regulator to control automatic repressuring of tank with field gas, usually from the gas-oil separators. This will prevent air from entering tanks when vacuum develops. (4) Compressor to compress vapors from collecting system battery and send them into the field low-pressure gathering system. If there are vacuum gathering lines near the tank, compressors are not required. Figure 6-2 shows a manifolded tank system. A pipe manifold interconnects the tanks vapor-wise. The manifold is fitted with breather valves which serve to relieve excessive pressure or vacuum, which may be created as a result of abnormal system operation or malfunction of pressure-control equipment. Each breather valve includes a flame arrester for fire protection when and if the breather valve functions. An emergency relief manhole cover is used on each tank for relieving very excessive pressure buildup. Vapor-control regulators are provided: one to control the release of vapor from the system when the normal operating pressure within the system reaches a predetermined level; the other to control the addition of vapor to the system when the normal operating vacuum within the system reaches a predetermined level. A manometer can be used to visually determine whether or not the system is operating within predetermined pressure and vacuum limitations. Control of product inpumping and outpumping operations has great effect upon the proficiency of the vapor control system. It is desirable to limit the withdrawal of vapors from the tank system to a minimum. The ideal operation, although impractical to expect from the operator, is to predetermine and schedule product pumping so that vapors will be transferred between tanks within the tank system by means of the manifold. This would involve continuous and equal rates of product inpumping
155
and outpumping. Inasmuch as this is impractical, the basic vapor system at the tanks must be pressure balanced. The various relief valves, relief manhole covers, and regulators must be set so that when the normal product inpumping rate causes the vapor space pressure within the tank system to exceed a predetermined limit, the vapors will be released through the pressure (wet gas) regulator (Fig. 6-3). Conversely, when the normal product outpumping rate causes a vapor space vacuum
Fig. 6-3. Typical installation of pressure (wet gas) and vacuum (dry gas) regulators. (Courtesy of Varec Division, Emerson Electric Co., 1976).
156
within the tank system to exceed a predetermined limit, vapors will be added through the vacuum (dry gas) regulator. Settings of the regulators must be closer to atmospheric pressure than settings of the breather valves and the emergency relief manhole covers. Under normal operating conditions, the regulators must control only the flow of vapor to and from the tank system. In the event that a slightly abnormal pressure or vacuum increase occurs, which exceeds the capability of the regulators to control, then the breather valves relieve the excess. In the case of a greatly abnormal increase
Fig. 6-4.Typical installation of fittings on a low-pressure, cone-roof tank. (Courtesy of Varec Division, Emerson Electric Co., 1976.)
157
Fig. 6-5. Float-actuated gauge, powered by negator motor (controlled power). (Courtesy of Varec Division, Emerson Electric Co., 1976.)
158
in pressure or vacuum, which the regulator and breather valves cannot handle, the emergency manhole covers relieve the excess. Figures 6-2 and 6-3 show typical settings for the various equipment involved in the vapor control system at the tanks. In Fig. 6-4, a typical installation of fittings on a low-pressure, cone-roof tank is presented. A gas-tight, float-actuated gauge, powered by a motor, is presented in Fig. 6-5. It measures changes in liquid level as a function of float travel. The float acts upon a counterbalanced, nongraduated, perforated tape which moves a dial counter. Dial-counter reading minimizes the possibility of error, which was inherent in reading devices using a graduated tape.
DESIGN OF VAPOR RECOVERY SYSTEMS
In planning the installation of a vapor recovery system, one has to consider several factors: (1) volume of gas; ( 2 ) specific gravity of gas; ( 3 ) allowable tank pressures (allowable pressure drop and pressure drop actually required are both considered); (4) available vacuum for removing vapors; (5) available market or facilities for processing surplus gas; ( 6 ) availability of gas for repressuring for a fully closed system; and (7) type of tank installation (individual or manifold system). In arriving at the volume and'content of the vapors, one can use recording orifice meters to determine flow rates and the gas chromatograph test to establish their average gasoline content. It is believed advisable to make these surveys over a period of several days in order to offset variations in volume and composition caused by temperature changes. Where field vacuum lines do not exist, small compressors driven by electric motor or gas engine, usually ranging in horsepower input from 5 to 25 hp, may be used to supply vacuum for removing the vapors. These compressors are either air- or liquid-cooled, with single or double cylinder, and are single-stage. Where discharge pressures are required, two-stage equipment is employed. These small, fairly cheap compressors, operating continuously day after day, have permitted the installation of many systems which otherwise would not have been possible because of the necessity for running large lines considerable distances to a field vacuum line. To prevent any liquid, which may condense in the intake line, from being pulled into the compressor cylinder, scrubbers are placed just ahead of the unit. Whereas they are equipped with a bleed at the bottom so that any accumulated liquids can be drained as often as needed, mercury high-level switches are provided as a safety precaution. Likewise, to prevent overheating of the compressor, mercury switches are used to shut down the motor when temperatures become too high (Varec, 1970). To collect vent gas satisfactorily and safely, regulators (Fig. 6-5) must operate dependably at predetermined, supersensitive pressures, frequently as low as one- or two-tenths of an inch of water above and below atmospheric. At maximum flow, pressure drop through the vapor line (between the tank and regulators) plus dead weight of the vapor in the regulator control line must not exceed vent valve opening pressure.
159
Regulators should be capable of handling the pressure and vacuum requirements of the tanks as determined by calculations. Venting formulas [see American Petroleum Institute (API), 19681 or actual oil/gas ratios are guides to be used. Regulator sizes are obviously smaller than the vent unit sizes because of the required differences in valve design and pressure drop. As an example: the vent units may be required to handle normal venting of 30,000 cu ft/hr of gas at 0.35-02differential pressure drop and 20,000 cu ft/hr of inhaled air at 0.50-02differential drop. The wet-gas regulator would be required to handle the same 30,000 cu ft/hr at from 1 to 20 in. of mercury pressure drop, and the dry-gas regulator, 20,000 cu ft/hr of makeup gas at from 1 to 30 psi pressure drop. Thus, whatever capacities are
A
2.0
'-EMERGENCY
b
1.8-
RELIEF MANHOLE
FULL FLOW OPEN
1.61.4
U
BREATHER VALVE
-
W
a 3
1.2-
.r'
1
1.0
1
W
3
0.8-
a
0.6-
v) W v)
PRESSURE (WET GAS) REGULATOR
-
0.4
I
0.2-
ATMOSPHERIC
I
FULL FLOW OPEN
t
+
I
o 0.21-
FULL FLOW OPEN
U
lW
a 3
0.865 in. WATER
.-c
I.o
5
1.2
0.5 ounces/%q in.
:
EMERGENCY RELIEF MANHOLE
IBREATHER
U
a
VALVE
\'
1.8
'
+
1.73 in. WATER = 1.00ounces Irq in. FULL FLOW OPEN
PRESSURE OR VACUUM SETTING
Fig. 6-6. Pressure balance for a typical vapor recovery system. (After Hein et a].. 1969. fig. 5, p. 76.)
160
required of the regulator, are also required of the vent valve for any emergency eventualities. After regulator requirements have been determined, capacity charts for regulators must be consulted to select the correct size. Size of pressure- and vacuum-relief vents must be sufficient to insure against damage to tanks. They should have sufficient capacity to exhaust incoming dry gas in case the repressuring regulator sticks open, and to admit air to vent the suction line in case the wet-gas regulator sticks open. With the pressure-vacuum relief manhole covers set to pop open in the event that a trap valve sticks and throws the entire gas load on the battery, there is little danger of tank rupture. Certain auxiliary equipment for additional safety to the system is recommended. A sediment trap is required in order to protect the dry-gas regulator against foreign matter such as sand, millscale, etc. Foreign matter passing through the valve at high velocities can seriously damage the valve seats, thus requiring their repair. A back-pressure check valve should be installed downstream of the wet-gas regulator for the purpose of protecting the valve and system against downstream flame ignition, which might spread back up through the wet-gas regulator. A manometer should be installed in the common control line and not in any portion of the vertical riser, because the pressure at the "sensing point" is the true critical value. A manometer installed on the vertical riser would be subject to varying flows and would tend to give erroneous readings. The manometer serves both in initially setting the regulators and also in observing the condition of the system. An automatic drip trap, which is installed at the low point of the common control line, serves the purpose of draining condensate. An explosion- or pressure-relief valve ROOF
A b
CONNECTIONS
4 k
JL
Pressure drop Pressure drop P r e ssu r e d r o p equals 40diameters. equals 20diameters. n e g l i g i b l e .
VAPOR 9 0 "ELBOW WELDED
LINE CONNECTIONS 9 0 ' ELBOW WELDED 9 0 ' FORGED OR SWEDGED ELBOW
662 RATIO.!
Pressure drop Pressure drop Pressure drop equals 40 diameters. equals 20 diameters. equals 10 diameters. FULL
PIPE
AREA
ELBOWS
Ratio 2 pressure drop equals 5 diameters. Ratio 1.1 pressure drop equals 6 diameters. Ratio 1 .25 pressure drop equals 9 diameters. Ratio 1 .O pressure drop equals I0 diameters. Ratio .75 pressure drop equals 16 diameters. Ratio S O pressure drop equals 30 diameters.
-
S Q U A R E WELDED TEES
i t r Pressure drop eqsuls 40 diameters.
ROUNDED CORNER TEES
7 > I f Pressure drop equals 6 diameters.
%if
LINE
BALANCE
BRANCHES
ANGLE BRANCHES
R O U N D E D CORNER BRANCHIS
-=qqjG$ 90' angle pressure drop equals 40 diameters. 60' m g k pressure drop equals 17.5 diameters. 4 5 O angle pessure drop equds 8.8 diame8ers. 30' angle pressure drop equals 6.8 diameters. 13' angle pressure drop equals 3.9 diameters. M A I N
Ratio 2 pressure drop equals 5 diameters. Ratio 1.3 persure drop equals 6 diameters. Ratio 1.25 pressuredrop equals 9 diameters. Ratio 1.0 pressure drop equals 10 diameters. Ratio .75 pressure drop equals 16 diameters. Ratio .50 pressure drop equals 30 diameters.
VAPOR
LINES
WHEN CONNECTING LAPGB A N D SMALL LINES.
4
q-
'p'
< T A P E R 7
D' TAPER-,
.
5
-I. \
D
-
1'42 0
Pressure drop Pressure drop Pressure drop equds 40 diameters. equals 20 diameters. equds 9 diameters.
Fig. 6-7. Gas pressure drop in tank roof fittings and pipe bends. Equivalent length of straight pipe expressed in pipe diameters. (Courtesy of Varec Division, Emerson Electric Co., 1976.)
161
should be installed at the top of the vertical riser for the purpose of relieving excessively high pressures caused by an ignition emanating downstream of the wet-gas regulator. When hydrocarbon vapor ignites, it expands approximately 15 volumes (sometimes almost instantaneously), depending upon the composition. The size of the explosion-relief valve, therefore, must handle 15 times the volume of the vacuum pipe, calculated at atmospheric pressure. An expansion of 15 volumes increases the pressure 15 atm. This increased pressure will travel down the line in two directions from the point of origin, ahead of any flame, and will be relieved at the explosion-relief valves. A pressure balance chart for a typical vapor recovery system is presented in Fig. 6-6. In the design and installation of vapor recovery systems, it is important to keep in mind that the pressure drops involved are small. Vacuum- and pressure-relief valves are set to function somewhere between 0.5 and 2.0 oz/in.2 (about 1-4 in. of water). The total pressure drop through a vapor recovery system must be kept well within the limits of pressure- and vacuum-relief valve settings. For this reason, it is necessary that the header, laterals, and all other piping be properly sized and streamlined. If maximum benefit of the recovery system is to be gained, the piping must be of sufficient diameter to permit the vapors to be readily withdrawn. Also, angles, bends, and other resistances to flow should be kept at a minimum (Fig. 6-7). Although most engineers, experienced in flow calculations, have preference regarding flow formulas, the following formula may be used for estimating pipe sizes: q = 3550
/
Apd’ GI [l + ( 3 . 6 / d ) 0.03d]
+
where: q = quantity of gas, cu ft/hr; A p = drop in pressure, in. of water; G = specific gravity of gas at flowing conditions, air having sp. gr. = 1; d = internal diameter of pipe, in.; 1 = length of pipe, ft. The above formula applies only for straight pipes, with fairly smooth interior surface, and should only be used for flow of gas where pressures do not exceed 1 psig. Having determined the line sizes and pressure drops, allowances should be made for possible further expansion. In order to handle sudden overloads, it is usually advisable to slightly oversize the piping. Ordinarily, to hold the weight of the header and laterals to the minimum, the lightest-weight pipe obtainable is used. In some cases, sheet metal duct piping is satisfactory. The use of this light-weight pipe in flanged, shop-fabricated sections will materially reduce installation costs. STORAGE PRESSURES
When low vapor pressure products are being stored, the construction of most tanks is such that vent valves should not open at pressures and vacuum higher than
162 TABLE 6-1 Table of weight conversions for steel tank roofs (After Varec. 1976, p. 3)
a
Gauge and thickness, steel deck materials
Weight (Ib/sq ft)
Weight (oz/in.2)
Weight (in. H20/in.2)
16-gauge (0.0625”) 14-gauge (0.0781”) 12-gauge (0.1094”) 11-gauge (0.1250”) 10-gauge (0.1406”) 9-gauge (0.1562”) 8-gauge (0.1719”) 7-gauge (0.1875”) 6-gauge (0.2500”)
2.553 3.187 4.473 5.107 5.740 6.374 7.00 7.65 10.20
0.284 0.354 0.497 0.568 0.638 0.708 0.778 0.850 1.133
0.490 0.613 0.860 0.982 1.103 1.225 1.346 1.471 1.961
~
a
~~
~~
~~
~
Based on US. Standards.
0.5 oz/in.2. The internal pressure at which a tank may be maintained, however, is dependent upon design and condition of the tank. Generally speaking, cone-roof or low-pressure fixed-roof tanks should not be operated at pressures beyond dead weight loading of the deck. Vent valves for such tanks, therefore, should be sized and pressure settings so fixed that maximum normal relief requirements are attainable within this limitation (Table 6-1). It is equally important that the required flow capacity be obtained when the tank is breathing-in without developing a vacuum, which will cause damage. This requires consideration of tank size, design, the possibility of external loading, and other considerations, which make a general rule-of-thumb method for determining maximum safe working vacuum and maximum allowable vacuum impossible. According to Factory Mutual Engineering Division, Loss Prevention Bulletin No. 13.23, 1-f in. H,O negative pressure is the maximum allowable for vertical cone-roof tanks having 3/16-in. roof plates. No recommendation, however, is made for tanks of heavier construction and it is, therefore, suggested that established company policy or the recommendation of the tank manufacturer be followed. For high vapor pressure products, entirely different vent valve designs are used. These valves are equipped with dead weight vacuum pellets for settings of 0.5 oz/in.2 to 0.5 psig, depending upon tank construction. Higher pressure settings are obtained by weight-and-lever arm, spring loading, or even diaphragm-operated relief valves. It is important to note that increasing storage pressures reduces storage losses. This, however, is complicated by the changes in temperature and the consequent changes in vapor pressure, which may lead to vapor release during the day and air admittance at night. VENT VALVE PRESSURE SETTINGS
Dalton’s law of partial pressures states that the total pressure of a mixture of gases equals the sum of the pressures that each gas would exert if present alone at
163
the same temperature in the volume occupied by the mixture (provided, of course, that there is no chemical reaction or that there is no tendency for one gas to dissolve in the other). Partial pressure of an individual component in a gaseous mixture is equal to the product of the total pressure and the mole fraction of that individual component:
where: p,Y = partial pressure of a component in the vapor phase; 'TT = total pressure of the system; and y,, = mole fraction of the component in the vapor phase. For example, if a tank contains a volatile product and equilibrium exists between the liquid and vapor at an absolute temperature T, and at atmospheric pressure (14.7 psia), the vapor space will contain a mixture of gas having a pressure of ( pmin) and air at a pressure of (14.7 -pmin). If the temperature in the vapor space is increased to some other absolute value ( T 2 ) ,the partial pressure of the air will increase to (14.7 - pmin)T2/Tl, At this increased vapor space temperature, the product-surface temperature will also increase, resulting in an increased vapor pressure ( p , , ) . This makes a total pressure (absolute) in the vapor space equal to [(14.7 - pmi,)T,/T1 + p,,]. Thus, the theoretical storage pressure ( p s ) , at which equilibrium is re-established and at which no breathing losses occur, is equal to:
Obviously, this equation applies to standing storage only and its use is restricted to tanks which' are constructed to withstand pressures so calculated (Varec, 1976). In order to make use of this equation, it is necessary to obtain data on vapor space and liquid-surface temperature. According to Rogers (in Varec, 1976), maximum liquid-surface temperatures in the United States vary from 85 to 115°F. The maximum vapor space temperature is approximately 40°F higher than the maximum liquid-surface temperature, whereas the minimum vapor space temperature is 15°F lower than the maximum liquid-surface temperature. According to information published by the A.P.I. (1965), the maximum liquid-surface temperature on the Gulf Coast, Atlantic seaboard, and northern Middlewest in U.S.A. is about 100°F. In the Mid-Continent area and the arid Southwest of the U.S.A., temperatures as high as 115°F are encountered. On the West Coast of the U.S.A., at locations directly tempered by the Pacific Ocean, the maximum liquid-surface temperature may be as low as 80'F. The size and shape of a tank, product outage, color and condition of exterior surface paint, length of daily exposure to direct solar heat, and heat input from the introduction of warm product, all have a definite effect on inside temperature. The maximum liquid-surface temperature, resulting from atmospheric conditions only, is equal to or is slightly above the maximum atmospheric temperatures. If more accurate information is not available, therefore, it is suggested that the maximum
TABLE 6-11 Daily average temperature relationships between atmospheric temperature and those in the tank (All other sources of heat disregarded) (After Varec, 1976, P. 5 ) Tank
Maximum atmospheric temperature ( O F )
60 Max.liquidsurfacetemp. Min. liquidsurface temp. Max.vaporspacetemp. Min. vapor space temp.
65
60
65
50 100
55
45
105 50
70
70 60 110 55
75
80
85
90
95
100
75 65
80 70
85 75
90
95 85
100
80
115
120
125
130
135
140
60
65
70
75
80
85
90
105
105 95 145 90
115
120
125
130
125
130
150
115 105 155
120
110 160
115 165
120 170
95
100
105
110
115
110
110 100
165
atmospheric temperature be used as the maximum liquid-surface temperature and that the preceding data be accepted as a basis for estimating probable vapor space and liquid-surface temperatures. These temperature relationships are presented in Table 6-11. Example 6-1 Gasoline having Reid vapor pressure of 9 Ib/in2 is stored in a tank. Determine storage pressure (p,) required to eliminate standing storage losses from a tank operating at sea level and in a climate where maximum and minimum liquid-surface temperatures, as the result of atmospheric conditions, are estimated at 100°F and 90°F, respectively. Under such conditions, according to Table 6-11, maximum vapor space temperature will be 140'F and minimum temperature will be 85°F. Product has a vapor pressure of 8.1 psia at 90°F ( pmin)and 9.6 psia at 100°F (p,,). Thus: Tl = 460" 85' = 545"R T2= 460" 140" = 600"R By substituting in eq. 6-3:
+ +
p, = (14.7 -phn)-
T2 600 + p , , - 14.7 = (14.7 - 8.1)Tl 545
+ 9.6 - 14.7 = 2.2 psig
Altitude should be taken into consideration if storage tanks are located at altitudes where barometric pressures vary appreciably from the sea level pressure of approximately 14.7 psia. For example, atmospheric pressure at 4000-ft elevation is approximately 12.7 psia and, thus, required storage pressure to prevent standing losses from the same product subject to identical temperature variations at this altitude is equal to: 600 p , = (12.7 - 8.1)545
+ 9.6 - 12.7 = 2.0 psig
VALVE FLOW CAPACITY
The free gas capacity of a valve varies inversely with the square root of the gas gravity (with respect to air = l.O), Gg. In addition, qg varies directly as the square root of the ratio of the absolute standard temperature (520"R) to the absolute valve inlet temperature. Thus:
166
where: qg = free gas capacity of the valve; qa = free air capacity of the valve; Gg = specific gravity of gas (air = 1); and Ti = absolute valve inlet temperature, OR. The qg and qa must be expressed in the same units, e.g., ft3/min or m3/hr.
VENTING
Conservation vent valves mechanically limit the loss of vaporized product to the atmosphere. The same variables that make evaporation loss calculations difficult, must also be considered in arriving at vent valve flow rates required to protect tanks under conditions requiring maximum normal pressure or vacuum relief. Conservation vents also serve as safety equipment and are, therefore, designed for the worst conditions anticipated in order to safeguard against tank damage due to underventing. But in several instances, the conservation role is overemphasized, leading to a tendency toward size reduction to conform more nearly with capacity requirements based on normal operating conditions. This is a very poor economic decision, because damages to a tank may result in repair costs which are several times the cost of additional venting equipment. Also, installation of equipment, after a tank has been placed in service, is very often hazardous. Many formulas, charts, and tables have been developed, based on experience and available data, for vent valve flow capacity requirements. The American Petroleum Institute (A.P.J.) has reviewed all of these, and the A.P.I. Venting Guide (A.P.I., 1968) presented in the Appendix of this chapter, serves as the general guide for the petroleum industry at the present time.
FAST PAYOUTS FROM VAPOR RECOVERY SYSTEMS
Conservation venting equipment, properly installed and maintained, will often pay for itself several times a year in evaporation savings alone. Studies have shown that lease-tank evaporation is often reduced by as much as 50%, solely through the use of gas-tight tanks equipped with vapor conservation devices. There have been several reports claiming total return of conservation vent valve and installation costs on the first shipment of crude oil from gas-tight field production tanks. In addition, by reducing air content, corrosion of the tank and its equipment is reduced and fire prevention is enhanced. Another advantage is the fact that properly equipped tanks result in insurance savings and offer other safety features which cannot be estimated on a strictly monetary basis (Varec, 1976). In 1951, a survey conducted by Oil and Gas Journal (Stormont, 1951a, b) covering vapor recovery systems in California showed that very rapid payouts can be obtained where substantial volumes of vapors are available. They also indicated that in a number of cases reasonable payouts can be made from rather small volumes. The following illustrative examples are taken from the survey.
167
Example (A)-System payout from vapor recovery only
There were 4 tank batteries, each equipped with a conventional vapor system. Batteries consisted of 1600-bbl steel tanks in groups of 3, 5, 6, and 13. With the exception of the number of tanks, other system characteristics, such as gas content, trap pressures, etc., were similar. Cost of the vapor recovery equipment per battery ranged from $2000 for 3 tanks to $5500 for 13 tanks. In this example, $101,007 was realized from the recovered vapors from a total of 27 tanks. Total cost of construction and equipment was less than $14,000 (engineering time was not included). This was the case of an exceptional payout (Table 6-111). Example (B)
In this example, the recovered tank vapors amounted to 17,000 Mcf or 32 cu ft/bbl of crude oil produced (Table 6-IV). Average liquid content of the vapors was 14.4 gal per 1000 cu f t accounting for 244,800 gal or 7.5% of the 3,256,000 gal contained in all the gas. TABLE 6-111 Performance of Company A, Fresno County, California (After Stormont, 1951a) Month
Vapor recovered (Mcf/mo)
Liquid content (isobutane plus) (gal)
Value of liquid fractions a
Value of stripped gas
Total value of vapors
(9
($1
($1
1950
March April May June July August September October November December
8518 10,554 14,777 11,776 15,914 16,082 13,452 10,583 11,120 10.896
73,136 111,084 133,819 147,008 191,053 210,482 168,393 108,642 110,538 97.808
4388 6665 8029 8820 11,463 12,629 10,104 6519 6632 5868
886 1098 1537 1537 1655 1673 1399 1101 1156 1133
5274 7763 9566 10,357 13,118 14,301 11,503 7619 7789 7002
10,228
94,194
5652
1064
6175
136,900
1,446,157
86,769
14,238
101,007
1951
January Total a
Estimated on the basis of 6c/gal for butane and gasoline content. No credit given for propane content. Calculated on the basis of 16t/1000 cu ft for dry gas, allowing 35% for shrinkage and lease fuel. For period March 7-31 inclusive.
168 TABLE 6-IV Vapor recovery operations in Company B during January, 1951 (After Stormont, 1951a) (524,600 bbl of crude oil and 1,614,000 Mcf of gas produced) Monthly volume (Mcf) Tank vapors Low-pressure gas, 30 psi High-pressure gas, 480 psi
17,000 220,000 1,377,000
Totul
1,614,000
Average liquid content (gal/Mcf) 21-lb butanes propane gasoline
3.75 1.15 0.22
4.55 1.40 0.56
6.10 2.50 0.60
Total liquid total
‘Ontent
(gal)
14.40 5.05 1.38
244,800 1,111,000 1,900,200 3,256,000
Example (C)
As shown in Table 6-V, tank vapors recovered amounted to 18 cu ft/bbl of produced crude oil and 2.1% of the total gas gathered. The collected vapors contained 5.5 % of all recoverable liquid hydrocarbons. TABLE 6-V Vapor recovery operations at Company C during January, 1951 (After Stormont, 1951a) Monthly volume
Average liquid content (gal/Mcf)
Total liquid
(Mcf)
21-lb gasoline
butanes
propane
total
‘Ontent
Tank vapors Low-pressure gas, 30 psi High-pressure gas, 450 psi
8,500 339,000 49,000
3.90 1.28 0.70
3.40 1.21 0.78
3.50 1.75 1.20
10.80 4.24 2.68
91,800 1,437,360 131.320
Torul
396,500
(gal)
1,660,480
Example (0)
Although the percentages of gasoline fractions are not shown in Table 6-VI, one can see that 13,149 Mcf of recovered tank vapors amount to 52 cu ft/bbl of produced crude and 1.2% of the total gas gathered. Tank vapor accounts for approximately 12% of the butanes plus recovered. At the time of the survey in 1951, the cost of vapor recovery equipment was in the range of $400 to $500 per tank. Some installations, however, ran over $1000 per tank because a relatively larger compressor was required.
169 TABLE 6-VI Vapor recovery operation at Company D unit during October, 1951 (After Stormont, 1951b) (254,800 bbl of crude oil and 1,095,639 Mcf of gas produced) Monthly volume
Average liquid content (gal/Mcf)
(Mcf)
isobutane-plus
Total liquid content (gal)
13,149
8
105,000
Low-pressure and High-pressure gas
1,082,490
-
748,919
Total
1,095,639
Tank vapor
853,919
SUMMARY
It is estimated that, on the average, rapid payouts occur on installing vapor recovery systems, many in less than one year. Although vapors ,represent only a small fraction of the total gas recovered, the gasoline content may run as'high as 10% or more of the liquid in the total gas from field. Vapor recovery offers oil operators important additional revenue by recovery of light hydrocarbons, dollar value of natural gas saved, and increased value of stored oil due to preservation of product gravity. Other considerations, not directly measurable dollarwise, are reduction of fire hazard, oxidation, and corrosion. Fortunately, a good conservation practice can yield attractive profits to the oil operator who installs a complete, gas-tight, vapor recovery system.
SAMPLE QUESTIONS "\
(1) List five reasons for installing vaflor recovery systems on lease-flow tanks. (2) List four factors which affect the amount of gas that can be recovered. (3) Describe the equipment required for vapor recovery. (4) In outline form, present steps followed in designing vapor recovery equipment. ( 5 ) Draw a schematic diagram (flow diagram) of all the surface equipment; start at wellhead.
ACKNOWLEDGEMENTS
The help extended by Robert Siler, District Manager of Varec, and B.C. Wride is indeed greatly appreciated by the writers.
170 REFERENCES American Petroleum Institute (A.P.I.), 1965. Evuporution Loss of Petroleum from Storuge Tunks. American Petroleum Institute (A.P.I.), 1965. Guide for Venting Atmospheric and Low-Pressure Storuge Tunks. API RP 2000. 10 pp. American Petroleum Institute, 1968. Venting Atmospheric and Low-Pressure Storuge Tunks (Nonrefrigeruted). API Standard 2000, 1st ed., 18 pp. Hein, W.G.. Johnson, J.L. and Chilingar, G.V., 1969. Vapor recovery. In: G.V. Chilingar and C.M. Beeson (Editors), Surfuce Operutions in Petroleum Production. Am. Elsevier, New York, N.Y., pp. 67-86. Rogers, W.F., 1976. Method of calculating oil evaporation losses. In: Tunk Venting und Guuging: Tunk Equipment, Gus Control and Safety Devices. Handbook and Catalog No. P-8. Varec Inc. Stormont, D.H., 1951a. Conservation of lease-tank vapors. Oil Gus J., 50(2): 93-96. Stormont, D.H., 1951b. Tank-vapor recovery at Guijarral Hills effects payout in few months. Oil Gas J., 50(4): 85-87. Varec Division, Emerson Electric Co., 1910. Pollution und Gus Control Equipment. Bull. CP6003-A, 61 pp. Varec, Inc., 1976. Tunk Venting und Gauging: Tunk Equipment, Gus Control und Safety Devices. Handbook and Catalog No. P-8. Varec, 1979. Pollution und Gus Control Equipment, Vupor Recovery Systems, Gus Piping Systems, Technical Reference. Bull. 6003-B, Cypress, Calif., 52 pp.
171 A P P E N D I X 6 . I - V E q T I N G A M O S P H E R I C A N D L O W - P R E S S U R E STORAGE T A N K S ( N O N R E F R I G E R A T E D ) * (d N D A R D 2000, 1st ed., M A Y 1968.)
Scope This standard applics to the normal and emergency venting requirements for aboveground liquid petroleum s t o m tanks and aboveground and belowground refrigerated storage tanks designed for operation from 46 oe per sq in. vacuum through 15qsig pressure. The rcqwkements of this standard do not apply to floatingocMer-root tanks. Bn@neeringstudies of a particular tank may indicate that it is desirable to use a venting capacity other than thstatimated in accordance with this standard. Part I of this standard applies only to aboveground notrefrigerated liquid petroleum storage tanks. It out1$4(safe and reasonable practices for the normal temPART I-NONREFRIGERATED Determination of Venting Requirements Venting requirements are set forth for the followin4 conditions: 1. Inbreathing resulting from maximum out5ow of oil from tank. 2. Inbreathing resulting from contraction of vapors ca)lsed by maximum decrease in atmospheric tem1.0
pctatun. 3. Outbreathing resulting from maximum inflow of oil &to tanks and maximum evaporation caused by such iO8OW'
4. Outbreathing resulting from expansion and evaporation which result from maximum increase in atmospheric temperature (thermal breathing).
5. Outbreathing resulting from fire exposure.
2.0 Normal Venting Capacity Requirements Normal venting capacity shall be obtained without cxcceding the operating pressurc or vacuum which may be applied regularly to a tank without causing physical damage or permanent deformation to the tank. Total normal venting capacity shall be. at least the sum of the venting requirements for oil movement and hmaleffect.*
* Reprinted with permission of
perate-zone climate and normal operating conditions. The many abnormal variables which must be considered in connection with tank-venting problems make it impracticable to set forth definite simple rules which are applicable to all locations and all coaditions. Larger vents may be required on tanks in which oil is heated, on tanks whizh receive oil iron] wells or traps, and on tanks subjected to pipeline surgcs. Similarly, the use of flame arresters or other restrictions, which may build up pressure under certain conditions, m y requirc the use of larger vents on tanks. Part I1 of this standard applies only to aboveground and belowground refrigerated liquefied hydrocarbon storage tanks.
ABOVEGROUND TANKS However, lhe required capacity may be reduced for those prcducta whose volatility is such that vapor generation or
condensation, within the permissible vessel pressure operating
range, will provide all or part of the vcnthg requirewnts. Where noncondcn~blesM pmt, chi3 should be taken into .ccounL
2.1
Inbreathing Nacuum Relief)
Venting capacity requirement for maximum oil movement out of a tank should be equivalent to 560 cu ft of free air per hour for each 100 bbl (4,200 gal) per hour of maximum emptying rate, including gravity flow rate to other tanks, for oils of any flash point. 2.1 1
2.12 Venting capacity requirement for thermal inbreofhing for a given tank capacity for oils of any flash point should be. at least that shown in column 2 of Table 1. 2.2
Outbreathing (Pressure Relief)
Venting capacity requirement for maxinium oil movement into a tank and resulting evaporation: 1. For oil with a flash point of 100 F or above, should be equivalent to 600 cu ft of free air per hour for each 100 bbl (4,200 gal) per hour of maximum filling rate. 2.21
American Petroleum Institute, New York, N.Y.
172 2. For oil with a flash point below I00 F, should be equivalent to 1,200 cu ft of frce air per hour for each 100 bbl (4,200 gal) per hour of maximum EUing rim. 2.22 Venting capacity requirement for thermal ouibreathing, including thermal evaporation, for a given tank capacity:
3.0
Emergency Venting Capacity Requirements
When storage tanks are exposed to fire, the venting rate may be in excess of that resulting from a combination of normal thermal effects and oil movement. In such cases, the construction of the tank will determine whether additional venting capacity must be provided.
1. For oil with a dash point of 100 F or above, should be at least that shown in column 3 of Table 1.
3.1
2. For oil with a dash point below 100 F, should be at least that shown in column 4 of Table 1.
(maximum %-in. single-fillet weld) as described in Par. 3.5.2(c) and (e) of API Standard 650: Welded Steel Tanks for Oil Storage, the roof-to-shell connection will fail preferentially to any other joint, and excess pressure will be safely relieved, if the normal venting capacity should prove inadequate. In tanks built to these specifications, consideration need not be given to any additional emergency venting requirements.
TABLE 1-Thermal Venting Cepadtj Repoirementi (Expressed in cubic feet of free air per hour14.7 psia at 60 F.) Outbreathing (Pressure)
--
Tank Capacity (Barrels)
(Gallons) 1
60 100 500 1,000 2,000 3,000 4.000 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 60,000 70,000 80,000 90,000 100,000 120,000 140,000 160,000 180,000
2,500 4,200 21,000 42,000 84,000 126,000 168,000 210,000 420,000 630,000 840,000 1,050,000
Inbreathing (Vacuum) Flash Poinr All 100 F or Stocks Above 2 3 60 100 500 1,000 2,000 3,000 4,000 5,000 10,000 15,000 20,000 24,000 28,000 3 1,000 34,000 37,000 40,000 44,000 48,000 52,000 56,000 60,000 68,000 75,000 82,000 90,000
40 60 3 00 600 1,200 1,800 2,400 3,000 6.000
9,000 12,000 15,000 17,000 19,000 2 1,000 23,000 24,000 27,000 29,000 3 1,000 34,000 36,000 41,000 45,000 50,000 54,000
Flash Point Below 100 F
4 60 100 500 1,000 2,000 3,000 4,000
s.000 10,000 15,000 20,000 24,000 28,000 3 1,000 34,000 37,000 40,000 44,000 48.000 52,000 56,000 60,000 68.000 75,000 82,000 90,000
lnterpolile for intermediate sira.
NO=% I. For tanks with 8 capacity of more than 20,000 bbl (840,000gal). the reguircmenls for the vscuum condition are very close to the theorcticslly computed YSIYC of 2 EU f l of air per hour per squaw foot of total shell and roof area.
2. For tanlo with a capacity of lus than 20,OW bbl (840.000 gal). the thermal inbrcathing requirement for the vacuum condilion haa been based on 1 CY n of free sit per hour for each banti of lank capaciw. This is substantially equivalent to P mean rate of YPPDI 8paEbtcmwraturc changc of 100 F per hr. 3, For slocki with P flash point at 100 F or ~ ~ O V Cthe . outbreathing requirrmenl has k e n amumcd a8 60 percent of the inbreathing capacity requirement. I h c tank roof and shell tempemure E B M O ~ iise as rapidly under MY condition as they c a n drop. rush as during I sudden cold rain. 4. For stacks with P flash mint below IM) F. lhe thermal ~ i e w u r c venting requirement has bem ksumed equal to ihc vacuum rcqhiremeat in Order to ~ l l o wfor ~apoI(2~Uon at the liquid curface and far the higher specific grwity of the tnnk vapors.
Tanks With Weak Roof-to-Shell Attachment
On fixed-roof tanks with a roof-to-shell attachment
3.2
tanks Without Weak Roof-to-Shell Attachment
Where the tank is not provided with a weak roof-toshell attachment as described in Par. 3.1, the following procedure shall govern in evaluating the required venting capacity for fire exposure. 3.21 For tanks designed for pressures of 1 psig or below, the total rate of venting shall be determined in accordance with Table 2. (No increase id venting is required for tanks with more than 2,800 sq ft of exposed wetted surface. The basis for Table 2 is given in the Appendix.) 3.22 For tanks and storage vessels designed for pressures over 1 psig, the total rate of venting shall be determined in accordance with Table 2, except that when the exposed wetted area of the surface is greater than 2,800 sq ft, the total rate of venting shall be calculated by the following formula: CFH = 1,107
173
L = latent heat of vaporization of the specific liquid,
TABLE %-Total
Rate of Emergency Venting Required for Fire Exposure Vs. Wetted Surface Are. (Wetted area versus cubic feet of free air per hour14.7 psia at 60 F.)
Wetted Area
Venting Requirement (Cubic Feet perHour)
(Square
Fat) 20 30
21,100 31,600 42;lOO 52,700 63,200 73,700 84,200 94,800 105.000 126;OOO 147,000 168.000 190,000 211,000 239,000 265,000
40 50 60
70 80
90
Wetted Area (Square Feet)
Venting Requirement (Cubic Feet per Hour)
350 400 500 600 700 800 900 1,000 1.200 1;400 1,600 1.800 2,000 2,400 2,800 Over 2,800 *
100 120 140 160 180 200 ZSO 300 ' FW expo(ed wetted aurfrcca with more
288,000 312.000 354;000 392,000 428,000 462,000 493,000 524.000 557.000 587;OOO 614,000 639.000 662;OOO 704,000 742,000
3.26 T h e vent size may be calculated on the basis of the pressure which the tank can safely withstand.
lhm 2,8W 8q 11. l e e Par. 3.21,
3.22, .nd 3.24.
1. 1ntCfloate for htcrmcdiate Y B I Y U . area for B e lank or storage
2. I h c weNd
YCSIC~
shall be cslcylatcd
a8
SQhnO a d rphemld: the total cxposcd Surface Up to the mPxlmum horizontal diameter or to a height of 25 ft. whichever is srcaler. Horimntal tank: 75 pcrccnl of the total exposed surface. Vmical unk: the tom upmed area of L c shell within a muimum height of 30 fl above grade.
WheW CFH = venting rcquircmcnt, in cubic feet of free air per hour-14.7 psia at 60 F. A = exposcd wcttcd surface, in square feet. N m I: The foregoing forn~ulais based on Q = 21.000 A',. at given in API RP 520: Design and Installation of PressureRdieVing Systems in Refineries, Parf I-Design. The total heat absorbed, Q, is in British thermal units per hour. The constant, 1,107, is derived by converting the heat input value of 21,000 Bhl per hr per sq ft to standard cubic feet of free air by using tbs latent heat of vaporization and molecular weight of hexanc (aCr Apwndix of this standard for further detail).
3.23 The total venting requirements in cubic feet of free air determined from Table 2 and the formula in Par. 3.22 are based on the assumption that the stored liquid will have the characteristics of hexane, since this will provide results which are within an acceptable d e p e of accuracy for almost all liquids encountered. However, if a greater degree of accuracy is desired, the total emergency venting requirement for any specific liquid may be determined by the following formula:
Cubic feet of free air per hour = Y E L a Where: V = cubic feet of free air per hour from Table 2 or the formula in Par. 3.22.
provided for normal venting, since the normal thermal effect can be disregarded during a fire, and it can also be assumed that there will be no oil movement into the tanks. 3.25 If normal vents are inadequate, additional emergency vents of the type described in Par. 4.2 shall be provided so that the total venting capacity is at least equivalent to that required by Table 2.
NOTBl:
fouom:
in British thermal units per pound.
M = molecular weight of the specific liquid. 3.24 Full credit may be taken for the vent capacity
3.27 T h e total rate of emergency venting determined by Par. 3.21 or 3.22 may be multiplied by the appropriate one of the following factors when additional protection is provided: 0.5 when drainage away from the tank or vessel is provided. 0.3 when 1 in. thickness of external insulation is provided. 0.15 when 2 in. thickness of external insulation is provided. 0.075 when 4 in. thicknesa of external insulation is provided. Note 2: The values for insulation are based o n an arbitrary conductance value of 4 Btu per hr per sq ft per deg F per in. of thickness. Insulation shall resist dislodgement by fire-hose streams and shall be noncombustible. Note 3: Water films covering the metal surfaces can, under ideal conditions, absorb substantially all incident radiation. However, the reliability of effective water application is dependent upon many factors. Freezing weather, high winds, clogging of the system, unreliability of water supply, and tank surface conditions are a few factors which may prevent adequate or uniform water coverage. Because of these uncertainties, the use of an environmental factor other than 1.0 for water spray is generally discouraged.
4.0
Means of Venting
4.1
Normal Vents
Normal venting shall be accomplished by a pilotoperated relief valve, prcssure relief valve, pressure vacuum (PV) valve, or an open vent with or without a flame-arresting device. in accordance with the following requirements: 1. A pilotsperated relief valve, if used, shall be so designed that the main valve will open automatically and protect the tank in the event of failure of the pilot valve diaphragm or other essential functioning device. Relief valves equipped with a weight and lever preferably should not be used. 2. A pressure relief valve is appliciible on tanks operating above atmospheric pressure; in cases where a
174 vacuum can be created within a tank, vacuum protcction may be required. 3. PV valves are recommcndcd for usc on atnlospheric storage tanks in which oil with a flash point bclow 100 F is stored and for use on tanks containing oil which is heated above the flash point of the oil. A flanic arrester is not considered necessary for use in conjunction with a PV valve. 4. Open vents with a flame-arresting dcvice may bc used in place of PV valves on tanks .in which oil with a flash point below 100 F is storcd and on tanks conlaining oil which is heated above the flash point of the oil. 5 . Open vents may be uscd to provide vcnting capncity for tanks in which oil with a flash point of 100 1: or above is storcd, for heated tanks whcrc thc oil stordg': tempcrature is bclow the oil flash point, for tanks with a capacity of less than 59.5 bbl (2,500 gal) used for the storage of any product, and for tanks with a capacity of less than 3,000 bbl (1 26,000 gal) used for the storage of crude oil. 6. In the case of viscous oils, such as cu:baA and penetration grade asphalts, whcre the danger of t x k collapse resulting from sticking pallets or from pluggin2 of flame arresters is greater than the possibiiity of i l a m transmission into the tank, open vcnts may be used as an exception to the requirement for PV valvcs or flamearresting devices as called for in items 3 and 4.
Testing of Venting Devices The capacity of venting devices shall be established by any of the following: 1. In accordance with Par. UG-131 of Unfred Pressure Vessels, Sect. VIII of A S M E Boiler and Pressure Vessel Code (1965), with the exception that the determination of theoretical flow for the valve (actual discharge area) and the application of any coefficient to determine flow capacities shall be based on formulas which describe flow rates occurring below the critical pressure drop, rather than those shown in Par. UG131(e), item 2, which describe theoretical flow rates above the critical pressure drop. 2. By determining flow capacities of manhole covers with long bolts and similar venting devices by calculation, using a flow coefficient of 0.5, rather than by flow test. The flow formula used shall be suitable for noncritical flow and shall give proper consideration to actual flow area, flowing pressure, and features of the vent which would affect flow capacity. Data and calculations to show how capacities were determined shall be available. 3. By flow-testing at least one production model of every type and size of venting device under the conditions listed hereinafter. Tests may be made by the manufacturer if certified by a qualified impartial observer, or tests may be delegated to an outside agency. 5.0
5.1 4.2
Emergency Vents
Emergency venting may be accomplished by the use of: 1. Larger or additional open vents as limited by Par. 4.1, items 3 and 4. 2. Larger or additional PV valves or pressure relief valves. 3. A gage hatch which permits the cover to lift under abnormal internal pressure. 4. A manhole cover which permits the cover to lift under abnormal internal pressure. 5 . A connection between the roof and the shell which is weaker than the weakest vertical joint in the shell or shell-to-bottom connection. A tank with a roof-to-shell attachment (maximum ?bin. single-fillet weld), as described in Par. 3.5.2(c) and ( e ) of API Standard 650, is recognized as having a weak seam connection and, therefore, will not require emergency vents. 6. Other forms of construction demonstrably comparable for pressure relief purposes. 4.3
Vent Discharge
For tanks located inside buildings, discharge from vents shall be to the outside of the buildings. A weak roof-to-shell connection shall not be used as a means for emergency venting a tank within a building.
Capacity Data
Capacity data shall be presented in the form of curves or tables which give the volume of flow through both vacuum and pressure ports, and which cover the full range between the opening pressure (or vacuum) and the pressure (or vacuum) at which the ports are fully open. Capacity data for pilot-operated vents or devices which open fully at set pressure (or vacuum) may be expressed as a flow coefficient, this coefficient being the ratio of the flow of the vent to the flow of a theoretically perfect nozzle of the same diameter. 5.1 1
5.12 Capacity data shall indicate points of initial opening and final closing of the venting device; the closing noted as pressure (or vacuum) is decreased after fully opening the ports. 5.13 Capacity data shall be expressed in terms of cubic feet of free air per hour at 60 F and at a pressure of 14.7 psia. 5.14 Pressures shall be expressed in inches of water; however, auxiliary scales shall be expressed in ounces per square inch, and other units of measurement may also be included if desired. 5.15 Sufficient measurements shall be made at pressures in the viciuity of the opening points, particularly at 1.15, 1.25, and 1.50 times the opening pressure or vacuum, in order to clearly establish the flow capacity at those points.
175 5.16 The pressure or vacuum at which the valve disk reaches its fully open position shall be noted in the capacity data sheet. 5.17 Capacity data shall include a statement of the manner in which the valves were mounted and tested. If any fluid other than air is used in the test, this fact shall be noted on the test report, together with the temperature of the fluid actually used and its specific gravity at standard conditions. 5.2 Mounting of Venting Device for Test 5.21 To minimize the effect of entrance losses, the
venting device shall be mounted on the top of the test tank at a location near the center of an area which is essentially flat. The flat area shall have a diameter at least five times greater than the nominal diameter of the device to be tested. 5.22 The valve shall be mounted for test on a
straight-pipe nipple which has the same nominal diameter as the valve and a length one and one-half times the nominal diameter. The pipe nipple shall squarely enter the top of the test tank near the center of the flat portion, with the end of the nipple machined to 90 deg with the axis and flush with the inside of the tank. Rounding of the entrance in excess of a %-in. radius shall not be permitted.
5.23 Valves to be used on productica tanks or to be mounted on special nozzles or fittings shall be mounted on the test equipment in the same manner as they are to be mounted in the field, with their axes in the position normally used on a tank. 5.3
Test Tank
The test tank shall be so constructed as to prevent high-velocity jets from impinging on the venting device. 5.31
5.32
Provisions shall be made to dampen pulsations
in the test medium supply in order to avoid errors in flow metering. 5.4
Flow Metering
Air or other suitable gas shall be employed in testing the venting device. 5.42 Air or gas flow shall be measured in accordmce with Chapter 4, “Flow Measurement by Means of Thin Plate Orifices, Flow Nozzles, and Venturi Tubes,” of Part 5, “Measurement of Quantity of Materials,” of the supplement on “Instruments and Apparatus” to the ASME Power Test Codes. 5.41
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
177
Chapter 7
NATURAL GAS AND NATURAL GAS LIQUIDS BRUCE A. ECKERSON, ARNOLD L. JOHNSON and GEORGE V. CHILINGARIAN INTRODUCTION
The contribution of natural gas to the national supply of energy in the U.S.A. is presented in Figs. 7-1 and 7-2, with forecast through the year 2000. The demand for natural gas is now greater than the supply during periods of cold weather. 100
LEAR 90
80
70
i3
E
a 8"
AND GAS LIPUIDS
b
3 ;so
' a
J
40
3! I-
8'
6L 30 20
10
0
1920
1930
1940
I950
1960
1970
K)
Fig. 7-1. Distribution of United States energy market among various fuels, 1920-1980.
178
In the United States, over 75% of the required energy comes from petroleum and natural gas (DeGolyer and MacNaughton, personal communication, 1976). Natural gas and natural gas liquids contribute approximately 32% of this total. Figure 7-1 shows the distribution of the total United States market from 1920 to 1980: (1) since 1920, the water power has remained approximately constant at 4%; (2) nuclear energy has reached about 3% in 1975; (3) from 1920 to 1975, coal's share of the market has decreased from 78% to 18%;(4) from 1920 to 1975, oil's share has increased from 13% to 42%; (5) from 1920 to 1975, the natural gas share has increased from 4% to 30%; and ( 6 ) the consumption of natural gas liquids have increased from a negligible amount in 1920 to 3% of the energy market in 1975. Marginal gas reserves containing large quantities of inerts, such as carbon dioxide and nitrogen, are now being considered for development. Removal of inerts is expensive and these resen-s can be economically produced when natural gas prices are allowed to seek t h c - proper level. It is imperative that all available gas be produced in the most efficient way, which requires an understanding of natural gas and its properties.
NATURAL GAS
Natural gas is a naturally-occurring mixture of hydrocarbon and nonhydro,. carbon gases found in porous formations beneath the earth's surface, often in association with crude petroleum (AGA, 1965). Primarily, natural gas is a mixture of hydrocarbon molecules belonging to the paraffin series. The simplest hydrocarbon is methane, CH,; followed by ethane, C,H,; propane, C,H,; butanes, C,H,,; and heavier components as shown in Table 7-1, and in Fig. 7-3. These compounds have the chemical formula of CnH2,,+,. Natural gas is principally composed of methane with decreasing amounts of ethane, propane, and heavier components. It normally is partially or completely saturated with water vapor and may contain inert gases such as nitrogen and helium, and acid gases such as carbon dioxide, hydrogen sulfide, and mercaptans. There are many dry gas fields in which no liquids are produced, and the only processing required is dehydration, or perhaps, heating value adjustment. The other fields, in which a clear condensate is produced with the gas, are called condensate fields. In these fields the phenomenon of retrograde condensation often occurs; that is, liquid condenses out of the gas as the pressure is reduced (Katz et al., 1959). Thus, when the high-pressure gas is produced into a lower-pressure system through a choke, liquid forms. Liquid also condenses in the formation as the pore pressure drops and, unfortunately, does not completely revaporize before abandonment pressure is reached. Cycling plants are installed to prevent this loss of product. The produced gas is processed to remove the heavy ends; the residue gas, rather than being sold, is injected to maintain reservoir pressure. When the reservoir has
179 TABLE 7-1 Composition of natural gases Component
Type of gas field Dry gas, Los Medanos
Hydrogen sulfide Carbon dioxide Nitrogen and air Methane Ethane Propane Isobutane n-Butane Isopent ane n-Pent ane Hexane Heptane Octane Nonane
'
"
(mole ')
(mole ')
0
3.3 6.7 0 84.0 3.6 1.o 0.3 0.4
0 0.8 95.8 2.9 0.4 0.1 Trace 0 0 0 0 0 0
Gas condensate, Paloma
Sour gas, Jumping Pound
L(
%)
0 0.68 0 74.55 8.28 4.74 0.89 1.93 0.75 0.63 1.25
* 0.7
Natural gas separated from crude oil Ventura " 400 lb (mole %)
50 Ib (mole %)
Vapor (mole %)
0 0.30 0 89.57 4.65 3.60 0.52 0.90 0.19 0.12
0 0.68 81.81 5.84 6.46 0.92 2.26 0.50 0.48
0 0.81 2.16 69.08 5.07 8.76 2.14 5.02 1.42 1.41
0.15
1.05
4.13
100.00
100.00
100.00
6.30
100.0
100.0
100.00
" In California. ' In Canada. I
% ANNUAL CHANGE
1979
O
'
O
~
~
2000 ~ E
~
85-2000
YEAR
Fig. 7-2. Outlook on energy consumption in U.S.A.; forecast through the year 2000 (Copyright", 1985 by Chevron Corporation.)
180
HYDROCARBONS IN NATURAL GAS
METHANE
Ci CH4
ETHANE
c2
PROPANE
C3 C3He
C2H6
" YC -H
H-C-
k k Vt 'i' H-C-C-C -H k k h M
PENTANE
Ca
CsH12
Y H Y Y Y HC-~-C-C-C-H
~ A IH l H l H
Fig. 7-3. Principal hydrocarbons present in natural gas and their structural formulas. With few exceptions, natural gas consists of at least 95% hydrocarbons. The remainder is nitrogen, carbon dioxide, and, sometimes, small proportion of hydrogen sulfide. The principal hydrocarbon is methane with heavier hydrocarbons, i.e., ethane, propane, butane, pentanes, hexanes, and heptanes, being present in decreasing proportions.
been swept of heavy ends so that retrograde condensation no longer can occur, the field is produced in a normal manner. Gas is also produced with crude oil. This associated gas is normally rich in recoverable liquids, and construction of a gas processing plant may be economically justifiable even at relatively low gas production rates. At lower oil-gas separator pressures the heavy ends content of the associated gas is higher. Several gas analyses are given in Table 7-1 including dry gas, sour gas, gas from a condensate field, and oil-well gas produced at different pressures. Although these compositions are typical of many gases, it should be pointed out that nonhydrocarbon contents may be many times those shown. Methane-ethane ratio may also be as low as 4 : 1. This wide diversity of composition means that each gas must be individually evaluated and properly handled. This is the function of a skilled gas engineer, which is discussed later.
181
GAS PROCESSING PLANTS
A typical gas processing plant produces residue sales gas and a variety of liquid products including ethane, liquefield petroleum gas (LPG), and natural gasoline (Table 7-11). Inasmuch as early plants were intended to remove only heavier components intended for blending into motor fuel, the term gasoline plant came into being. Since World War 11, recovery had emphasized both on LPG and natural gasoline. LPG is usually defined as propane, butanes, or mixtures thereof. In recent years, the extraction of ethane for petrochemical feedstocks has become an important function of the gas processing plant. Operation of the plant involves the removal of impurities such as water, carbon dioxide, and sulfur compounds. The number and types of hydrocarbon products produced depends on the size of the plant and its location with respect to other facilities. Whereas older plants generally have their own fractionation facilities, newer plants more often produce a single demethanized or deethanized product, which is shipped by truck or pipeline to a central fractionating facility. Older plants had a relatively long life expectancy, being built during times of restricted production, and were designed for handling relatively small volumes of gas over long periods. Modern plants, on the other hand, have shorter lives, which is a result of unrestricted production, and the economics of larger short-lived facilities are less acceptable. A centrally-located facility can handle products from many sources and can, therefore, have a long economic life. Natural gas may be liquefied and transported by ship. Liquefied natural gas, normally referred to as LNG, is becoming an important source of energy for European, American, and Japanese markets. Large LNG plants are operating or planned for construction in Alaska, Algeria, Indonesia, Iran, Saudi Arabia, and
TABLE 7-11 Composition of natural gasoline (liquid volume per cent) Ten Section Gasoline Plant
Ventura Gasoline Plant 100 psia
Reid vapor pressure
38 psia
60 psia
Ethane Propane Isobutane n-Butane Isopentane n-Pentane Hexane Heptane Octane Nonane Decane
Trace 1.1 19.0 41.0 13.2 11.3 6.8 5.3 1.2 1.1
0.5 16.0 16.0 34.7 11.2 9.5 5.7 4.4 1.o 1.0
0.7 43.8 10.7 23.0 7.4 6.3 3.8 2.9 0.7 0.7
Trace
Trace
Trace
0 0 0.2 22.7 24.1 21.0 12.6 13.7 4.1 1.2 0.4
100.0
100.0
100.0
100.0
22 psia
182
other countries. Production of LNG, however, is not a normal surface operation in petroleum production. GAS SPECIFICATIONS
Sales gas specifications for natural gas include one or more of the following: water content, hydrocarbon content, heating value, specific gravity, acid gas content, temperature, and pressure. Water content
Water content is ordinarily expressed as pounds of water per million standard cubic feet of gas; however, dew point temperature and pressure also are used. The two methods have a definite relationship as shown by curves of water content as a function of saturation temperature and pressure (N.G.P.S.A., 1966). Common specifications are 1-, 4-,or 7-lb gas (i.e., lb water/Mscf gas) depending on the severity of conditions to which the gas will be exposed. In some warmer areas a maximum dew point of 50°F at delivery pressure is specified to assure that no water will condense in underground lines. Ground temperatures seldom fall below that level in those areas. Hydrocarbon content
Hydrocarbon content is usually indicated indirectly by either heating value or specific gravity. This procedure is not entirely accurate because composition can vary widely in a multicomponent system without changing either property significantly. Hydrocarbon dew points are sometimes specified or limits placed on gas enrichment with reference to specific components, expressed as gallons of liquefiable material per thousand cubic feet of gas (GPM, G/M, or gal/Mcf). Of particular importance are hexanes and heavier components which may condense or otherwise create problems in gathering or distribution systems. If significant amounts of carbon dioxide or nitrogen are present, neither gravity nor heating value will indicate hydrocarbon content. If both of these properties are measured, presence of one or both of these impurities will be indicated because either will raise gravity and lower heating value. Hydrocarbon content is not specified as often as it was in the past. The demand for energy has made purchasers less demanding. Gas is seldom sold strictly on a volume basis today, the price being adjusted for the heating value of the gas. The contract gas price is expressed as $/MMBtu; the price per Mcf is equal to $/MMBtu X gas Btu per cf/1000. Acid gas content
Acid gas content is specified according to the particular impurity. The usual specification for hydrogen sulfide is 0.25 grain/100 scf, although specifications as
183
high as 1.0 grain/100 scf are sometimes found. As a comparison, current OSHA standard for H,S in ambient air is 20 ppm (1.2 grains/100 scf); at that level, protective gear is required. Mercaptans are also expressed as grains/lOO scf; however, they do not often present a problem in gas sales because mercaptans are added as a warning odorant for natural gas. Carbon dioxide content may also be specified; an upper limit is commonly 5% by volume. There are a few reported cases of carbonyl sulfide (COS), although its occurrence in natural gas is rare. GAS TEST METHODS
The value of any specification depends on the availability of reliable test methods to determine the specific property. Reference is made to published test methods of the American Society for Testing Materials (A.S.T.M), Gas Processors Association (G.P.A.), and the Pacific Energy Association (P.E.A.; previously Western Gas Processors and Oil Refiners Association: W.G.P. and O.R.A.). Specific references are given at the end of this chapter (G.P.A., 1980a-d; N.G.P.S.A., 1962,1966,1972; P.E.A., 1943, 1950, 1966; W.G.P.&O.R.A., 1950a,b, 1955, 1956, 1965). The following methods are the most common. Water content
Water content is most often determined by measuring the dew point temperature at a fixed pressure with a commercial device (Bureau of Mines Tester). This indicator consists of a pressure chamber with a thermometer and a mirror that can be cooled with a refrigerant. The dew point is visually observed and the water content read from any standard chart (e.g., N.G.P.S.A., 1966). Some experience is required to differentiate between the water and the hydrocarbon dew points. Water content and dew point are relatively independent of gas composition. At very low water contents, a. suitable refrigerant may not be readily available and, in some cases, a continuous record is desired. In these cases a recorder using the conductivity of a hygroscopic salt is used. Carbon dioxide and air
Determination of carbon dioxide and air, as well as hydrocarbons, is most frequently done by gas chromatography. Carbon dioxide is also determined by Orsat analysis which is based on volume reduction of a known volume of gas after reaction of the carbon dioxide with sodium hydroxide solution. Air is sometimes estimated by determining oxygen by Orsat analysis with a special reagent (pyrogal101) and assuming a normal air/O, ratio. Hydrogen sulfide
Hydrogen sulfide is determined by the cadmium sulfide test in which a measured volume of gas is first bubbled through a cadmium solution to precipitate cadmium
184
sulfide and then titrated iodometrically. In the presence of very high concentrations of H,S the Tutweiler method is used by allowing it to react directly with iodine solution. A qualitative test for the presence of H,S is the use of moist lead acetate paper. Semiquantitative tests can be made with any of several “length of stain” tubes, in which a substrate is impregnated with a reagent that turns dark on contact with H,S. The length of darkening and the volume of gas are a measure of the H,S con tent. Specific gravity
Specific gravity, whch can be determined with a gravity balance, is commonly measured with a Ranarex or calculated from the gas analysis. Heating value
Calorimeters are used for direct determination of heating value. Inasmuch as the equipment is very expensive, however, it is more common to analyze the gas by chromatography and calculate the heating value from known properties of the individual components. A method for this calculation has been published by the G.P.A. This method is also used for specific gravity and compressibility determinations. Gas measurement
Measurement of natural gas usually involves inserting a restrictive orifice in the line and measuring the pressure drop across the orifice. Basic measurement data were developed by the American Gas Association, which appears in the publications of G.P.A., P.E.A., and the Southern California Meter Association. Chapter 2 of Volume I1 discusses the equipment and procedures involved in gas measurement in detail. NATURAL GAS LIQUIDS
Natural gas liquids can be classified as (1) ethane, (2) LPG (liquefied petroleum gas), or (3) natural gasoline. The LPG is normally restricted to propane and butane or mixtures thereof, with small amounts of ethane and pentane being present as impurities. Natural gasoline is considered by many to consist of pentane and heavier hydrocarbons, but the term is also applied to mixtures of LPG and pentanes plus (i.e., pentanes and heavier fractions). Liquid specifications
Liquid specifications as set by mutual agreement between buyer and seller vary widely, but approximate limits for commercial products can be summarized as follows:
185
Ethane Maximum methane content = 1.5%by volume. Maximum carbon dioxide content = 0.28% by volume. It is noncorrosive by using copper strip method. Propane A minimum of 95% propane by volume, a maximum of 1-2% butane, and a maximum vapor pressure which limits ethane content. The currently used vapor pressure is 208 psig, which is limited by the working pressure of DOT shipping containers (300 psig at 130°F.) Corrosivity, sulfur content, dryness, and specific gravity also may be specified. If propane is to be used as a motor fuel, the propylene content is limited because of its low octane rating. Butane The percentage of one of the butane isomers is usually specified along with the maximum amounts of propane and pentane. Other properties that may be specified are vapor pressure, specific gravity, corrosivity, dryness, and sulfur content. Butane-propane mixture In the case of butane-propane mixture, in addition to limits on nonhydrocarbons, the maximum isopentane content is usually stated. The particular mix is identified by vapor pressure or percentages of the components. Natural gasoline In the midcontinent area, natural gasoline is designated or sold on the basis of vapor pressure or, sometimes, by grade. The grade is defined by the vapor pressure and the percent vaporized at 140'F and 740 mm Hg. On the Pacific Coast, gasolines are usually sold on the basis of actual composition, which is determined from the Reid vapor pressure-composition curves prepared for each product source. Specifications for natural gasoline limit the Engler distillation end point to 375°F. End points of 300-32OoF, however, are more common. Liquid testing
Standard tests for LPG and gasoline are given in the technical bulletins referred to above. The P.E.A. Bulletin TS-352 presents an empirical method for computing Reid vapor pressure from analytical data. Engler distillation and copper strip corrosivity tests are A.S.T.M. methods. The G.P.A. has published an improved copper strip method using instrumental readings to replace colorimetric visual evaluation, GAS TREATING
Water removal
The most common impurity in natural gas requiring treatment is water. Water removal is necessary to prevent condensation of water and formation of ice or gas
186
PUMP
Fig. 7-4. Fluid process for gas treating.
hydrates. Liquid water can cause corrosion or erosion problems in pipelines and various equipment, particularly in the presence of carbon dioxide and hydrogen sulfide. Solid formation can plug pipelines, block control valves, and cause other operating problems. The simplest method of water removal is to cool the gas to a temperature equal to or below the required dew point. The range of this method can be extended if cooling can be done at higher pressures. For example, if it is necessary to produce a gas with a dew point of 50°F at 135 psig (water content of 60 lb/MMscf) when the best available cooling is 80"F, the required dew point can be realized by cooling to 80°F at 460 psig or higher, under which conditions the water content will be 60 lb/MMscf or lower. In a majority of cases, cooling alone is insufficient and, in field applications, usually impractical. Most dehydration requires the use of solid adsorbents or hygroscopic liquids. Solid desiccants include alumina, silica gel, and molecular sieves. Liquid agents for countercurrent contact are usually di- or triethylene glycol. Ethylene glycol can be directly injected into the gas stream in refrigeration-type plants (see Fig. 7-8). Figure 7-4 shows a typical fluid process for gas treating which may be used for glycol dehydration. Field units normally do not have a reflux drum or pump. The regenerator overhead is cooled with air fins at the top of the column or by an internal coil through which the feed flows. Reflux condenses and flows downward by gravity. Countercurrent vapor-liquid contact between the gas and the glycol produces an outlet dew point that is a function of the contact temperature and the residual water content of the stripped or lean glycol. Stripping (regeneration) of glycols is limited by the temperatures to which they can be heated. Both di- and triethylene glycol tend to decompose before they boil. The boiling point composition curves of both glycols are almost identical; better stripping and, hence, lower dew points can be
187
COOLING GAS-OUT
Y
m,
BED
BED
COOLER
n7-+by A
WATER
-@
HEATING
COOLING
GAS - O U T
+ ,
-
r
GAS-IN
I INLET
I
V
PROCESS
GAS-IN
obtained with triethylene glycol which has a higher decomposition temperature. Normal dew point depressions are 50-60°F below contact temperature for diethylene glycol and 70-75°F for triethylene glycol. Special techniques such as stripping of hot triethylene glycol with dry gas give dew point depressions up to 100°F or more. The Drizo Process, which uses heavy hydrocarbon vapors as a stripping medium for triethylene glycol, reportedly can give equally low dew point depressions. Vacuum distillation can also be used. Figure 7-5 shows a typical two-bed solid adsorbent treater used for dehydration. While one bed is removing water from process gas, the other is being heated and cooled. Sometimes a three-bed system is used: one bed is adsorbing, one is heating, and one is cooling. An added advantage is that the three-bed system can be used as a two-bed system while the third bed is being maintained or replaced. For this reason, a third bed is most often used where a dehydrator failure can result in a costly plant shutdown. Desiccants, as mentioned above, may be alumina, silica or molecular sieve. Silica gel and aluminas have capacities for water adsorption in the order of 7-8% by weight; cheaper bauxite (crude alumina): 4-6%; and molecular sieves: up to 15%. These are long-term design capacities and not the higher initial capacities often quoted. Molecular sieves are severalfold more expensive, but provide very low dew point outlet gas and are used almost exclusively for cryogenic plant feed preparation. Because of its high tolerance to hydrogen sulfide, silica is usually selected for sour gas dehydration. To protect molecular sieve beds from plugging by sulfur, alumina beds are sometimes placed ahead of the molecular sieves to remove the sulfur compounds. Regeneration is usually carried out with gas heated to 350-400'F. Heating is continued until the gas leaving the bed reaches 300-375°F. Although lower temper-
188
atures result in longer bed life, higher temperatures regenerate better. Molecular sieves require regeneration gas having temperatures of 450-500'F. Downflow is most commonly used during adsorption, with regeneration flow being in the opposite direction and cooling in the same direction as adsorption. This flow pattern requires the smallest vessel diameter because of higher permissible gas velocities. There is no tendency to lift the bed and cause breakage. Upflow, although more expensive because of initial vessel cost, is more conducive to trouble-free operation. Acid gas removal
Treatment of natural gas to remove the acid gas constituents (carbon dioxide and hydrogen sulfide) is most often accomplished by contact with an alkaline solution. The basic flow is the same as that shown in Fig. 7-4. Common treating solutions are aqueous solutions of the ethanol amines or alkali carbonates. A number of special treating agents have been developed in recent years the action of which is based on physical absorption and chemical reaction. Most of the newer agents are economically competitive only when the ,partial pressure of the acid gas is high (50-75 psi or higher). When only carbon dioxide is to be removed in large quantities, or when only partial removal is necessary, hot carbonate or one of the physical solvents is the most economical selection. The hot carbonate process operates at about 200'F and both the heat exchanger and the solution cooler are eliminated (Fig. 7-4). Hydrogen sulfide is sometimes removed by the iron oxide or dry box method. The gas is passed through a bed of wood chips or shavings impregnated with iron oxide, while the bed is being kept moist by circulation of a small stream of soda ash solution. The hydrogen sulfide reacts with the iron oxide to form iron sulfide and is regenerated by passing air through the bed, either continuously or on a batch basis. Aeration of the bed converts the iron sulfide to elemental sulfur and iron oxide. The method is suitable only for small quantities of sulfur, inasmuch as only about 90% removal per bed can be realized. If several beds in series are required, the process is not economic. The total sulfur removed by a bed is limited, because it becomes clogged with elemental sulfur and must be discarded. Modern environmental considerations may require that impurities should not be discharged to the air. Most hydrogen sulfide removal processes return the hydrogen sulfide unchanged. If the quantity involved does not justify installation of a sulfur recovery plant, usually a Claw plant, then a process must be selected which produces elemental sulfur directly. The dry box is suitable for the removal of small quantities of hydrogen sulfide, whereas larger quantities require a continuous process such as Ferrox or Stretford. Ferrox process is based on the same reactions as the dry box method except that it is fluid and continuous. Stretford is a licensed process using a solution containing vanadium salts and anthraquinone disulfonic acid (ADA). In an excellent book, Dr. R.N. Maddox of Oklahoma State University discusses these treating methods in detail (Maddox, 1974).
189
Nitrogen removal
Nitrogen is sometimes found in sufficient quantities to lower the heating value of the gas. Such a gas can sometimes be sold at reduced prices if it can be blended with a gas having a higher heating value. Some gas reserves were left undeveloped in the past because the low energy content in the gas would not justify treatment. At current high gas prices, however, plants for the removal of nitrogen are being considered and a few have been installed. Nitrogen removal requires liquefaction and fractionation of the entire gas stream which is quite expensive.
LIQUID EXTRACTION
Recovery of liquid hydrocarbons can be justified either because it is necessary to make the gas salable or because it is economic to do so. With the increasing scarcity of natural gas, purchasers are less critical and processing for liquid recovery is usually based on economic considerations alone. Figure 7-6 shows the value of recoverable liquids as natural gas energy. The justification for building a plant depends on the spread between the value of enriched gas containing heavier hydrocarbons and the price of lean gas plus the value of extracted liquid. The spread must be sufficient to pay operating costs, amortize the plant cost, and provide an adequate rate of return on capital. In many cases, particularly when the volume of gas is small, processing cannot be justified. If salability of the gas is the only reason for processing, then liquid removal usually is a field operation using either crude oil enrichment, adsorption, or
30
m
c
8
10
0
I.o
1.5
Dollar8 per million BTU
Fig. 7-6.Value of liquids as gas.
21)
2.5
190
refrigeration processes. If liquid recovery is economic, the extent to which extraction is carried out will depend on the availability of a market for the products and heating value limitations of the residue gas. Most of the liquid recovery plants recover a substantial portion of the propane and essentially all of the butanes and heavier hydrocarbons comprising gasoline. These products can be readily moved by rail, truck, and pipeline. Ethane recovery depends on availability of a product pipeline, although small amounts of ethane are moved by truck or rail when mixed with heavier hydrocarbons. Sufficient amount of ethane, however, must be left in the gas to meet contractual requirements for heating value. Crude enrichment The purpose of crude enrichment is to produce two products: sales gas and enriched tank oil. The tank oil contains more light hydrocarbon liquids than the virgin crude oil and the residue gas is drier (leaner). Inasmuch as crude oil is finally separated at atmospheric pressure, only those fractions can be added that can be retained at that pressure. Every crude oil enrichment process, therefore, must, in some manner, remove light ends from the oil to make room for the gasoline-and the LPG fractions. One of the simplest and yet least often recognized methods is manipulation of the number and operating pressures of the gas-oil separators (traps). In rare cases, separator temperatures also can be varied. Selection of trap pressures will also affect the nature of other processing steps and significantly affect the amount of gas compression. One method of removing light ends is using pressure reduction (vacuum conditioning). A typical process of this type is shown in Fig. 7-7. Heating, stripping with SALES GAS
p
bR
COOLER I
COMPRESSOR
CRUDE OIL
I
-k
1 PUh
Fig. 7-7. Crude oil enrichment process.
TO FUEL OR COMPRESSION
191
dry gas, or a combination thereof are also used. Generally, stripping is done at low pressure, after which the crude so stripped is pumped to high pressure to act as an adsorbent. The enriched crude oil is then reduced to atmospheric pressure in stages or using fractionation (rectification). Crude oil enrichment is used where there is no separate market for light hydrocarbon liquids, or where the increase in API gravity of the crude will provide a substantial increase in the price per unit volume as well as volume of the stock tank oil. Adsorption
When gases are relatively lean, adsorption-type units are installed for liquid recovery. Equipment is similar to that used for dehydration (Fig. 7-5),and operation is identical except for the length of the cycle. Dehydration is normally accomplished in eight hours or longer and permits saturation of the bed with water, resulting in the displacement of adsorbed hydrocarbons. Hydrocarbon recovery requires 30-60 min per cycle and most of the adsorbed hydrocarbons are retained. The size of the unit depends on the amount of liquid to be adsorbed and, therefore, the process is not economically feasible for rich gases. Refrigeration processes
Refrigeration processes can be classified according to the source of the refrigeration and the nature of the basic separation step. A typical flow sheet is shown in Fig. 7-8. In this case, mechanical or compression-type refrigeration is used to reduce
I
SALES GAS
I
f FUEL GAS
t
192
the temperature, and the basic separation is simply a phase separation followed by liquid stabilization. When wellhead pressures are high and large pressure drops can be used, expansion across a choke (Joule-Thomson effect) will supplement mechanical refrigeration or even supplant refrigeration by an outside medium. Propane and ammonia are the most common refrigerants. Ethylene glycol is injected into the system at points where icing or gas hydrate formation can occur. The glycol is recovered from the main separator and regenerated in the manner shown in Fig. 7-4. The most common process used today, when liquid recovery can be economically justified, is the one which uses a turboexpander to produce the necessary refrigeration. Very low temperatures and high recovery of light components, such as ethane and propane, can be attained using this process, a typical flow chart of which is presented in Fig. 7-9. Inlet gas is dehydrated in molecular sieve beds and cooled by heat exchange. The separated liquid containing most of the heavy fractions is fed to a demethanizer column. The cold vapors are expanded through a turbine which is loaded by a compressor wheel on the same shaft. This removal of energy from the gas results in much lower temperatures than are possible by ordinary Joule-Thomson expansion. The expander outlet is a two-phase stream that is fed to the top of the demethanizer column, which serves as a separator. The liquid is used as the column reflux and the separated vapors combined with vapors stripped in the demethanizer are exchanged with the feed gas. The heated gas, which is partially recompressed by the expander compressor, is further recompressed to sales pressure in a separate compressor. Inlet gas is used for heating reboiler, which makes the
193
FGAS UEL GLYCOL I SALES
INTAKE GAS
J
GAS
t
v HEAT EXCHANGER
a
2
s
-43
CONDENSATE
Ir \--1
-To
GLYCOL WATER
I
I
REBOILER
1
Fig. 7-10. Vapor rectification process.
process very efficient, i.e., the fuel is only required for dehydrator regeneration and for recompression. When the inlet gas is rich in liquefiable hydrocarbons, the inlet heat exchange is supplemented by mechanical refrigeration ahead of the primary separator. Vapor rectification processes, typified by Fig. 7-10, also employ mechanical refrigeration, but the basic separation is accomplished in a series of steps in the vapor-rectified column. Refrigeration is applied to the column overhead to produce reflux and, sometimes, to partially condense the feed. If the recovery of lighter liquids is sufficiently high, the reboiler may be placed on the vapor rectifier and the stabilizer is not used. The process is very flexible within the limits set by heating medium, refrigeration system, and equipment size. Production can be varied simply by resetting the top and bottom temperatures. Absorption
Up until the early 1970s most hydrocarbon recovery plants involved oil absorption process. These complex plants tend to consume excessive fuel and are difficult to operate. In a time of energy conservation and the trend towards construction of short-lived plants, oil absorption process has become economically undesirable in most new applications. Although the majority of operating plants still use this process, very few newly-constructed plants utilize it. Oil absorption process involves countercurrent contact between stripped or lean oil with the incoming wet gas, (Figs. 7-11 and 7-12), at such conditions of temperature and pressure that the desired amounts of the liquefiable components are dissolved in the oil (Figs. 7-13 and 7-14) together with lesser amounts of lighter components. Refrigeration is frequently used to obtain lower (more favorable) temperatures. The remainder of
194
R e s i d u e G o s lo S a l e s
Wire Mesh Demirler
Lean Oil
Troy
Bubble C o p
Down Spoul
W c l Gor
Rich Oil
u Fig. 7-11. A schematic diagram of bubble tray absorber.
the plant can be divided into the following functional sections: (1) separation of light ends from the oil; (2) separation of absorbed materials from the oil; (3) final removal of light ends from the raw product; and (4) separation of the raw product into various finished products. A typical oil absorption system is illustrated in Fig. 7-15. The extraction of a given component in the absorber depends on (1)the number of equilibrium contacts (theoretical trays) in the absorber during countercurrent flow of gas and liquid and (2) the absorption factor A defined as L / V K , where L = moles of liquid, V = moles of vapor, and K = equilibrium ratio. Inasmuch as the L / V is a molal ratio, a low-molecular-weight oil will yield a large L for the same volume of oil circulated. Refrigeration, intercooling, and presaturation of the oil with light components are all used to reduce effective absorber temperature and increase the absorption factor by reducing K,the equilibrium ratio (mole fraction of a component in the gas: mole fraction of component in the oil). Increase in pressure also reduces K up to a point;
195
Fig. 7-12. Bubble cap (a, b) and bubble tray (c, d).
LEAN OIL RATE O A L S ~ C F
Fig. 7-13. Relationship between the propane recovery (in 76) and lean oil/gas ratio (in gal/Mcf). Ventura gasoline plant, California.
196
o/o PROPANE RECOVERED
Fig. 7-14.Relationship between the propane recovery (in %) and recovery of butane and pentane. Ventura gasoline plant, California.
however, after reaching a minimum at some high pressure (500-600 psig), the value of K starts to increase. Absorber pressure is usually set by the sales or end-use pressure. Absorbers used in cycling plants are operated in the 1400-2000 psig range, because it is more economic to cool or circulate more oil than to greatly increase the recompression horsepower. The rejection of unwanted light ends may be carried out in one or more steps. Methane is nearly always removed in the rich-oil rectifier by pressure reduction and heating. In many plants this column also removes the ethane. Following rich-oil rectification, the absorbed material is removed from the oil in the stripper or a still. If a heavy absorption oil is used, stripping is commonly done by preheating the oil and countercurrently contacting it with steam. Steam is used because it is easily condensed and is immiscible with the raw product. With low-molecular-weight oils,
197
I
S
CONDENSER
COOLER
Fig. 7-15. Oil absorption process.
the still feed is heated by ‘exchange with the still bottoms and a fired heater is installed as a reboiler. Low pressure is conducive to good stripping, whereas high pressure aids condensation. Although most operations represent a compromise between these two factors, some plants use two stills in series: one at high pressure
ISOBUTANE
PROPANE
BUTANE
XED N4TURAL GASOLINE FROM DISTILLATION AREA REBOILER
DEPROPANIZED GA5OLINE
I
I Fig. 7-16. Flow diagram of fractionation area.
I
198
to condense the light ends and the second at low pressure to insure good stripping of the heavier gasoline fractions. If the oil is not well stripped, the lighter components in the oil will be vaporized in the absorber and lost in the residue gas stream. Fractionation As was mentioned earlier, the trend is toward construction of large, centrally located, fractionating systems; however, at times local systems may be justified. The order of fractionation (i.e., which product is produced first) can vary widely. There are two general rules of thumb which can be substantiated by economic evaluation: (1) remove any unwanted light ends in the first column so that the other columns can run totally condensing; and (2) remove the largest product stream first. In practice this means that the first column will be a deethanizer unless (1) all ethane was removed in the rich-oil rectifier or (2) an ethane product is being made and shipped as an ethane-propane mix. Normally propane or ethane is the largest volume product and the depropanizer is the next column; they are followed by the debutanizer and the butane splitter (Fig. 7-16). Using this design each column can operate at a successively lower pressure and streams will flow from column to column without pumping. On the other hand, if propane recovery is low, it may be desirable to debutanize first in order to get rid of the large gasoline stream and then depropanize, followed by a splitter if there is one.
GAS GATHERING
Custody transfer of the gas normally takes place immediately downstream of the trap, and it is the responsibility of the gas plant operator to gather the gas for processing. Substantial savings in compressor horsepower can be realized if multiple stages of separation are used. This means a partial duplication of the gathering system to bring gas at two different pressures to a compressor station. A careful economic study must be made to evaluate the advantages of one system over the other during the life of the operation. Generally, the processor gathers the gas at low pressure and compresses it to the processing pressure which is usually the residue gas sales pressure (see Fig. 7-27). In most gas processing units the greatest capital investment is for the gas compressors and auxiliaries. If the gas gathering lines are short, it is advantageous to locate compressors at the plant where they can be attended for more efficient operation. It is a good rule of thumb that length of gathering lines to a compressor suction should never exceed five miles. Field compressors, designed for unattended operation, should be used for greater distances. In the later stages of a field’s operation, when gas volumes have declined and gathering lines may be oversized, it may be possible to bypass the field compressors. Laying of pipelines (pattern-wise), size of pipelines, and location of compressors should be based on economic considerations, both current and future.
199
i COOL1
n
Fig. 7-17. Example of a gas-gathering system for an oil field.
If the wells are in a gas field, rather than an oil field, the gathering lines are normally-sized to deliver the gas at utilization pressure. As the wellhead pressure declines, compression is added to maintain the flow. Hydrate formation in gas gathering lines is often a serious problem, particularly if high pressures, low temperatures, and long distances are involved. Gas hydrates, which resemble wet snow and have a crystalline structure, can plug lines and valves if allowed to form. Hydrates may form at temperatures considerably above the freezing point of water; for example, at a pressure of 800 psi a water-wet, 0.7-specific gravity gas will form hydrates at a temperature of approximately 65°F. There are three methods commonly used to prevent hydrate formation in gas gathering lines: (1) Operation at temperatures and pressures at which hydrates cannot form. Wellhead gas heaters are sometimes used to heat the gas entering the gathering system, such as at the Shell’s Jumping Pound Plant in Canada. (2) Dehydration of natural gas at the source before it enters the gas gathering lines. This system is used at Prudhoe Bay, Alaska (Wilson, 1974) and at the Ekofisk Field in the North Sea (Kennedy, 1972). (3) Injection of hydrate inhibitors into the gas gathering lines. Methanol is commonly used where only a small amount of inhibitor is required. It is sometimes used for large systems particularly when salt water is present. For example,
200
methanol is injected into the gathering line at the offshore production platform for the Viking Field in the North Sea. A methanol-water mixture is separated from the well fluid at the associated onshore facilities. This stream is subsequently fractionated to recover methanol as a top product and salt water as a bottom product. When salt is not present, ethylene glycol is sometimes injected into the gathering line to prevent hydrate formation. Dilute glycol is separated at the gas processing plant for reconcentration and is recycled back to the inlet of the gathering system. This procedure was followed in the offshore Molino Field near Santa Barbara, California. Two-phase flow is frequently encountered in gas gathering lines. This occurs when warm gas enters the system and subsequently cools to a temperature below its dew point and/or when well fluid consisting of gas, condensate, and water enters the gathering system. Under certain flow conditions, long sections of line may fill with liquid, thereby increasing the pressure drop in the system. Liquid can be displaced out of the line to improve its efficiency by running spheres (“pigs”) through the line. It may be necessary to install a large vessel or “slug catcher” at the outlet of the gathering lines to collect these large volumes of liquid. The slug catcher acts as a surge to level out the liquid flow to the processing facilities. An example of this is the 3000-bbl capacity slug catcher at a pressure of 1050 psig installed in the onshore facilities of the Viking Field. Although only gas transmission lines and gathering lines in incorporated areas are subject to regulation by the Federal Office of Pipeline Safety, prudent operators should design and install all pipelines in accordance with those standards. Steel lines must have certain minimum wall thickness and be designed for the required operating pressure and density of population. They must be coated, wrapped, cathodically protected against corrosion, and buried with the proper amount of cover to avoid rupture by outside operations. Many operators are now using PVC or other plastic pipe for gas gathering lines below a pressure of 50 psig. These pipes are less expensive, require no coating, and are less expensive to install.
GAS INJECTION
Produced gas may be processed and injected back into the formation from which it is produced. Additional gas from other formations may also be injected to supplement the gas produced from a given oil zone. This is done for two reasons: (1) reservoir engineers have determined that the production of oil can be maximized (maximum recovery) by maintaining the field pressure and/or (2) the gas cannot be sold because an outlet is not available and venting is not permitted because of air pollution regulations. Injection pressures of 3000-4000 psig are common.
201 GAS ENGINEERING
The gathering and processing of natural gas is very complex, and careful evaluation of all engineering and economic factors is essential for an efficient and profitable operation (see Fig. 7-18). This is the function of a gas engineer with the assistance of specialists from other fields. The first step in the evaluation of a new project is to determine gas reserves. This is usually done by a reservoir engineer with assistance from a geologist. The results of this study must also include an estimate of deliverability versus time. Although the well test is helpful, test volumes tend to be higher than can be realized in sustained production. The well test results also include compositional data. A complete analysis is much more desirable than some quick test methods used for product allocation. Consideration must be given to probable changes in composition in oil well gases as gas/oil ratios increase. On the basis of the gas analysis and the market conditions, the process engineer will prepare a flowsheet for the processing scheme and the mechanical engineer will prepare detailed designs and mechanical equipment specifications. Preliminary cost estimates are prepared by the process and/or mechanical engineer. The gas engineer will make an economic evaluation outlining production, income, expenses and profitability. If the profitability meets the requirements of the management of a OF ENGINEERING DEPT.
FUNCTIONS
QE
NATURAL GAS AND GASOLINE DIVISION
j -1
I
I PROCESS
ENGINEERING
I
1
1 JUSTIFICATION,
I
MECHANICAL ENGINEERING
I
PROFITABILITY, PAYOUT
.
t
,
I
I
(PROJECT) PROJECT ~
I
I
HYDROCARBON BALANCE
I L
I
PROCESS
P L O T PLAN EQUIPMENT RATING
I
1 1
I
I
I
ENGINEERING
1
1 - 1
CONSTRUCTION SUPERVISION
PROCESS PROBLEMS AND INSTRUMENTATION
I
START
UP
]
Fig. 7-18. Example of functions of engineering department of the natural gas and gasoline division
202
particular company, the project will proceed. Economic criteria vary widely; among those used are payout, rate of return on capital, ratio of cash flow to investment, and the net present value at a specified discount rate. Final engineering design may be completed in-house and the engineering drawings submitted to contractors for construction bids, or the entire project may be submitted to contractors for both final design and construction. The size of the engineering group determines how much work is done in-house. In very small companies, a few engineers perform all the engineering functions and preliminary as well as final design may be made by an engineering contractor. The various facets of gas engineering are so interrelated that even a specialist, such as the process or mechanical engineer, should be as familiar as possible with all aspects of a project, including field production, so that the proper decisions can be made. The production engineer responsible for field operations should also be familiar with the problems of the gas processor in order to make the entire operation run smoothly.
SAMPLE PROBLEMS AND QUESTIONS
(1) Draw a simplified flow 'diagram of absorption, stripping, and stabilization operations. (2) Describe three general methods of concentrating the components of gases. (3) List agents which can remove hydrogen sulfide from natural gases. Also show corresponding reactions. (4) Draw a simplified flow diagram and cross section of an ethylene recovery unit. (5) For a specific period of operation of the ethylene recovery plant, the compositions of the various streams are given below in mole%: Component
Feed 29.2 2.9 1.1 1.3 58.4 1.2 5.9 100.0
Make
Discharge
Purge
0.0 13.9 86.2
23.3 2.5 1.1 1.3 71.5 0.2 0.0
42.6 3.9 1.3 1.6 49.2 0.5 0.8
100.1
99.9
99.9
0.0 0.0 0.0
0.0
For the period covered by the table, the number of moles of discharge gas was 1.447 times the number of moles of purge gas, and 94.8% of the C,H, was recovered in the make gas. For each 100.0 moles of feed gas, compute the number of moles of each component and the total number of moles in each of the other streams, entering the answers on a skeleton-flow diagram.
203
(6) For a specific period of operation of the natural gasoline plant, the compositions of the various streams are given below in mole%: Raw gas Cl c2
c3 c4
c:
Absorber residual gas
Reabsorber residual gas
Stabilizer residual gas
Finished gasoline
0.0
0.0 0.0 55.4 44.6 0.0
0.0 0.0 53.1 46.9 0.0
0.0 0.0 0.0 34.4 65.6
100.0
100.0
100.0
100.0
44.5 16.9 16.2 12.1 10.3
60.3 22.9 14.2 2.6
100.0
For each 110.0 moles of the raw gas, compute the number of moles of each component and the total number of moles in each of the other streams, entering the answers on a skeleton-flow diagram. (7) Crude oil is heated in exchangers by the counter-current flow of fuel oil. Measured at 60°F, the exchangers handle 30,500 gal/hr of the crude and 10,250 gal/hr of the fuel oil. The gravity and U.O.P. characterization factor of the crude are 35" API and 11.3, whereas those for the fuel oil are 20" API and 11.3, respectively. The fuel oil is cooled from 750 to 250"F, and the crude enters the exchangers at 85°F. Compute the heat lost by the fuel oil in Btu/hr. Estimate the outlet temperature of the crude oil, assuming all of the heat lost by the fuel oil is picked up by the crude. (8) Engineers were studying the advisability of deasphalting lube oil by contact with a mixture of propane and normal butane. A mixture containing 32.1 mole% propane, 31.2 mole% butane, and 36.7 mole S lube oil was flashed at 25 psig and 100°F. Under these conditions, the equilibrium ratios ( K= y / x ) were 3.87 for propane, 1.28 for butane, and 0.00 for the lube oil. For 100.0 moles of the original mixture, compute the number of moles and mole % of each constituent in the equilibrium liquid. For the first approximation, assume that the total number of moles of liquid is 1.75 times the total number of moles of vapor, or L / V = 1.75. (9) Outline a satisfactory sampling procedure for a nonretrograde wet gas and sketch the equipment required. What precautions should be taken? (10) A mixture of nitrogen and propane having a gross heating value of 1035 Btu/cu ft is used as a fuel in lieu of natural gas. It is transported in a buried line. If the ground temperature is 50"F, what is the maximum permissible pressure that will avoid condensation of liquid in the line? (11) (a) Define "equilibrium ratio". @) What factors affect the value of the equilibrium ratio? (c) What methods can be used to correlate the most complex of the factors listed in (b)? (12) On purchasing propane as a motor fuel, what properties should one specify? Why? Describe testing procedures.
204
APPENDIX 7.1 - SAMPLE PLANT FLOW PROBLEMS (After Chilingar and Beeson, 1968, pp. 360-365)
Ethylene recovery plant problem For a specific period of operation of the ethylene recovery plant, the compositions of the various streams are given below in mole %: Component
Feed
Make
30.3 3.1 1.0 1.5 56.7 1.3 6.1
0.0 0.0 0.0 0.0 0.0 12.1 87.9
Discharge 22.5 2.2 0.7 1.5 72.5 0.5 0.2
49.7 5.3 1.8 1.8 40.0 0.6 0.6
Purge
100.0
100.0
100.1
99.8
For the period covered by the table, the number of moles of discharge gas was 1.765 times the number of moles of purge gas, and 95.1% of the C,H, was recovered in the make gas. For each 100.0 moles of feed gas, compute the number of moles of each component and the total number of moles in each of the other streams. Enter the answers on a skeleton-flow diagram.
Description of plant used for ethylene recovery The feed gas stream, which is piped directly from the demethanizer column, is introduced about midway in the tower. It moves upward in contact with (and countercurrent to) the activated carbon bed, moving downward through the tower. At 140°F temperature, which is maintained at the feed point, the carbon selectively adsorbs the ethylene and other gases of higher molecular weights. The unadsorbed lighter gases, with predominating methane and hydrogen, pass upward and are removed as a discharge gas. This removal occurs at a disengaging plate located just below the cooling section at the top of the tower. This discharge gas is then used around the plant as fuel. In order to keep the system free of an excessive amount of 60-mesh or finer carbon particles, a portion of the unadsorbed gases is permitted to pass up the tower, and past the cooling section, as purge gas. The adsorbed gases (mostly ethylene) pass downward with the carbon, into the stripping section, where the temperature of the carbon is increased to about 510°F. At this temperature the adsorbed gases are desorbed from the carbon, pass back up the tower, and are removed as make gas at the second disengaging tray situated just above the stripping section. A small portion of these gases, however, continues on up past the disengaging tray and serves as reflux in the section of the tower below
205
the feed tray. These gases serve to liberate the lighter components which may contaminate the adsorbed material. The last traces of adsorbed gases are removed by the steam (at 150 psig), which is introduced below the stripping section. This steam passes upward and out with the make gas, and is then condensed in a cooler. In order to remove the remaining moisture from the product, the make gas is finally passed through an alumina dehydrator. As the carbon leaves the base of the tower, it is moved back up to the top through a pipe. A sealing leg at the bottom of the tower prevents the stripping steam from entering the lift section. The gas pressure for lifting the carbon is supplied by a centrifugal blower. On reaching the top of the tower the carbon is dehydrated and cooled to about 120°F in the cooling section.
Solution: Moles of C,H,
=
(95.1)( 6.1) = 5.8 moles 100
5.8
Total moles in make gas = -- 6.6 moles 0.879 Material balance:
100.0= 6.6 + P + 1.765P solving for P (moles of purge gas)
P = 33.8 moles and
D = 100 - 33.8 - 6.6 = 59.6 moles (discharge gas) Component
H2 NZ 0 2
co CH, CZH, C2H4
Feed
Make
Mole %
Mole %
30.3 3.1 1.0 1.5 56.7 1.3 6.1 100.0
Discharge Moles
Mole %
Purge Moles
Mole %
Moles
0.0 0.0 0.0 0.0 0.0 12.1 87.9
0.0 0.0 0.0 0.0 0.0 0.8 5.8
22.5 2.2 0.7 1.5 72.5 0.5 0.2
i3.4 1.3 0.4 0.9 43.2 0.3 0.1
49.7 5.3 1.8 1.8 40.0 0.6 0.6
16.9 1.8 0.6 0.6 13.5 0.2 0.2
100.0
6.6
100.1
59.6
99.8
33.8
206
Purge Discharge Make
Moles 33.8 59.6 6.6
-=59*6 - 1.765 P 33.8
100.0
5.8 Recovery of C,H, in make gas = - = 95.1% 6.1
Natural gasoline plant problem
The compositions of various streams (mole %) in a natural gasoline plant are as follows: Component
c, c2
c3 c4
c;
Raw gas
Absorber residual gas
76.0 9.5 6.0 4.5 4.0
84.0 10.5 3.9 1.7 0.0
100.0
100.1
'
Reabsorber residual gas
Stabilizer residual gas
Finished gasoline
0.0 0.0 78.6 21.4 0.0
0.0 0.0 73.7 26.3 0.0
0.0 0.0 0.0 35.5 64.5
100.0
100.0
100.0
Assume that the liquid volumes are additive and that 1 gal of gasoline can be obtained from 37.0 scf of C, vapor, 31.0 scf of C, vapor, or 26.0 scf of C: vapor. (a) Calculate the number of moles (based on 100 moles of raw gas) for each stream and each component. (b) Calculate the total number of moles per day (based on 15,000 standard Mcf per day of raw gas being processed) of residual gases from the absorber, the reabsorber, and the stabilizer. (c) Calculate the composition of the raw gasoline fed to the stabilizer. (d) Compute the number of gallons per day of finished gasoline produced (based on 15,000 standard Mcf per day of raw gas being processed). Description of natural gasoline plant The natural gasoline plant (see Figs. 7.1-1, 7.1-2, 7.1-3, and 7.1-4) is processing gas which is pumped from the casing heads of oil wells. The raw gas is compressed, cooled, and then sent upward through an absorbing column, countercurrent to a stream of cold absorption oil. The residual (or fuel) gas leaving the absorber contains almost all of the C, and C, present in the raw gas. As the gasoline-rich absorbing oil leaves the bottom of the absorber, it is pumped through a heat exchanger, and through a steam preheater, into a common stripping still and
207
FIELD FUEL
r
COMPOSITIONS VOLUME %
METHANE
WET GAS
RESIDUE GAS
LlOUlD HYDROCARBONS
FIELD FUEL
80.38
90.87
0
70.55
2.24
.02
2 3 14
100 00
100 00
ETHANE PROPANE 9 58
ISO- BUTANE
NORMAL BUTANE PENTANE
+
I69
100 00
17 4 4
100 00
Fig. 7.1-1. Example of compositions of various flow streams in a natural gasoline plant.
rectifier. The stripped absorption oil, which is devoid of all gasoline fractions, is pumped from the bottom of the still through the above mentioned heat exchanger, and through a cooler, back to the top of the absorption column.
4
-
LIGHT FRACTIONS
!I
ABSORPTION AREA
RIC
DlSTILLATION AREA
FRACTIONATION AREA
t
TO STORAGE
UTILITIES STEAM WATER E L E C T R I C POWER COMPRESSED A I R
SPEC I F IC A T ION N A T U R A L
HYDROCARBONS
Fig. 7.1-2. Schematic flow diagram showing the various areas of gasoline plant.
208
,
r----*
-
Fig. 7.1-3. Flow diagram of absorption area
The raw gasoline vapors from the top of the still are partially condensed and collected in a small rundown tank. Inasmuch as the uncondensed vapors contain sufficient amounts of pentane and heavier hydrocarbons, they are sent up through a small reabsorber in order to relieve the main absorber of a heavy recycle load. The reabsorber is fed at the top with gasoline-free absorption oil; and the rich oil from the bottom is added to the rich oil from the absorber. The reabsorber does not absorb propane, but absorbs some of the butane and all of the pentane. The unabsorbed gas from the reabsorber is a residual or fuel gas. The condensed raw gasoline from the still is pumped through the second heat exchanger, and a steam preheater, to a stabilizer. The residual or fuel gas taken overhead contains all of the propane, part of the butane, but none of the pentane.
Gosolip Vapors and Stwm
t
Hot Lton Oil Cold Leon Oil
COOLER
Fig. 7.1-4. Flow diagram of distillation area.
c
I
ACCUMULATOR
209
The finished, propane-free gasoline is pumped from the bottom of the stabilizer through the second heat exchanger and a cooler to storage. Solution: Absorber residual gas:
C,
= 84.0%= 76.0 moles
loo x 76.0 = 90.5 moles Total moles = 84.0
3.9 x 76.0 = 3.5 moles, etc. c, = 84.0 Finished gasoline:
C: = 64.5%= 4.0 moles loo x 4.0 = 6.2 moles Total moles = 64.5
35.5 x 4.0 = 2.2 moles c, = 64.5 Reabsorber residual gas and stabilizer residual gas: If R and S are the total number of moles in these two streams, respectively, then
0.786R+0.737S=2.5; 0.214 R + 0 . 2 6 3 S = ( 3 . 0 - 2 . 2 ) = 0 . 8
On solving these two equations simultaneously, S = 1.9 moles and R
= 1.4 moles
(a) The number of moles of C , , C , , C , , C,, and : C in: (1) absorber residual gas are 76.0, 9.5, 3.5, 1.5, and 0.0, respectively; (2) reabsorber residual gas are 0.0, 0.0, 1.1,0.3, and 0.0, respectively; (3) stabilizer residual gas are 0.0, 0.0, 1.4,0.5, and 0.0, respectively; and (4) finished gasoline are 0.0, 0.0, 0.0, 2.2, and 4.0, respectively. (b) Moles/day raw gas =
15,000,000 = 39,600 379
Absorber residual gas = 0.905 X 39,600 = 35,800 moles/day Reabsorber residual gas = 0.014 X 39,600 = 550 moies/day Stabilizer residual gas = 0.019 x 39,600 = 750 moles/day Note: At 60'F and a pressure of 760 mm Hg,the volume of 1 Ib-mole of a perfect gas is 379.4 cu ft (379 is commonly used).
210
(c) The number of moles of C , , C,, C,, C,, and C: in raw gasoline fed to the stabilizer are 0.0, 0.0, 1.4,2.7, and 4.0, respectively. Thus, Feed to stabilizer = 0.081 X 39,600 = 3210 moles/day. (d) Finished gasoline: C , = 0.022 x 15,000,000/31 = 10,600 gal/day C: = 0.040 X 15,000,000/26 = 23,100 gal/day Total
33,700 gal/day
REFERENCES American Gas Association, 1965. Gas Engineers Handbook. Industrial Press, New York, N.Y., pp. 2-13. Chilingar, G.V. and Beeson, C.M., 1968. Surface Operations in Petroleum Production. Am. Elsevier, New York, N.Y., 397 pp. Gas Processors Association, 1980a. LPG Specifications and Test Methods. Publication 2140. G.P.A., Tulsa, Okla. 42 pp. Gas Processors Association, 1980b. Calculation of Gross Heating Value, Specific Gravity and Compressibility of Natural Gas Mixturesfrom Compositional Analysis. Publication 2172. G.P.A., Tulsa, Okla., 6 pp. Gas Processors Association, 1980c. Method for Determination of Hydrogen Sulfide and Mercaptan Sulfur in Natural Gas. Publication 2265. G.P.A., Tulsa, Okla. Gas Processors Association, 1980d. Natpral Gasoline Specifications and Test Methods. Publication 3132. G.P.A., Tulsa, Okla., 33 pp. Katz, D.L., Cornell, D., Kobayashi, R., Poettmann, F.H., Vary,J.A., Elenbaas, J.R. and Weinaug, C.F., 1959. Handbook of Natural Gas Engineering. McGraw-Hill, New York, N.Y., 802 pp. Kennedy, J.L., 1972. Ekofisk plans include gas injection at 9200 psi. Oil Gas J., 70(9): 69-73. Maddox, R.N., 1974. Gas and Liquid Sweetening (Petroleum Series). Campbell, Norman, Okla., 300 pp. Momson, W.E., 1964. Summary of Energy Balance for the U.S. I.C. Bur. Mines Inf. Circ. 8242. Natural Gas Processors Suppliers Association, 1962. Liquefied Petroleum Gas Specifications and Test Methods. N.G.P.S.A., Tulsa, Okla., 36 pp. Natural Gas Processors Suppliers Association, 1966. Engineering Data Book. N.G.P.S.A., Tulsa, Okla., 310 pp. Natural Gas Processors Suppliers Association, 1972. Engineering Data Book. N.G.P.S.A., Tulsa, Okla., 425 pp. (Latest revision). Pacific Energy Association, 1943. Determination of Hydrogen Sulfide in Natural Gas. Bull. TS-431. P.E.A., Vernon, Calif. Pacific Energy Association, 1950. Determination of Carbon Dioxide and Oxygen in Natural Gases. Bull. TS-501. P.E.A., Vernon, Calif. Pacific Energy Association, 1966. Methods of Test for Liquefied Petroleum Gas. Bull. TS-441. P.E.A., Vernon, Calif., 81 pp. Western Gas Processors and Oil Refiners Association, 1950a. Method for Determining the Specific Graviv of Gases. Bull. TS-391. W.G.P & O.R.A., Long Beach, Calif. Western Gas Processors and Oil Refiners Association (formerly California Natural Gasoline Association), 1950b. Tentative Standard Procedure for the Determination of Carbon Dioxide and Oxygen in Natural Gases. Bull. TS-501. W.G.P. & O.R.A., Los Angeles, Calif., 16 pp. Western Gas Processors and Oil Refiners Association, 1956. Tentative Standard Procedure for the Measurement ofNatural Gas with Orifice Meters. Bull. TS-561. W.G.P. & O.R.A., Los Angeles, Calif., 89 PP. Western Gas Processors and Oil Refiners Association, 1965. Tentative Standards of Test for Liquefied Petroleum Gas. Bull. TS-441. W.G.P. & O.R.A., Los Angeles, Calif., 81 pp. Western Gas Processors and Oil Refiners Association, 1955. Tentative Standard Method of Test for the Reid Vapor Pressure of Natural Gasoline. Bull. TS-352. W.G.P. & O.R.A. Los Angeles, Calif., 22 pp. Wilson, H.M., 1974. Three flow stations will handle Prudhoe East Production. Oil Gas J., 72(11): 57-72.
211
Chapter 8
OIL AND GAS TRANSPORT SANJAY KUMAR and GEORGE V. CHILINGARIAN
MTRODUCTION
Inasmuch as it is usually much cheaper to use pipelines than to use barges, tankers, trucks, etc., complex transcontinental piping systems have been developed. Some of the major advantages of using pipelines include: (1) Economy. (2) Reliability under almost all conditions (e.g., adverse weather, breakdowns, non-availability of tankers, etc.). (3) Control: An installed pipeline can usually handle a wide range of flow rates. (4) Continuity of flow, which is highly desirable in modern continuous-flow oil refineries, because a minimum of storage facilities are required at either end. The designs of simple systems are briefly discussed here. The fundamental concepts involved are very useful and their knowledge is absolutely essential for any practicing petroleum engineer.
FUNDAMENTALS OF FLOW IN PIPES
Reynolds applied dimensional analysis to the phenomenon of the transition from laminar to turbulent flow. He concluded that it occurred at a fixed value of a dimensionless group, which is called the Reynolds number ( N R e )in his honor: NRe = dVp/p
= d V / v = inertia
forces/viscous forces
(8-1)
where d = diameter of the conduit through which the fluid is flowing, V = velocity V = O AT PIPE WALL
I
LAMINAR FLOW
TURBULENT FLOW
Fig. 8-1.Velocity profiles in a circular conduit for laminar flow (left) and turbulent flow (right).
212 2
of the fluid, p = density of the fluid,’ p = dynamic viscosity of fluid, and Y = kinematic viscosity of fluid. For laminar or viscous flow, N,, < 2100; for transitional or intermediate flow, 2100 < N R e < 10,000 3 , and for turbulent flow, N,, > 10,000.3 For cases where the flow is not through a cylindrical pipe (the cross-section is other than circular), the concept of effective diameter has been developed: d, =4R,
(8-2)
where hydraulic radius, R , = (area of flow)/(wetted perimeter). If the cross-section of a pipe is square, for example, cross-sectional area of flow = a’, wetted perimeter = 4 a , and R ,
=
d 4
a2 4a
= -=
a 4
-. Thus: d , = 4 R , = a .
In the case of circular cross section, area of flow, A = r d 2 / 4 , wetted perimeter R , = ( m d 2 / 4 ) / ( r d )= d / 4 , and d , = 4 R , = d . The schematic diagrams of velocity profiles for laminar and turbulent flows are shown in Fig. 8-1. =rd,
Allowable working pressure of pipeline The allowable working pressure of a pipeline, p a , is equal to: St’Fe ( d o - d , ) Pa =
do
(8-3)
where S = specified minimum yield strength, psi; t’ = temperature derating factor, dimensionless; F = design factor (construction type); e = longitudinal joint factor (i.e., anomaly due to weld seam); it is equal to 1.0, except for butt-welded ASTM-A53, API-5L ( = 0.6), fusion-welded A 134 and A 139 ( = 0.8), and spiralwelded A 211 (= 0.8); d , = inside diameter of pipe, in.; d o = outside diameter of pipe, in. The specific weight, y , is the weight per unit volume and can be expressed in terms of lb/cu ft. The density, p’, which is mass per unit volume and is equal to y / g , where g is the gravitational acceleration ( = 32.174 ft-sec-’), can be expressed in terms of slugs/cu ft. For example, pure water, which has a specific weight of 62.4 lb/cu ft, has a density, p ’ , of 1.94 ( = 62.4/32.17) slugs/cu ft. In other words, the mass is attracted by the earth with a force of magnitude p’g. Inasmuch as 1 slug = 32.17 lb,, a density of 1 slug/ft3 = 32.17 lb,/ft3. The density expressed in lb,/ft3 will be designated by p throughout this book. Inasmuch as p = shear stress/shear strain = T/(dV/dy), the units of dynamic viscosity are Ibfsec/ft2 or slugs/ft-sec. 1 poise = 100 centipoises = 1dyne-sec/cm2 = 1 g/cm-sec = 0.00209 slug/ft-sec = 0.00209 lb,sec/ft’ = 0.0673 Ib,/ft-sec. 1 slug/ft-sec = 479 poises. 1 stoke -100 centistokes -1 cm2/sec = 0.001076 ft‘/sec. Value of 3100-4000 is commonly assigned to this limit.
213
Horsepower
The hydraulic horsepower, P,, is equal to 0.000017 X bbl/D gal/min X psig.
X
psig = 0.00053 X
Friction
Obviously, friction is associated with any kind of flow. Pipes are defined as smooth, if the relative roughness k, (= r/di) is < ( c = absolute surface roughness in inches and di = inside diameter of pipe in ft). The Fanning equation for steady state flow in uniform circular pipes which are full of liquid under isothermal conditions is as follows:
where A p , = pressure loss due to friction, lb/ft2; f = Fanning. friction factor, dimensionless; I = length of pipe, ft; V = fluid velocity, ft/sec; y = fluid specific weight, lb/ft3; p = density (mass/unit volume), lb,/ft3; g = gravitational acceleration, 32.2 ft/sec2; g, = dimensional constant, 32.2 lb,-ft/lb-sec2; and d = diameter of pipe, ft. If A p , is expressed in psi, the eq. 8-4 becomes:
The Fanning friction factor, f , is equal to: 16
f = - for laminar flow, ' NRe
f = 0.1419( N R e ) -0.3192 for intermediate flow, f = 4 log- NReJT for partially turbulent flow, and 1.4126
According to Hagen and Poiseuille, at N R e ranging from 2000 to 2300, f = 64/NR,. Thus, extreme care must be exercised on using values of f available in the literature, because they may be multiples of the f given in Fig. 8-2. The equation f = 64/( V d p / p ) is widely accepted.
214
1 = 4 log- NReJT -
JT
1.255
for fully turbulent flow (Prandtl-Khrmhn formula) For rough pipes, using the absolute roughness, E in microinches, the above relationships are slightly modified: 1
2.51
(8-9)
for intermediate flow (Colebrook's relation), 1
- = 4 log-
sr
3.71di NRe
f=f, + 0.68NRef?
for turbulent flow (Prandtl-Khrmhn equation),
i i i )
(Supino formula),
(8-10)
(8-11)
where f, = friction factor for smooth pipes. These equations present fairly good approximations over a wide range of NRe. The most widely used correlation between NRe and f (Moody, 1944) is shown in Fig. 8-2. For an excellent discussion of fundamentals of flow in pipes, the reader is referred to Szilas (1975) and Craft et al. (1962, pp. 1-100). Friction head loss in fittings and connections The Crane Company, which conducted exhaustive tests to find the resistance of valves and fittings to single-phase flow, classified all valves and fittings as follows: ( 1 ) Branching- tees, crosses, side-outlet elbows, etc. (2) Reducing or expanding-swages, reducers, chokes, bushings, etc. ( 3 ) Deflecting-elbows, bends, return bends, etc. Crane Company also included an equivalent length concept, i.e., expressing head loss due to the friction in valves and fittings in terms of equivalent head loss due to the friction in a straight pipe. In 1911, Blasius formulated the following empirical equation for turbulent flow:
f = 0.316/N;<'. This equation is valid for smooth pipes at Reynolds numbers up to about lo5. Formula 8-8 is commonly called Prandtl's formula and is expressed as follows:
215
Fig. 8-2. Friction factors for various types of commercial pipes. (From Moody. 1944 p. 671; courtesy of the Am. Soc. Mech. Eng.) NRr= V D p / p . where V = velocity in ft/sec, D = pipe diameter in ft, p = fluid density, Ib/cu ft. and p = viscosity of fluid in Ib/ft-sec (cP/1488).
216
I'(
- ")*
FEET OF FLUID
SEE ALSO EQUATION (5) IF A, = 00 SO THAT V, = 0
h = - 'I'
FEET OF FLUID
V' h = K - FEET OF FLUID 29 Fig. 8-3. Resistance coefficients for valves and fittings (reprinted from Engineering Data Book, 1979, p 75, table 32 (a); courtesy of the Hydraulic Institute, Cleveland, OH).
I FI
217
REGULAR SCREWED 45O ELL.
-
BELL MOUTH INLET OR REDUCER K=0.05
K
SQUARE EDGED INLET K = 0.5
INWARD PROJECTING PIPE K = 1.0
I
NOTE: K DECREASES WITH INCREASING WALL THICKNESS OF PIPE AND ROUNDING OF EDGES
LINE FLOW 1 3'.
.5
I
n
2
A
SCREWED TEE BRANCH FLOW
h=K Fig. 8-3 continued.
V' FEET OF FLUID 2a
21 8
0.wm 0.w15 0.w10 D O.OW5 SMOOTH
-
0
1
2
3
4
s
(I
7
8
8
10
R D
Fig. 8-4. Resistance coefficients for 90" bends of uniform diameter. (From Engineering Data Book, 1979, p. 79; courtesy of the Hydraulic Institute; Cleveland, OH).
The decrease in static head due to velocity is expressed as: V2
and if there is a valve or fitting in the line, then the head loss due to the friction h , is
t
06
04
K
03
02
01
00
10
15
20
25
30
35
40
Fig. 8-5. Resistance coefficient for sudden reducers. (From Engineering Data Book, 1979, p. 82; courtesy of the Hydraulic Institute. Cleveland, OH).
219
equal to: V2 hl=K2g
(8-12)
where K = resistance coefficient, defined as the number of velocity heads lost due to the valve or fitting. Inasmuch as the same head loss can be also expressed by the Fanning equation: h
1 v2 d 2g
=f--
1 K=fd
(8-13)
The ratio l / d is the equivalent length (in pipe diameters) of straight pipe, which will cause the same pressure drop as the fitting under the same flow conditions. Thus, the equivalent length (in ft) of a straight pipe, I,, is equal to: (8-14)
where ( I / d ) = equivalent l / d for a given fitting, and d = diameter of the pipe in which the fi&g is installed. The K values for various valves and pipe fittings are presented in Figs. 8-3, 8-4, 8-5, 8-6. PRINCIPLES OF PUMPING
Pumping mechanisms
The six basic mechanisms of artificially-induced fluid flow are (Perry and Chilton, 1973): (1) the action of centrifugal force, (2) volumetric displacement (mechanical or by another fluid), (3) mechanical impulse, (4)transfer of momentum by another fluid, (5) electromagnetic force, and (6) gravity. The mechanisms (1) and (2) are commonly used in the petroleum industry. Measurement of performance
The amount of useful work performed by a fluid-transport device is the product of the rate (capacity) at which fluid is transmitted and the head (the height of a column of fluid equivalent to the pressure differential between inlet and outlet ends of the device). Capacity can be expressed in ft3/min, whereas head can be expressed in ft.
220
GL
h
NDTF:
GATE VALVE '14 CLOSED '/2
'
CLOSED
-FULLY
OPEN
-3,000
- 2,000 --~1,000 -
50
7 500
ANGLE VALVE, OPEN
I
STANDARD TEE
-
- 300 - 200
30
30 22
/SQUARE
SWING CHECK VALVE, FULLY OPEN
F++ ,BORDA
20
I-
=loo
!
16 14
ENTRANCE
2
--- 0.5 - 0.3 - 0.2 -I
LONG SWEEP ELBOW OR' RUN OF STANDARD TEE
-
I
0.1 0.5
Fig. 8-6. Resistance of valves and fittings to flow of fluids. (Reprinted from Technical Paper No. 409, 1942; courtesy of Crane Co., New York).
221
The overall efficiency of the pumping system is defined as the ratio of useful hydraulic work performed to the input work to the device. Hydraulic horsepower, P h =
8.33HGQ --H p 33000 1714
(8-15)
where H = total dynamic head, ft of liquid; H p = total dynamic head, psi; Q = volumetric rate of flow, gal/min; and G = specific gravity with respect to water ( G , = 1). The brake horsepower of a pump, P b , is greater than the theoretical or hydraulic horsepower by the amount of losses in the pump due to friction, leakage, etc. The efficiency of the pump, vp, can, therefore, be defined as:
v,
'h
(8-16)
='b
The total dynamic head H is equal to the total discharge head h , minus the total suction head h , . The total suction head, h,, is equal to: h , = h,,
+ ha + h,,
(8-17)
where h,, = potential energy head at suction end, i.e., pressure head recorded on a pressure gauge at the suction end in ft of liquid; h,, = velocity head, ft of liquid, at the point of gauge attachment; and h a = atmospheric pressure head, ft of liquid. Total discharge head, h,, is equal to: h,
= h,,
+ h a + h,,
(8-18)
where h,, = pressure gauge reading at discharge end, ft of liquid, and h,, = velocity head at discharge end, f t of liquid. Velocity head ( V 2 / 2 g )is equal to the vertical distance necessary for a body to fall to acquire the velocity V. There are limitations on the net positive suction head of the pump ( N P S H ) . The maximum theoretical vertical distance between the pump and the level of suction exposed to atmosphere is given by the following formula: N P S H = h,, - h , - h ,
(8-19)
where h,, = static suction head, which is equal to the absolute pressure at the source (ft of liquid) plus the vertical distance from this level to the pump center line; h,, = friction head; and h , = head equivalent to the vapor pressure of the liquid. OIL PIPELINE TRANSPORTATION
The flow of crude oil in a pipeline may be assumed to be isothermal if the oil viscosity is low and the inflow temperature is close to the surrounding soil temperature.
222
The design of a pipeline for hilly terrains has been discussed in detail by Beggs and Brill (1972) and several others (see Chapter 9). Assuming a Newtonian liquid, the principles are presented here. As mentioned previously, the friction loss over a length of uniform pipe is given by eq. 8-4:
This equation can be presented as follows: (8-20)
where: (8-21)
and h h is the hydraulic gradient. Thus:
I
h , , =ALP = h Y
h
(8-22)
where h,, = the total suction head at any point x along the length of the pipe and y = specific weight of fluid. One can then use the criterion of hydraulic gradient to determine the possible flow over a given terrain, as shown in Fig. 8-7. Gradient h , is calculated using eq. 8-21, and a line with a slope equal to this gradient is drawn from the outlet end 0 to the inlet end. If this line A’ intersects the ground profile, there are two options: (1) Increase the inlet head to h , . In this case, A’ is shifted upwards until it just touches the profile at M. Because of safety considerations (safety factor), h , must be higher than this minimum by 30-50 m. (2) Increase the head h , to hma. This results in a line B having a different slope. The point M is called the critical point, where the pressure head of the flowing liquid is the lowest. If no throttling is applied at the tail end 0, the oil arrives at atmospheric pressure, whereas if throttling is applied, the pressure head at the outlet is h , . The point M‘ in the valley is also critical, because one must calculate the pressure head h and limit it to the maximum allowable operating pressure of the pipeline ( P= b ’ g ) . The maximum flow rate in the pipeline can be determined as follows: (1) Calculate the maximum head, h,, corresponding to the maximum allowable operating pressure of the pipeline.
223
-
7
i
1
hl
09 Distance, f ?
-
f'
Fig. 8-7. Pressure profile for a pipeline laid over a hilly terrain. (Modified after Szilas, 1975, p. 479, fig. 7.1-1; courtesy of Akademiai Kiad6, Budapest, Hungary).
( 2 ) Determine the constants a and b in the equation f = a N i : for the flow regime anticipated and the pipe in question. These are usually given in the pipe manuals. ( 3 ) Substitute the above equation for f in eq. 8-21:
Inasmuch as V = -and g
= 32.2
ft/sec-',
d277/4
(8-24)
or
(8-25) This equation can be used to determine the maximum Q for a maximum h,, if the only pump station is at the head end.
224
Example 8-1
Determine the maximum throughput of a 30-in. API 5LX grade X46 pipe (maximum allowable stress, Sa(,,) = 21,000 psi), if the safety factor, F, = 1.2, relative roughness c / d = l o p 6 (smooth pipe), d o = 30 in., d i = 28 in., Go,= 0.84, poi,= 4 cP, pipe length = 40 miles, and outlet is located 400 ft above inlet. Solution:
S,
= Sam,/< = 21,000/1.2 = 17,500 psi.
Maximum allowable working pressure is given by eq. 8-3:
For additional safety, t h s h,, = 3205
h,,
(0.95) = 3045 ft
Then, hh(,=) p,
=4
= (3045 - 400)/(40)(5280) = 0.012524 ft/ft
CP= 4 x 6.72 X
p .. . v = o = O
po
is, reduced by about 5%:
lb,/ft-sec
(4)(2.09)(10-5) (0.84)(62.4)/32.2
=4
= 5.2 x
x 2.09 x l o T 5lb,sec/ft2,
l o p 5 ft2/sec.
Using a = 0.05 and b = 0.19 in eq. 8-24, Q can be calculated:
= 50.225
ft3/sec
= 32,200 bbl/hr.
Increasing flow capacity of pipelines
There are two ways of increasing the flow capacity of an existing pipeline, which should be achteved with a least-cost objective in mind: (1) By installing one or more intermediate pumping stations along the pipeline. (2) By installing a second pipeline alongside the already existing pipeline. This so-called looped system is of two kinds:
225
(a) Complete loop-installation of a new line of the same length, but not necessarily of the same diameter. (b) Partial loop-installation of a new line shorter than the old line but of the same inside diameter. This new line is preferably started at the tail end, in order to achieve a lower average pressure in the entire system. For short sections of the Moody friction factor chart, one can express f as an exponential function of NRe: f = aNi:
where a and b are constants. Assuming a = 0.05 and b = 0.19,
f = 0.05Ni:.19
(8-26)
From eq. 8-21: 2 f V 2 = 2(0.05)N;f,0.'9V2 hh= gdi 32.2di - 0.05V2 diV --
16.1di
(v)
~ 1 . 8 v0.19 1
= 0.0031
d;.l9
(8-27)
Inasmuch as V = Q / ( n / 4 ) d ? , in terms of Q, h , is equal to:
( :*;:::)
h h = 0 . 0 0 3 9 5 ~ ~"~
(8-28)
Thus: Ql.81
=
252.95hhd,!.81 v0.19
or (8-29)
and (8-30)
226
These equations for h,, Q , and di can be used to calculate the parameters of the new pipeline required. ( I ) Complete loop Complete loop is regarded as two independent pipelines. In order to calculate the diameter of the new pipeline, the following procedure must be followed: If Qlmax=maximum flow rate in old pipeline (line l), Q, = total flow rate desired, and Q, = flow rate in new Line 2 only, then Q , = Q, It is assumed that hhmaxis the same for both pipelines 1 and 2. Equation 8-30 can be used to get the diameter of the new line d , (Q, = Q , , and
elmax.
h h = hhmax):
di, = 0.3165( ~ ~ . ~ ~ ~ ~ Q 2 0 . ~ ~ ~ ) / h 2 ~
(2) Partial loop In the case of partial loop for delivering at a rate Q, > line would be h , > hhmax.Using eq. 8-28:
hhl
(8-31)
elmax, the gradient in each
= 0.00395(
Inasmuch as the two pipelines have equal diameter and the same inlet pressure, each will carry oil at a rate of Q,/2. Thus: h,,
= 0.00395(
Q,/2)1'81/d4.81
Dividing hhl by h,, and solving for h,: h,,
= h h l ( i ) 1 . 8= 1 0.285hh1
(8-32)
After determining the two gradients, the length of the new pipeline can be either calculated or determined graphically. Graphical method. If the new pipeline is started at the head end Z, which is done when pressure is to be maintained in the line due to hilly terrain, trace a gradient h,, from the head end I and a gradient h,, from the tail end 0 as shown in Fig. 8-8. The intersection point K enables determination of the length of the new pipeline. If the new pipeline is to be started from the tail end, which is preferable, then the two gradients are reversed, as shown by the dashed lines meeting at point K'. Calculation method. Refemng to Fig. 8-8: h = h h m a x l = h h l ( l - l x+) h h , / ,
227
Dis?ance(l), miles W -
Fig. 8-8. Pressure profile for a twin looped pipeline over a hilly terrain. (Modified after Szilas, 1975, p. 482, fig. 7.1-2; courtesy of Akademiai Kiad6, Budapest, Hungary),
Thus:
(8-33) Example 8-2. Propose a feasible design for increasing the capacity to 75 ft3/sec, using the data in Example 8-1.
Solution: (1) For a complete loop: Q , = 75 - 50.87 = 24.13 ft3/sec.
Using eq. 8-31:
gz
d , , = (0.3165v0.0395Q,0~376)/h
- (0.3165) [ (5.128)(10 - )] 0.0395( 24.13)0.376 ( 0.012784)0'208 = 1.756 ft = 21
inches.
The closest standard pipe size is Schedule 80, 24 in. OD, 21.562 in. ID, with a wall thickness of 1.219 in.
228
(2) For a partial loop system: Using eq. 8-28: h hl
= (0.00395v0.19Q:.81)/d4.81 = ((0.00395) = 0.02543
[ (5 .128)(10-5)]0.19(75)'.8')/(28/12)4.8'
ft/ft.
Using eq. 8-32: h h , = 0.285hh, = (0.285)(0.02543) = 0.007248 ft/ft.
The length of the new pipeline in miles can be determined using eq. 8-33:
[x=
]
0.02543 - 0.012784 (40) - 0.007248
[ 0.02543
= 27.82
miles.
Booster pump stations In designing booster pump stations, a simple graphical method can be used to determine the minimum number of stations required. Assuming that the oil is injected into the pipeline at pump stations at the maximum allowable pressure of the pipeline, input pressure is the same at all stations, except (usually) the last one. In designing for a minimum number of pumping stations, the pressure gradient in the pipeline must reach the ground zero level at the next pumping station along the line. As shown in Fig. 8-9, starting from h,, at I , the gradient hh is plotted until it intersects the ground profile. At this point, the second pumping station (the first booster pump station) must be installed. Again the pressure at point B reaches h,, and the same procedure is repeated for all the intersections with the ground profile, until the outlet end 0 is reached. In Fig. 8-9 the last booster station will deliver oil beyond point 0 (see Szilas, 1975, pp. 484-485). Branching pipelines Often, a single pipeline (called the main line or trunk line) is employed to transport oil over large distances to a central location, from where it branches out to various processing units. In order to determine the unthrottled throughputs of the branch lines for any flow rate Q in the main line, both calculation and graphical techniques can be employed (Szilas, 1975). It is much easier, however, to use a simple log-log plot. The procedure is outlined below. Consider a branched pipeline (Fig. 8-10.a) with flow rates Q , = 70 ft3/sec, Q2 = 100 ft3/sec and Q3 = 150 ft3/sec. Then, using the h versus Q relationship for these lines, draw the three h versus Q lines for the three branching pipelines. These
229
I
Distonce(l), miles
-
0 0'
Fig. 8-9. Pressure traverse with booster stations. (Modified after Szilas, 1975, p. 485, fig. 7.1-3; courtesy of Akadtmiai Kiadb, Budapest, Hungary).
will all be parallel and are equivalent to a single line with a throughput.of Q,, which is the sum of Q,, Q2 and Q3 (see Fig. 8-10.b). From eq. 8-24:
3
O3
'I
@
10 100 FLOW RATE, ~ ( f t J / r e c )
100
G
Fig. 8-10. (a) Branching pipelines. (b) Flow rate versus head relation for branching pipelines. (Modifie after Szilas, 1975, p. 487, fig. 7.1-5; courtesy of AkadCmiai Kiadb, Budapest, Hungary).
230
or h,
=KQ2-b
(8-34)
where K=
2avb 32.245 - b (
is constant.
;)2-
For a horizontal terrain, head loss h , is equal to discharge head h at the junction B:
or log h = log K
+ ( 2 - b ) log Q .
(8-35)
Equation 8-35 indicates that on a log-log paper, Q versus h relation is a straight line. Inasmuch as ( 2 - b ) is a constant, the Q versus h relations for the branching pipelines are parallel to each other, but have different K values. The Q versus h relations for different pipelines are plotted on the log-log paper. Adding the flow rates Q,, Q 2 ,. . . Q , of these pipelines yields a parallel line of the same slope. This line is called the equivalent line and gives the sum of the throughputs of the branch pipelines for any given head h. Thus, for any given total flow rate Q , , one can get the throughputs Q,, Q 2 , ... Q , for the branch pipelines. This is illustrated in Fig. 8-10.
NONISOTHERMAL FLOW
Pipelines transporting oil are commonly buried in the ground. If the oil viscosity is high and the temperature of the flowing oil is significantly different from that of the surrounding medium, the flow cannot be regarded as isothermal. Fundamentals of heat transfer
Heat transmission can occur in three different ways: (1) conduction; (2) convection: (a) free and (b) forced; and (3) radiation. When two bodies at different temperatures are in contact, conduction is the dominant mechanism, whereas if the two bodies are not in contact but are separated by a fluid body, the dominant mechanism is convection. In the latter case, the temperature is transferred by the fluid in motion. If the two bodies are separated by a vacuum, the mechanism of heat transport is radiation. Only conduction and convection, however, are discussed in this chapter.
231
Conduction Fourier's law is the fundamental differential equation for heat transfer by conduction:
dT _ dq - - kAdt
(8-36)
dx
where dq/dt is the rate of flow of heat (quantity per unit time), Btu/hr; A is the cross-sectional area across which heat flows, ft2; dT/dx is the rate of change of temperature with distance in the direction of flow of heat (the temperature gradient), "F/ft; and k is the constant called thermal conductivity, which is a characteristic property of the material through which the heat is flowing and varies with temperature, Btu/ft-hr-OF. Equation 8-36 can be used to derive the following unsteady state energy equation for static fluids or solids in 3 dimensions: cp(
g) =
&(kg)
+ $(k
g )
+
A(
) :k
+ q'
(8-37)
where x, y , z are the distances along the x, y , z axes in Cartesian coordinates; c = specific heat, Btu/lb,"F; p = fluid density, lb,/ft3; and q' = rate of heat generation by chemical or nuclear reaction, electric current, etc. In order to conform more closely to the physical shape of the system, this equation can be transformed into cylindrical or spherical coordinates. In vector notation, it may be written as follows: (8-38) Thermal conductivity, k, is approximately constant over a small range of temperature. Thus, eq. 8-38 can be simplified: CP-
aT at
+ 4'
= kV2T
(8-39)
Steady-state conduction. For steady-state heat flow, dq/dt is constant and aT/at = 0. Thus, eq. 8-39 becomes:
V2T= - q ' / k
(8-40)
and eq. 8-36 may be written as follows: dq/dt= -kA(dT/dx)
(8-41)
where dq/dt = q / t = constant. Unsteady stute conduction. In the case of unsteady state conduction, temperature is a function of both time and space. Equation 8-37 gives a general 3-dimensional equation for such a situation. There is a numerical solution to this equation, necessitating the use of digital computers. A variety of simplifications are incorpo-
232
rated to arrive at analytic solutions. One such case is the assumption of constant physical properties c, p, and k , resulting in eq. 8-42: aT 4' at = a( V2T) + -
(8-42)
CP
where
(Y
=k/pc,
called thermal diffusivity.
Convection Convection is an important factor in many cases of heat transfer involving liquids and gases. When a fluid flows past a solid surface, in the immediate neighborhood of the surface there is a film of fluid that does not contribute (or very little) to the total flow. This film clings to the surface as shown by the velocity distribution plot in Fig. 8-1. The film is in the laminar portion of the flow (the laminar sublayer), through which heat is transferred by molecular conduction. The turbulent core and the buffer layer between the laminar sublayer and the turbulent core transfer heat by simultaneous conduction and convection. If physical properties are assumed to be constant, an energy balance equation on a flowing fluid through which heat is being transferred is as follows:
(8-43)
5,
where V,, V, are the velocity components in the x, y , and z directions and @ is the energy dissipation due to the fluid viscosity, which is significant in the case of flow of highly-viscous liquids and high-speed gas. The local heat transfer coefficients were devised because of the impracticability of measuring thicknesses of the several layers and their temperatures in a turbulent flowing stream: dq = hidAi(Ti - T,) -
(8-44) = h , dAo(T, - To) dt where h i and h , are the local heat transfer coefficients inside and outside the wall, respectively (Btu/ft2-hr-"F), dA, = area element on the inside pipe wall; dAo = area element on the outside pipe wall across which heat transfer is taking place, ft'; T, = wall temperature, OF; To = outside temperature, O F ; and T, = inside temperature, OF. Natural convection Natural convection occurs when a solid surface is in contact with a stagnant fluid having a temperature different from that of the surface. The general equation used is the Nusselt equation:
(8-45) where NNu= Nusselt number, hI/k, (dimensionless); NGr = Grashof number,
233
I3p2gfiAT / p 2 (dimensionless); NPr= Prandtl number, c p / k (dimensionless); fi = volumetric coefficient of thermal expansion, ( O F ) - ’ ; p = fluid density, lb,/ft3; A T = bulk temperature gradient, O F ; 1 = length in contact, ft; g = acceleration due to gravity, 4.18 x lo8 ft/hr2, and a and m are constants. For cylinders at 1< NPr< 40, Kato et al. (1968) presented the following relations: NNu= 0 . 1 3 8 N p (N:;17’ - 0.55)
(8-46)
where NGr> lo9, and NNu= 0.683N2~sN:;2s
0.861 + N,,
(8-47)
where NGrG lo9 Forced convection In the case of forced convection, the fluid is pumped across the solid surface and the rate of heat transfer is a function of the physical properties of-the fluids, the flow rates, and the geometry of the system. For laminar flow in circular pipes, there are several relationships depending upon the Graetz number, NGz= NReNp,d/f.For Ncz < 100, Hansen (1960) presented the following relation:
( NNu)= 3.66 +
(
0.085Ncz h)’.l4 1 + 0.047NZ3 P w
(8-48)
where d = pipe diameter, 1 = pipe length, p b = viscosity of fluid at the bulk fluid temperature, and p w = viscosity of the fluid at the pipe wall temperature. For NGz> 100, the Sieder-Tate relation is satisfactory: 0.14
(8-49)
NNu= 1.86Ng3
For the transition region, the following equation was presented by Hansen (1960): (8-50) For turbulent flow ( N , , > 10,000) and 0.7 < NPr< 700, and f/d > 60, the Sieder-Tate equation is recommended: 0.14
NNu=
(8-51)
234
Application of heat transfer concepts to buried pipelines Thermal properties of the soil must be evaluated when designing buried pipelines. Heat transfer through the soil is proportional to its thermal conductivity, k, under steady state conditions, whereas under transient conditions, the thermal diffusivity a = k/cp is the controlling factor. It is, therefore, necessary to measure these soil constants, k, p, and c, which can be done in the following two ways: ( I ) Estimation of thermal constants from soil properties Thermal conductivity of the soil depends on the conductivity of the soil matrix, grain-size distribution, bulk density of the dry soil, and humidity. Gemant’s pioneering work was followed up by the studies of Makowski and Mochlinski (1956) who derived an empirical expression for thermal conductivity of wet soil in Btu/hr-ft-OF:
k,
= 0.578( a
log m + b)1O2
(8-52)
where a = 0.1424 - 0.000465(%~1);b = 0.0419 - 0.000313(%~1);%cl= percent of clay in the soil; z = 0 . 0 1 ~ ~pd; = density of dry soil, lb,/ft3; and m = moisture content of soil as a percentage of dry soil weight, 96. In order to determine the thermal diffusivity it is necessary to calculate density of wet soil, p,, in lb,/cu ft and specific heat of wet soil, c,, in Btu/lb,-OF. In terms of pd and cd, these variables are equal to: p, = pd(1
+ 0.000624m)
(8-53)
and c,
= (cd
+ 0.01m)/(1+
0.01m).
(8-54)
Combining the above three equations:
k, a,=--P W C W
- 0.578( a log,,m + b)lOz(l + 0.01m) pd(l
+ o.o00624m)(Cd -k 0.01m)
(8-55)
(2) Estimation of thermal constants by direct measurement Thermal conductivity is usually measured in situ by a field conductivity probe. Heat is applied at a constant rate to this thin cylindrical probe which is pushed into the ground. The rise in temperature of the probe surface is measured as a function of time. The k value can be determined from the heat input and temperature-time relationship. Yearly average diffusivity of the soil can be obtained from the measurements of the soil temperature at various depths below the surface. If the annual temperature
235
T i m e , t, sec
+
u Tsmin
-Pt Reference Temperature (O'F)
Fig. 8-11. Fluctuation of soil temperature with time. Z = ground surface, ZZ = at a depth x . Amplitude decreases with depth. (Modified after Szilas, 1975, p. 493, fig. 7.2-1.)
cycle is assumed to be sinusoidal (see Fig. 8-11), then the temperature change on the ground surface, AT,, is equal to:
AT, =AT,,,
sin at
(8-56)
where A denotes temperature fluctuation and w = 277/t,, with t, = time for 1 cycle. Figure 8-11 shows how the amplitude decreases with depth and that the temperature wave I1 at depth x is displaced in phase with respect to the ground-surface wave (see Szilas, 1975, p. 493). The temperature variations can be expressed as: aT at
a2T
--=ff-
(8-57)
ax2
where x = depth from the ground surface (x = 0 at ground surface). Using eq. 8-56 in solving eq. 8-57,
AT, =AT,,,
exp( -x/+)
sin(21it/tp - x/-)
(8- 58a)
and phase shift At, = ( x / 2 ) { F
(8-58b)
When the sine term in eq. 8-58a is equal to unity, the maximum temperature fluctuation (amplitude) at a depth x is obtained:
AT,,,
=A
qmax exp I- /x *s/at,
(8-58~)
236
Example 8-3. What are the extremes of daily temperature and the phase shift (a) at a depth of 1 ft and (b) at a depth of 5 f t if a = 5.3 X lo6 ft2/sec and the surface temperature varies from 90 to 40°F during the day.
Solution: See Fig. 8-11. The surface amplitude of the temperature fluctuation is equal to:
Using eq. 8-58a, the amplitude at a depth of 1 ft is equal to:
AT,,,
= 25
exp[ - (l){r/(86,400 x 5.3 x
] = f1.82"F.
At a depth of 5 ft, the amplitude is equal to:
AT^,^ = f 5 . i x 10-SOF. In order to evaluate the temperature extremes, first it is necessary to calculate the mean temperature, which is approximately equal to:
Then, the temperature extremes at a depth of 1 ft are equal to:
+ AT,,,
= 65
+ 1.82 = 66.82"F
Tlmin= T, - AT,,,
= 65
- 1.82 = 63.18"F.
T,,,
=
and -
At a depth of 5 ft: TSmax = 65.00005"F and TSmin = 64.99995"F. By using eq. 8-58b, the phase shift at a depth of 1 ft is equal to:
At,
= (;)/86,40O/(r = 36,017.5
X
5.3 X
sec = 10 hr.
Similarly, at a depth of 5 ft:
At,
= 50
hr.
237
Steady-state flow in buried pipelines
Temperature of the oil injected into a pipeline usually differs from that of the soil. Soil is generally cooler than the oil. Some of the factors contributing to the variation of axial oil temperature in the pipe are: (1) A part of the potential energy of the oil flowing in the pipeline is transformed into heat, due to the shearing at the liquid-solid pipe wall interface and the liquid-liquid shear within the bulk liquid. This heat increases the oil temperature. (2) Temperature also increases due to the exothermic process of separation of solid components (such as waxes) from the oil. Inasmuch as extensive separation leads to reduced flow efficiency and eventual blocking of the oil, crude oils are often dewaxed before pipeline transportation. (3) Temperature decreases due to the transfer of heat from the pipeline to the lower-temperature soil. In calculating the heat generated by friction, one can assume an average constant friction gradient along the pipeline. A pressure differential, Aprriction, is equivalent to a force F, = A p , (7rd2/4) in lb, along the length I of pipe. (1 ft-lb, = 0.001285 Btu). If the flow rate is Q (ft3/sec) and the heat generated by F, is q, then velocity, V, is equal to: V = Q/( 7rd2/4)
and q = 0.001285 F,I. The heat generated by friction per unit time per unit distance of flow, (Btu/sec-ft), is equal to:
@ = 4 =0.001285FrI/t It
I
=
[0.001285(A p , )( 7rd2/4) V ]/ I
@ J
= 0.001285AprQ / l
(8-59)
Using constant w and A, which are actually temperature dependent, the heat liberated per one O F drop in temperature (Btu/'F-sec) is equal to:
where Y
X = latent heat of fusion of the solids (paraffins) that separate out of the oil, Surface
Fig. 8-12. Schematic diagram of an infinitesimal element of flowing fluid, exhibiting a temperature drop dq.
238
Btu/lb,; w = amount of solids formed per lb, of oil per O F drop in temperature, lb/lb,il-oF; p = density of the paraffin-containing oil, lb,/ft3; and Q = oil flow rate, ft3/sec. On considering an infinitesimal volume element, as shown in Fig. 8-12 where the temperature of the flowing oil decreases by dT, over a length d l of pipe, one can obtain, by heat balance: PQC, dT,
=
Change in heat content of the flowing liquid
@ dl
+
Heat generated by friction
QpwXdT, Heat generated by separation of paraffins
-
k(T, - T,) d l Heat lost to surroundings
where T, = axial temperature of the flowing oil in the pipeline, OF; T, = original soil temperature at the same depth, OF; k = heat transfer coefficient per unit length of pipe, Btu/ft-hr-OF; and c, = specific heat of the flowing oil, Btu/lb,-OF: Or : (Qpc,-AQpw)
dT,=(@+kT,-kT,) d l
(8-61)
On substituting Qpc, - QpwX = A (constant) and Qz + kT, = B (constant) in eq. 8-61, integrating (initial conditions: I = 0 and T = TI): / ( B - kA T,)
dT,=/dl
Solving, and resubstituting for A and B, one obtains: (8-62) The oil temperature also varies radially. In turbulent flow, the radial temperature variation is slight, whereas in laminar flow, it may be significant. In the latter case, the axial temperature Tfis somewhat higher than the average temperature. Examples of temperature profiles are shown in Fig. 8-13. In practical calculations, the latent heat of fusion of paraffin is usually not considered. As shown in Example 8-4, however, it may be significant. Example 8-4 Oil is injected at a flow rate Q = 700 bbl/hr into a 20-mile long pipeline having I.D. = 12 in. The oil gravity is 0.85, temperature T, = llO°F, viscosity is 4.2 CP
239
Velocity
+
Temperoturr
I b)
(a)
Fig. 8-13. Velocity (a) and temperature (b) profiles for oil flowing in a pipeline. (Modified after 0.317 m/sec, 69"C, T, = 55.5"C. Chernikin, 1958 in: Szilas, 1975, fig. 7.2-4, p. 498).
r=
r=
(8.778 X slugs/ft-sec), soil temperature T, = 40 OF, relative roughness c/d = 0.0001, w = 0.003/"F, c, = 0.455 Btu/lb-OF, X = 200 Btu/lb,, and k = 1.2 Btu/hrft-OF. There are 2 globe valves, 1 long sweep elbow, and 100 couplings present. Find the exit temperature Tn of the oil: (1) Assuming friction and paraffin deposition over the entire length of the pipe. (2) Assuming negligible paraffin deposition. (3) Neglecting the effects of both friction and paraffin deposition.
Solution: Velocity of oil in the pipe, V = Q/( 7r/4)dz = [700 X 5.615/ 3600 ft3/sec]/[(7r/4)(1)2 ft2] = 1.39 ft/sec. Oil specific weight, y = 0.85 X 62.4 = 53.04 lb/ft3, whereas oil density, p = 53.04/32.2 = 1.647 slugs/ft3( = 53.04 lb,/ft3).
N Re = -d=Vp
(1)(1.39) (1.647)
P
(8.778 X
= 26,100.
From Fig. 8-2, for relative roughness r/d = 0.0001 and NRe= 26,100, the Fanning friction factor f = 0.0242. From Fig. 8-6, I, for a globe valve = 320 ft, and I,, for a long sweep elbow = 20 ft. From Fig. 8-3, K for a coupling = 0.04. Thus, the head loss due to friction is equal to:
*Pf=
2f/V2y
V2y
gdi+ K - 2g
Using eq. 8-59:
=
2(0.0242)[(20)(5280)
+ 2(320) + 201 (1.39)2(53.04) (32.2) (11
240
(1) From eq. 8-62, the exit temperature is equal to:
1
- (1.2)( 20)( 5280) 110 - 40 - 0.783 x exp[ 1.2 (700 X 5.615)(53.04)(0.455
+ 0.6)
1
+--0.783 -79.63'F 1.2 (2) On assuming w = 0,
- (1.2)(20)(5280) 0*783 110 - 40 - 1.2 x exp[ (700 X 5.615)(53.04)(0.455)
1
0.783 +=58.88"F 1.2 (3) On assuming w = 0, and @ = 0,
T,, = 40 + (110 - 40) exp
[
- (1.2)(20)(5280) (700 X 5.615)(53.04)(0.455)
1
= 58.40"F
This example enables one to evaluate the significance of each of these assumptions. The paraffin deposition, which is neglected by many pipeline engineers, has a very significant effect. It is important to understand what the heat transfer coefficient, k, per unit pipe length signifies. As shown in Fig. 8-14, there are a series of resistances to the transfer of heat from the oil to the surrounding soil. First there is heat transfer from the bulk oil to the pipe wall, which is a process of forced internal convection (see eqs. 8-48, 8-49, 8-50, and 8-51). For Na -= 100 and NRe < 2100 (eq. 8-48): NNu=3.66
[
0.085NGz
+ 1 + 0.047NZ3 ](
e)
0.14
Fig 8-14 Cross-sectional view of a buried insulated pipeline carrying oil
241
For NGz > 100 and NRe < 2100 (eq. 8-49): NNu= 1. 8 6 N g 3
r4
i-:L
For 2100 < NRe < 10,000 (i.e., the transition region; eq. 8-50):
”,
= 0.116( N i l 3 - 125) N ; i 3
[ 1+ (
0.14
(”)
d / ~ ~ / ~ ] PW
For NRe > 10,000 (turbulent flow; eq. 8-51): 0.14
N,,
=
These equations correspond to heat transfer by convection in the three zones of flow, i.e., laminar, transition, and turbulent. Thus, one can evaluate the Nusselt number N N u using the appropriate equation for a particular flow regime. Inasmuch as:
hidi NNu= k
(8-63)
k h i = NNuX di
(8-64)
In this manner, the heat transfer coefficient h (Btu/hr-ft2-OF) can be determined. It is called h i , because it is the heat transfer coefficient inside the pipe. Thus, the net heat transfer per unit length of pipe (Btu/hr-ft-OF) due to the internal forced convection is equal to:
Another resistance to heat transfer is the pipe wall. If the thermal conductivity of the pipe wall material is k,, then its net contribution in Btu/hr-ft-”F is equal to 2rkw/ln( do/di). In the presence of thermal insulation around the pipe, another factor (2nkin/ln( din/d,) is introduced, where kin= thermal conductivity of the insulating material and din= diameter of the insulating material. Finally, the thermal conductivity of the soil surrounding the pipeline must be considered. Assuming a constant undisturbed soil temperature T, at a depth x , it
242
can be shown that the net heat transfer coefficient of soil, h,, is approximately equal to :
In order to obtain the net effective resistance, it is necessary to add these four resistances ( = l/conductance) in series together:
R=-
1 hi(rdi)
1 +
2 r k w ln(
3)+
1 In( 2rk,
3)+
1 -
(8-66)
hordin
Thus, the total heat transfer coefficient per unit length is equal to: k- = - 1=
‘
R
r
1
-+ hidi
1 - In(%) 2kw di
.
+
1 ln(2) 2kin
+
1 hodin
(8-67)
The kin values for some common insulating materials are presented in Table 8-1. Pipes, which are usually laid in ditches dug for that purpose, are covered with backfill. Inasmuch as the porosity of this backfill is greater than that of the undisturbed soil, its thermal conductivity is lower. The backfill compacts with time, however, causing the effective heat transfer coefficient to increase slightly. Other variables are wind, plant cover, snow, moisture content of the soil, etc. Under steady-state conditions, the heat transfer across the flowing oil into the soil is equal to:
= 2rkin(T0= hodin..(
qn)/ln(dh/do)
qn- T , )
(8-68)
One can assume that the inside and outside wall temperatures of the pipe are After Rohsenow and Hartnett (1973) who gave an appropiate relationship: h,
=
2ks
dincosh-’(2x/din)
where cosh-’(2x/din) is approximately equal to In(4x/din).
243
TABLE 8-1 Thermal conductivities of pipeline insulators (after Thomas and Turner, 1953; courtesy of Chemical Engineer) Material
Resistance towater
Cellular glass Cellular silica Cement (semi thermosetting) Cement (hydraulic setting) Diatomaceous earth Burnt diatomaceous earth bricks Burnt diatomaceous earth paste Glass fibers (formed into pipe insulation and blocks) Hydrous calcium silicate Perlite Polystyrene (expanded) Polyurethane (expanded)
Excellent Excellent Water will soften the dried cement Will not soften when wetted
Temperature limits Thermal conductivity (OF) (Btu/hr-ft 2-oF/in.)
Specific weight
Min.
Max.
32°F
70°F
212OF
500OF
-400 - 300 100
800 1600 1800
0.35 -
0.39 0.44 0.69
0.46 0.62 0.820
1.0 1.058
10 10-12 26
32
1200
0.46
0.525
0.610
0.840
49
Fair
22
1900
-
0.66
0.720
-
23
Moderate
22
1900
0.603
0.617
0.694
0.742
29.95
Moderate Excellent
22 - 300
1900 600
0.520 0.270
0.541 0.273
0.590 0.320
0.610 -
31.2 8
Good Excellent Excellent
100 32 - 300
1200 1200 175
0.23
0.37 0.38 0.26
0.41 0.43 -
0.52 0.58 -
12 10-12 2-2.3
- 50
230
-
0.17
0.35
-
Good
3
equal ( q= To),and then break up the above multiple equation into three independent equations:
k ( T , - T,) =mdihi(T,- Ti)
(8-69) (8-70)
and
hidi&- Ti) =hodinpin- T , )
(8-71)
If Ti is unknown, one should use a trial and error procedure. Assume Ti and establish the physical parameters of the oil (density, viscosity, specific heat, etc.) to
244
Fig. 8-15. Universal viscosity versus temperature chart for crude oils. (After Frick, 1962, p. 19-39, fig. 19-41; courtesy of McGraw-Hill, New York).
evaluate NRe, NP,, NGr,Na and NNu. Knowing NNu,calculate h i and then k using eq. 8-67. Another k value is obtained using eq. 8-69. If the k values obtained by the two methods agree, then the assumed value of Ti is correct. If not, the calculation procedure is repeated with a different value for Ti, until one arrives at an acceptable agreement. To carry out such calculations, the variation of the physical parameters (viscosity, density, etc.) of the oil with temperature must be known. Viscosity Temperature and liquid viscosity may be correlated within the accuracy of most experimental data (1-2%) with the de Guzman-Andrade Equation:
where A and B are constants. This equation implies that a plot of log p versus 1/T will yield a straight line. Figure 8-15 illustrates the above equation, which is generally used. Beggs and Robinson (1975) presented a more recent empirical correlation between viscosity and temperature, which gives better results than the frequently used Beal's (1946) correlations: pdo = 10" - 1
(8-73)
where pdo= viscosity of dead oil (gas-free crude oil), cP; and x = Y T - ' . ' ~ ~where , T - O F ; y = 10'; and z = 3.0324 - 0.02023 ("API).
245
Fig. 8-16. Viscosity of gas-free crude oil at reservoir temperatures. (After Beal, 1946, fig. 8, p. 103; courtesy of the S.P.E. of A.I.M.E.).
I00
80
60 40
20
10
8 6 4
2
I
08 06 04 I-
4, 0 2
1111 0.4
OK08
I
2
4
6
8 10
I
I I1 I IllllJ 20
40
60 BOIOO
VISCOSITY OF DEAD OIL, CENTIPOISES (AT RESERVOIR TEMPERATURE AND ATMOSPHERIC PRESSURE)
Fig. 8-17. The viscosity of gas-saturated crude oils at reservoir temperature and bubble-point pressure. (From Chew and Connally, 1959, fig. 2, p. 25; courtesy of the S.P.E. of A.I.M.E.).
246
Fig. 8-18. Relationship between the viscosity of oil and pressure (above the bubble-point). (From Bed, 1946, fig. 11, p. 109; courtesy of the S.P.E.of A.I.M.E.).
For crude oil containing dissolved gas, Beggs and Robinson provided the following correction for viscosity:
(8-74)
Po =A d o
+
where A = 10.715(RS+ 100)-0.515;B = 5.44 ( R , 150)-0.338;and R , = amount of dissolved gas in oil, scf/STB. This equation gives the viscosity of gas-containing (live) oil. This correlation was developed by plotting log,, T versus log log(pdo+ 1). Straight lines were obtained and it was found that each of these lines corresponded to crude oils having a particular API gravity, which led to the development of eq. 8-73. Because of their popularity, Beal’s correlations are also presented in Fig. 8-16 and 8-18, along with the Chew and Connally (1959) correlation (fig. 8-17). Density The following equation takes into account the effects of both temperature and pressure on the density of the oil:
or
where p T = density of oil at any temperature T, lb,/ft3; pn = density of the oil at. the base temperature T, and base pressure (usually at standard conditions); a T = temperature coefficient of oil density, Ib,/ft’-’F; ap = pressure coefficient of oil density, l/ft; p = pressure, lb/ft2; and y = specific weight, lb/ft3.
247
The coefficients aT and ap can be evaluated if the densities are known for any two known conditions of temperature and pressure. There are other empirical correlations which are not presented here. Kinematic viscosity A useful correlation between temperature and kinematic viscosity Y (which combines both the viscosity p and the density p of the oil) was proposed by Walther. This equation, however, neglects the effect of pressure, which is generally small:
log log(102v + a ) = b
+ c log T
(8-76)
where a , b, and c are constants, their value depending upon the particular oil in question. The value of a is around 0.8 for most of the oils. On incorporating this value into the above equation, one obtains: log log(102v + 0.8) = b + c log T where
Y = cm2/sec
(8-77)
(or stokes) and T = temperature, "Rankine.
Specific heat Cragoe (1929) presented a correlation for the estimation of the specific heat cp in Btu/lb,-'F for petroleum oils: c =
0.388
+ 0.00045T G,0.5
P
(8-78)
\
where T is the temperature in O F and Go is the liquid specific gravity at 60°F. The accuracy of this equation is +5%. Thermal conductivity The American Petroleum Institute (1966) recommended a single value of 0.077 Btu/hr-ft-OF at 30°C (86°F) for thermal conductivity of petroleum oils. The average and maximum deviations from this value are 7 and 308, respectively. At other temperatures, Cragoe's equation gives satisfactory results, with average and maximum deviations of 12 and 398, respectively: k= 0'0677 [ 1 - 0.0003( T - 32)] Go
(8-79)
where 32" < T -= 392°F. Example 8-5. Determine the heat transfer coefficient for a 10-mile pipeline buried 4 ft below the surface. Pipe having O.D. = 12.75 in. and I.D. = 12.438 in. carries a crude oil having specific weight of 53 lb/ft3 (Go= 0.85 or 35'API). Given: = 110'F; T, = 35'F; aT = 0.25 1b/ft2-'F; k , = 1.0 Btu/hr-ft-"F. Assume laminar flow: NRe= 2000.
248
Solution: For the first trial, Ti = 80°F is assumed. Using eq. 8-73:
z = 3.0324 - 0.02023(35) = 2.32435 y
= 10' = 102,32435= 211.033
x =yT-'.163 = (211.033)(80)-
...
= 101.2914-
1.163
- 1.2914
1 = 18.56 CP
and p,(at T,) = 6.8 cP. Neglecting the effect of pressure, the density is equal to (eq. 8-75): p r = p , - a T ( T - T,)=53.0-0.25(80-60)=48
lbm/ft3.
Using eq. 8-78, specific heat of oil is equal to:
The thermal conductivity of oil can be determined from eq. 8-79: k o = -[l - 0.0003(80 - 32)] 0.85
The Graetz number, N,(
= 0.0785
= N R e N P r d / l )is
Btu/hr-ft-"F
equal to:
1 [lzO]
(18.56 X 6.72 X 1 0 - 4 ) ( 0 . 4 7 1Btu 6)7X lbm Ibm ft-sec NGz = (2000) 0.0785 Btu 3600 sec-ft-OF Inasmuch as Ncz < 100, one can use eq. 8-48 to find NNu: NNu = 3.66
+
h i= NNux
k o = (0.844)[
di
1i:;l;i2]
= 0.0635
Btu/hr-ft2-OF
12.438
X
= 10.6
249
and
-
h, =
din ln(
2)
4x4 In( 12.75/12
= 0.6941 Btu/hr-ft2-"F
)
Thus, k can be determined using eq. 8-67: k=
1T
1/(0.06352
X
12.438/12)
+ 1/(0.6941
X 12.75/12)
= 0.0604 Btu/hr-ft-'F
and using eq. 8-69: k=
mdihi(T, - q ) ( T f - T,)
= 0.0827
=
[ ~(12.438/12)(0.0635)(110 - 80)]/(110 - 35)
Btu/hr-ft-"F
On assuming Ti = 82"F, as a further approximation, the above procedure is repeated : Equation 8-67 gives k = 0.0711 Btu/hr-ft-"F, whereas eq. 8-69 gives k = 0.0705 Btu/hr-ft-OF, which is a reasonable agreement. Thus, k = 0.071 Btu/hr-ft-"F and Ti = 82°F. Transient (unsteady state) flow of oil in buried pipelines
If the inlet temperature, flow rate, the physical parameters of the oil, and the soil temperature are constant over a comparatively long period of time, the heat flow in and around a buried pipeline must be steady. These conditions are, however, never satisfied. Slight departures from the steady-state behavior are approximated by the steady-state relationships. In numerous practical situations listed below, however, the departure may be sufficient to render the steady-state relationships useless for even an approximate quantitative evaluation: (1) Termination or beginning of flow in the pipeline. (2) Change in injection temperature of oil into the pipeline. (3) Fluctuations in temperatures. (4) Fluctuations in flow rates. Given sufficient time, any kind of flow (except the fluctuating type) will eventually stabilize and reach steady state. Numerous models and calculation methods have been devised to describe and solve unsteady-state systems. Szilas (1975) presented a transient model which describes the temperature changes in a pipeline shut down after steady-state flow.
' This book was originally published in Hungary in 1968
250
He called it a “cooling model” for an insulated pipeline. In the case of shutdown after flow, it is assumed that: (1) The pipeline is embedded in an infinite half-space filled with soil of homogeneous thermal properties. (2) The temperature of the soil in contact with the pipeline can be described by Chernikin’s (1958) model. Heat flowing through the wall of unity pipe length into the cooler ground over an infinitesimal period of time d t reduces the temperature of oil and pipe by dT:
[
rd?
4po~o
1
+ -774 ( d ? - df ) pPcp dTo = - k (T; - Tin) d t
where subscript o refers to oil, p refers to pipe, and in refers to pipe insulation; T,, and T[n are transient temperatures. Neglecting h i and h , and using eq. 8-67: k = r/ [ (1/2kin) In( din/do
)I
Setting:
77 [ dfpoc0 + (d,2 - df)PpCp] = 4 4k and rearranging: A[dTo/(T,’- T6)]
=
-dt
(8-80)
The Chernikin relationship is then used here to describe the change in the oil temperature, TA, and the change in the outer temperature of the insulation Tin. If point PI is the image of the projection of the linear heat source on a plane perpendicular to it (see Fig. 8-19), then the difference in temperature between the point Pz (lying in the plane of projection and defined by the coordinates y and z) and the undisturbed soil is equal to:
TL- T,= (@/4ak,)[Ei(-xz/r2X1/NF,)
-Ei(-1/4NFo)]
(8-81)
where NFo= q / r 2 = Fourier number (dimensionless), and Ei is the exponential integral function.
Fig. 8-19. The Chernikin model for unsteady state flow of oil in buried pipelines. (Modified after Chernikin, 1958; in: Szilas, 1975, p. 517, fig. 7.2-13; courtesy of Akadkmiai Kiad6, Budapest, Hungary).
25 1
If t = 00 (steady state flow), eq. 8-81 reduces to the following form:
To - T, = (@/21rk,) ln(2x/r)
(8-82)
Dividing eq. 8-81 by eq. 8-82:
Inasmuch as x, r , k,, ps and c, are constant for a given pipeline, the above equation can be written as follows:
where
The process of warming 'up is described in eq. 8-85. In the case of cooling down after steady-state flow:
( T; - T , ) / ( To- T,) = 1 - x or Td=(l-X)(To-T,)+T,
(8-86)
The relationship X = f(t) for a given case can be plotted on a graph using eq. 8-86. Sections of this curve can be individually approximated by the following relation: X=a+bht
(8-87)
In addition, an equation similar to eq. 8-86 can be written for the insulation around the pipe:
TL = (1 - X ) ( Ti,- T,) + T,
(8-88)
Combining eqs. 8-87 and 8-88:
Ti',= (1- a - b h t )( Ti,- T,) + T, = T, + (1- a ) ( Tin- T,) - b( Ti,- T,) In t or
Ti:, = B - C In t where B = T, + (1 - a)(Ti, - T,) and C = b(Ti, - T,).
(8-89)
252
Substituting
’[
q,,from eq. 8-89 into eq. 8-80:
T d - BdTo +Clnt
I
=
-dt
(8-90)
The general solution of eq. 8-90 is: T,’=B-C In t + [CEi(t/A)C’] e-‘IA
(8-91)
where C’ is the constant of integration. In the case of initial condition T,’ = at t = 0: Ti = B - C In t
+ [ CEi( t/A)To - B - C(0.5772 - In A ) ] e-r/A
(8-92)
The above relationship can be used to evaluate the variation of temperature Ti versus the time t elapsed after shutdown of the pipeline at any pipe section situated at a distance 1 from the head end of the line (See Szilas, 1975, p. 519.). Heating up of a cold line by introduction of hot oil When hot oil is pumped into a cold line, the initial transient heat loss-to the cold soil can be very much higher than the equilibrium steady-state heat loss. In large crude oil lines, it may take many days or even weeks before equilibrium conditions are established. It is, therefore, necessary to know how the transient heat loss varies as a function of time. (See Davenport and Conti, 1971.) In the simple case, the pipe is assumed to be surrounded by an “infinite sea” of soil having homogeneous thermal properties. The heat transfer between the ground surface and air is assumed to be infinite. Thus, the heat loss is directionally symmetrical and for this simple case, the Nusselt number is related to the Fourier number as follows (see Davenport and Conti, 1971): N,,
= 0.362
+ 0.953/N;:’
(8-93)
The soil will warm up continuously and will reach equilibrium with the pipe surface. It is assumed that equilibrium is reached when the Nusselt number from eq. 8-93 equals that from eq. 8-94: (N,,),
= 2/cosh-’(2x/d0)
= 2/ln(4x/d0)
if x >, do
(8-94)
Thus, it is possible to calculate the time necessary to reach equilibrium, which is a function of NFo and x / d o . Due to the approximation employed, NNuis a function of x/do only in eq. 8-94. Substituting eq. 8-94 into eq. 8-93, yields: r
7
(8-95)
253
where N,, = cy,t/d:; x = depth of burial of pipeline, ft; and do = outside pipe diameter, ft. This relationship gives the time of transient flow in the pipeline as a function of x and do. As the temperature wave from the pipeline reaches the surface and is reflected, the surface begins to play an important role and the infinite sea approximation is invalid. The heat loss exceeds the calculated value by 10 - 15%thereafter.
TRANSPORTATION OF HEAVY OILS IN PIPELINES
Heavy oils, characterized by high viscosities, high pour points, and low API gravities, are currently being transported only to a limited extent by pipelines. Development of the largely untapped world resources of heavy crude oils, however, is changing this. The methods for pipelining heavy crude oils are briefly discussed here. (See Barry, 1971.) The major problems are caused by: (1) pour point (wax crystallization problems) and ( 2 ) viscosity (flow problems). High pour point temperatures are caused by excessive formation of wax'crystals in the oil, which inhibit its ability to flow. Wax deposits form in the storage tanks as well as in the pipeline. With increasing viscosity, the head loss due to friction increases and, therefore, greater pump horsepower is required. Thus, it is necessary to reduce the viscosity and the pour point of the oil being transported, which can be achieved by the following means: (1) Use of additives (pour point depressants). ( 2 ) Preparation of an oil-solid slurry. (3) Heating the oil to keep the flow essentially above the pour point, and also to reduce viscosity. (4) Dilution of oil with a solvent or with another oil for reduction of viscosity and pour point. ( 5 ) Preparation of a lower-viscosity, unstable slurry-emulsion system by mixing water with the oil. The first two techniques are applicable only in cases where the oil viscosity is low enough to permit economical pipelining at the existing temperature. The third and fourth techniques are currently the major methods for the pipelining of heavy crude oils. The fifth technique has not been used yet on a commercial scale. The transportation of hot oil by pipelines, which has been discussed already, requires some means of heating the oil. Usually direct-fired heaters are used. The dilution technique involves the addition of low-viscosity hydrocarbons such as condensate, natural gasoline, and naphtha in order to reduce viscosity of heavy oil. In the case of oil fields located in remote areas, where the diluent may not be readily available, two major alternatives are available: (1) Use of a dual pipeline, so that one pipeline could carry the diluent back to the field. ( 2 ) Installation of a thermal cracker, catalytic cracker, or a hydrocracker to crack a portion of the heavy crude to
254
lighter components, which can then be blended with the remaining heavy oil for pipelining. The choice of a technique is determined by the oil viscosity, the geographic location, length of pipeline, process feasibility, and economic considerations. PIPELINE TRANSPORTATION OF NATURAL GAS
Physical properties of gases
Physical properties of natural gases (compressibility, density, viscosity, and specific heat), which affect gas transmission in pipelines are discussed first in this section. Gas compressibility Deviation from ideal behavior of natural gas is seldom negligible. For practical purposes, only compressibility factor Z is applied: (8-96)
p v = ZnRT
GAS
GRAVITY
(AIR=I)
Fig. 8-20. Pseudocritical properties of condensate well fluids and miscellaneous natural gases. (After Brown et al., 1948; courtesy of Natural Gasoline Association of America).
255
where p = absolute pressure of the gas, psia; T/;= total gas volume, ft3; Z = compressibility factor; n = number of moles of gas in volume y ; T = absolute gas temperature, OR; and R = universal gas constant = 10.73 psia-ft3/lb mole-OF.
PSEUDO REDUCED PRESSURE
PR
Fig. 8-21. Compressibility factor for natural gases as a function of reduced pressure and temperature. (From Standing and Katz, 1942, fig. 2, p. 144; courtesy of the S.P.E.of A.I.M.E.).
256
According to the law of corresponding states, the compressibility factors Z of two gases are equal if the reduced state parameters of these gases are equal. This is the basis for the generalized compressibility factor chart shown in Fig. 8-21. Reduced pressure p , is equal to p / p c and the reduced temperature T, is equal to T/T,, where p, and T, are the critical pressure and temperature of the gas, respectively. For a mixture of gases, the reduced state parameters are replaced by the pseudo-reduced parameters ppr and Tp,, which are defined as follows: n
~ p =c
C Yipci
(8-97)
i=l
and n
Tpc =
C YiTci
(8-98)
i-1
where yi = mole fraction of component i in a gas mixture. The molecular weight, M , is equal to: n
M= CyiMj
(8-99)
i=l
If the gas composition is not known, one can use approximate empirical correlations (see Fig. 8-20) developed by Brown et al. (1948). Natural gas often contains nonhydrocarbon gases such as N, and CO,. If the N, content is less than 8% and that of CO, is less than 108, one can use the additive rule to determine compressibility factor Z :
where ZHc is the compressibility factor for the pure hydrocarbon gas, which can be determined as before. The compressibility charts for N, and CO, gases are given in literature. Density The gas density, p , at a pressure p and temperature T can be obtained from eq. 8-96 as follows: n / y =p/ZRT
(8-100)
Inasmuch as number of moles n = mass/molecular weight = m / M , eq. 8-100 becomes: m/K=pM/ZRT=p
(8-101)
251
.E3
Fig. 8-22. Relationship between the viscosity of paraffin gases and molecular weight at a pressure of orre atmosphere and reservoir temperatures, with corrections for nitrogen, carbon dioxide, and hydrogen sulfide. (From Carr et al., 1954, fig. 6, p. 268; courtesy of the S.P.E. of A.I.M.E.).
258
Fig. 8-23. Viscosity ratio, p / p l , as a function of pseudoreduced temperature and pseudoreduced pressure. (From Carr et al., 1954 p. 267, fig. 4; courtesy of the S.P.E.of A.I.M.E.). p = viscosity of gas at reduced temperature, T,, and reduced pressure, P,; p1= viscosity of gas at atmospheric pressure and at temperature T,.
Gas gravity, Gg, is defined as the ratio of the density of the gas, pe, to the density of air, pa, under standard conditions: (8-102) because Z = 1 at standard conditions. Inasmuch as molecular weight of air is equal to 28.97, eq. 8-102 can be written as follows: Gg= M/28.97
Viscosity Gas viscosity decreases with increasing molecular weight and increases with increasing pressure and temperature. The pressure effect is the same as in liquids, whereas the temperature effect is just the opposite. The two most widely used correlations are those of Carr et al. (1954, see Figs. 8-22 and 8-23) and Lee et al. (1966). Beggs and Brill (1972), gave the following modified form of Lee's equation: 1.1= A x
exp( BpC)
where p is the viscosity in cP, A
(8-103) = (9.4
+ 0.02M)T1.s/(209 + 19M + T ) ; B = 3.5 t
259
Temperature,
OF
Fig. 8-24.Viscosity of natural gases at atmospheric pressure. (After C a r et al., 1954, fig. 7, p. 269; courtesy of the S.P.E.of A.I.M.E.). Data is based on N.B.S.-N.A.C.A. tables of thermal properties of gases and research work by J.O. Hirschfelder, R.B. Bird and E.L.Spotz (1949),M. Trantz and K.G. Sorg (1931),and A.O. Rankine and C.J. Smith (1921).
+
986/T 0.01M; C = 2.4 - 0.2.4; M = molecular weight of the gas; T is the temperature in OR; and p is the density in g/cm3. The viscosities of natural gases at atmospheric pressure are given in Fig. 8-24. Specific heat Specific heat is defined as the amount of heat required to raise the temperature of a unit mass of a substance by one degree. It can be measured at a constant pressure, cp, or at constant volume, co, resulting in two distinct specific heat values. For an ideal gas, the difference between the two is equal to the gas constant R: cP - C,
=R
(8-104)
The cp of a gas mixture can be calculated from the following formula: n
cp=
c
YiC,i
(8-105)
i-1
Product of gas constant R o and molecular weight M is called Universal Gas Constant ( R= R o M ) and is equal to = 10.732 p~ia-ft~/Ibmole-~ R = 1544 ft-lbf/lbmole-o R = 1.986 Btu/lbmole-o R.
260
.W
TEMPERATURE DEtREfS FAHRENHEIT
Fig. 8-25. Specific heat, cp, of hydrocarbon vapors at a pressure of one atmosphere. (After Brown, 1945, fig. 1, p. 66; courtesy of the S.P.E.of A.I.M.E.).
Brown (1945) has presented the molar specific heats at constant pressure for the individual hydrocarbon gases (Fig. 8-25). Equation 8-105 can be used to determine approximately the specific heat of a mixture of gases if the composition is known.
261
0.04 0.060.000.10
0.2 0.4 0.6 0.8 I Reduced pressure pr=P/P,
2
4
Fig. 8-26. Generalized relationship of heat capacity differences, cp - c,,, to reduced pressure and reduced temperature. (After Edmister, 1948, p. 613, in: Perry and Chilton, 1973, fig. 3-54, p. 3-238; courtesy of Petroleum Refiner.)
The specific heat capacity ratio, c,/c,, for an ideal gas can be easily obtained using eq. 8-104. For a real gas, CJC, ratio is equal to:
The ( c p - c u ) quantity, which is larger than R for a real gas, may be obtained from the generalized chart in Fig. 8-26. Gas flow fundamentals
Detailed treatment of gas flow fundamentals was presented by the Institute of Gas Technology (1972). Some important equations are presented here. Q b = 38.774-
1
p: -pz - 0.0375GgA~p,Z,,/(TZ) Os x d 2s G,ITZf
QbGgpb NRe = 0.4775 pdiTb
(8-107)
(8-108)
262
(8-109) r
1
(8-110)
(8-111) (8-112)
Pavg = 0.67(
-)
(8-113)
where Q,=volumetric flow rate at pressure Pb and temperature T,,, Mcf/D; BI = bend index, degrees/mile; Ff = drag factor; f = Fanning friction factor; fpt = Fanning friction factor for partially turbulent flow; fft = Fanning friction factor for fully turbulent flow; Gg = gas specific gravity (air = 1); A t = elevation, ft (positive for net uphill flow and negative for net downhill flow); di = internal pipe diameter, inches; I = pipeline length, miles; p b = base pressure, psia (usually atmospheric); pavg= average pressure, psia; p 1 = upstream pressure, psia; p 2 = downstream pressure, psia; T = flowing gas temperature, OR; Tb = base temperature, OR; Z = gas compressibility factor; p = viscosity, lb,/ft-sec; and E = effective or operating roughness, microinches: Steel pipe
Roughness E (microinches)
New 12-months old 24-months old Plastic-lined, or sand blasted
500-700 1500 1750 200-300
The roughness E for aluminium pipe is equal to 200 microinches. Weymouth approximation A very widely used formula for volumetric flow rate (Qb), which was developed by Weymouth, can be derived as follows:
f = 0.0035rK
(8-114)
and (8-115)
263
where M = molecular weight of the gas, lb/lb-mole; R = universal gas constant = 10.73 psia-ft3/lb-mole-"R; T= average temperature, OR; and average compressibility factor at average pressure, pavg,and average temperature, $? Rearranging eq. 8-115:
z=
(8-116a) Substituting eq. 8-114:
f = 0.0035&, and R
= 10.73 psia-ft3/lb-mole-"R
in eq. 8-116a:
or (8-116b)
Mean pressure eualuation
The mean pressure for an incompressible fluid is simp] the arithmetic average of the inlet and outlet pressures. As discussed below, this is not the case for compressible fluids, i.e., gases. One can derive a simple formula to determine the pressure at any fractional distance x from the inlet end in a pipe carrying gas (see Fig. 8-27). Using eq. 8-116b, at point C where l = l ( x ) from inlet end and I = l(1 - x ) from outlet end:
Fig.8-27. Mean pressure evaluation.
264
x +
Fig. 8-28. Pressure profile along the length of a horizontal high-pressuregas pipeline shown in Fig. 8-27. (After Szilas, 1975, p. 36, fig. 1.2-1; courtesy of Akadtmiai Kiad6, Budapest, Hungary).
and 10.5
Equating these two equations: p:-p,Z -X
P,Z-P22 1-x
Solving for p,: p,
=
[ P: - x ( P: - Pz’>I0.5
(8-117)
Equation 8-117 suggests a pressure profile as shown Fig. 8-28. The mean pressure is given by the following equation:
or
If in eq. 8-116:
265
di is expressed in inches, 1 in miles, and Qb in Mcf/D, one obtains the following equation:
Qb
[
43.487
[
= (52so)y’]
24 X 3600 1000
w]
] (12“))[ d8l3
-
0.5
(Tb’pb)
(8-119)
or (8-120)
where Qb is in Mcf/D at Tb and p b , d , is in inches, T is in OR, p is in psia, and I is in miles. The following approximate forms of the Weymouth equation are also used: (8-121)
and (8-122)
where Q is measured in Mcf/D at the average pressure and temperature. In eq. 8-122, the friction factor, f, has been included. It can be included in a similar way in eq. 8-120. The hydrate point for hydrocarbon gases (see Szilas, 1975)
Gases like methane, ethane, propane, and butane can enter the water lattice without forming a chemical bond. Upon sufficient decrease in temperature, a solid granular substance forms, which resembles snow or ice and is called hydrocarbon hydrate. Besides hydrocarbon gas molecules, hydrates can also be formed by nonhydrocarbon gases, such as nitrogen, carbon dioxide, and hydrogen sulphide. The conditions necessary for hydrate formation and stabilization are (Szilas, 1975): (1) The water must be in a liquid state during hydrate formation. (2) Low temperature and high pressure. (3) The gas must be of a covalent bond type with molecules smaller than 8 A units. (4) The hydrate must be water resistant. ( 5 ) The gas must be immiscible with water in the liquid state. (6) No van der Waals’ forces should arise between the hydrate molecules. Several methods of determining the hydrate point temperature for a given
266
pressure and gas composition have been described in literature. The Katz et al. (1959) procedure seems to be the best suited one for natural gases devoid of nitrogen and up to about 275 atm (4000 psi). Hydrates form when the following condition is satisfied:
(8-123) where y, = mole fraction of i-th component in the gas and Khi = equilibrium ratio for the i-th component obtained from the Kh versus temperature-pressure charts (Katz et al., 1959). Heinze (1971) gave the following relationship for the hydrate point for nitrogencontaining natural gases up to 395 atm (5800 psi):
Th=
[
!lyiKhi]
(8-124)
0.445
where Th is the hydrate-formation temperature in OK. Gas transmission systems Gas transmission systems are complex network (looped or loopless) systems, flow through which can not be treated like a simple single-line flow. Flow may be at a high pressure with significant increase in specific volume with declining pressure, as in the case of systems for bulk transport of gas from the field to the regional supply stations. Flow may also be at a low pressure with negligible specific volume changes, as in the case of supply from the regional stations to the consumers. Although the flow in a transmission system is invariably transient, the assumption of steady state flow is valid for many design and operation problems. System of parallel lines As in liquid flow,
(8-125) where QA, QB, and Q , are the flow rates in lines A, B and C, respectively, and QT is the total flow rate. Also:
(8-126) because the end points of all three lines are common. The total pressure drop is given by the following equation (see Fig. 8-29): A P T = ApA(or
A P B or A PC ) + A P D 9
(8-127)
267 A
D
l M
I
Section I
I I
io !
Section 2
~
Fig. 8-29. Pipelines in parallel for Example 8-6.
The Weymouth equation (eq. 8-120) can be written as follows: (8-128) where K includes the constant terms. Thus, one can use an equivalent single line having a length 1, and an equivalent diameter d , , that has the same capacity as the looped line under the same pressure drop. This is given in Table 8-11. Lines in series A system of lines in series is shown in Fig. 8-30. In this case, Q =Q = Q =Q 1
2
3
T
and
where Q , , Q 2 ,and Q 3 are the flow rates in lines 1, 2, and 3, respectively; A p , , A p , , Ap, are the respective pressure drops in these lines; A p , is the total pressure drop; and Q , is the total flow rate. For this system, it is very simple to define a single line having diameter d , that is equivalent to I, miles of line of diameter d 2 and I , miles of line of diameter d 3 . TABLE 8-11 Equivalent diameter and equivalent length for lines in parallel and lines in series Equivalent diameter Lines in parallel Lines in series a
Entire line looped
a
Q 3/16
It is assumed that lengths of all lines in the looped section are equal.
Equivalent length -
268 Line I
I
Fig. 8-30. Pipelines in series.
Alternatively, one can determine the diameter of a line having length I,, which is equivalent to the above mentioned lines (Table 8-11). Example 8-6
In Fig. 8-30, assume d, = 2 in.,d , = 2.5 in., d , = 3 in.,1, = 1, = 1, = 15 d e S , d , = 3 in. and I, = 2 miles. Determine the equivalent ~ f : / ' / l ; / ~ratio for use in the Weymouth equation for this system. Solution: Refer to Table 8-11. For the looped part,
d,"/'
(2)"'
+-( 2 .5)8/3 +--(3)"/' - 9.446-in.8/3 (15)'12 (15)'/2 (15)'12 ft'/2
-=-
Inasmuch as length 1, = 15 miles,
d,"/' = (9.446)(15)'/2
= 36.584
Thus: d , = (36.584)3/8 = 3.857 in. The resulting configuration is shown in Fig. 8-31. This represents two pipes in series. Choosing an equivalent length:
( 3*8357
IDe= 2 miles x -)16/'
= 7.639
miles.
Therefore, the equivalent length of the full system having diameter of 3.857 in. is equal to 15 7.639 = 22.639 miles.
+
Alternatively: Equivalent length of the 15-mile section as a 3-in. diameter pipe= 15 X (3/3.587)16/' = 3.927 miles. Hence, the equivalent length of the full system having diameter of 3 in. is equal to 2 3.927 = 5.927 miles.
+
D
C3in
t 15 miles
. 1-
2 miles
Fig. 8-31. Pipeline system for Example 8-6.
269
Thus, the system in Fig. 8-30 is equivalent to: (1) 22.64-mile long 3.857-in. diameter pipe; or (2) 5.93-mile long 3-in. diameter pipe. Flow in hilly terrain Frequently the terrain over which the pipeline is laid may not be horizontal. In such cases, the horizontal flow equations presented for gas flow have to be modified to account for the net head (or elevation) change from inlet to the outlet, as described earlier for flow of liquid oil. For gases, this is complicated by the fact that the gas properties are very sensitive to pressure. A simple, yet valid, approach is to modify the Weymouth equation by replacing the [( p : - p i ) / 1 ) ] term by:
PI 1
where 1 = total length of pipeline; 3 = 2Ggz/53.33TaZa; z = total elevation difference between inlet and outlet (= zOutle, - zinlet),ft. These relationships can be used to predict the performance of many systems. However, it must be realized that in real situations numerical simulation techniques may have to be employed to enable more accurate predictions. The interested reader is referred to Szilas (1975) and other references listed for a detailed treatment of the subject.
ACKNOWLEDGEMENTS
The authors are very thankful to Mr. Rajay K. Goyal for his valuable comments and suggestions during the preparation of this chapter.
APPENDIX 8.1
One of the most routine calculation procedures in a gas field is the determination of deliverability. This involves a trial and error type solution for the flow capacities of the tubing in the well and the surface flow pipes. Iterative solutions are introduced by the gas properties (viscosity, 2 factor, etc.), which have to be evaluated at the average pressure. The calculation procedure is outlined below: (1) Determine the gas flow in the reservoir:
'
=
kh ( F i - P $ ) 1422(pZ),,TR [1n(0.472re/rw) + S ]
(8 .I-1)
270
(2) Determine the gas flow in the tubing and casing from downhole to the wellhead:
Q 2 = (198.6)2
(8.1-2)
Express the flow for each well. (3) Determine the gas flow in the surface flowline from the wellhead to the supply terminal(s): (8 .I-3)
In all these equations, Q = gas flow rate, Mcf/D; pwr= bottomhole flowing pressure, psia; jiR= average reservoir pressure, psia; pwh = wellhead flowing pressure, psia; pd= pressure at supply end of pipeline, psia; k = reservoir permeability, md; h = formation thickness, .ft; p = gas viscosity, cP; Z = gas compressibility factor; Gg= gas gravity with respect to air = 1; T = temperature, OR; 1 = length of flow conduit, ft; f = friction factor, dimensionless; S = skin factor, dimensionless; d = diameter of flow conduit, in.; re = drainage radius for the producing well, ft; rw = wellbore radius, ft; and s^ = 0.0375 Gglt/TaZa. Subscripts: t = tubing from bottomhole to the surface; s = surface flowline; av = average conditions in the reservoir; a = average conditions between the bottomhole and the wellhead; and R = reservoir condition. The usual calculation procedure is to assume that the gas flow rate, Q, is equal throughout and establish the unknown wellhead pressure, pwh, through iterative techniques. First, assume a reasonable value of pwh, obtain the average pressure and temperature, and then evaluate the average p and Z for the gas. Use eqs. 8.1-1, 8.1-2 and 8.1-3 to get Q. If all three Q’s are equal, the assumption was correct. If not, assume a new value of pwhand repeat the calculation procedure. Dougherty (1982) presented a modified calculation procedure. He defined the following quantities: (1) For tubing, .the flow constant B, ( n is the number of wells) is equal to:
(8 .I-4)
and B,, = n2B1(for n wells)
(8.1-5)
271
(2) For flowline, E is equal to: E=
(198.6)’d:
(8.1-6)
G,T,ZS f s 4
(3):
A, = A,
kh
1422( P
= nA,
1,
(for 1 well)
(8.1-7)
TR
(for n wells)
(8.1-8)
(4) Drainage radius is equal to: re,l =
JA77. (for 1well)
(8.1-9)
and
(8 .I-10)
re,n = r e J 6
if there are n wells in the drainage area, A ft’. (5):
K , = Xl/[ln(0.472re,l/rw)
+S]
(8 .I-11)
+ S ] (for n wells).
(8.1-12)
and
K,
= An/[ln(0.472re,,/rw)
(6): C=ej
(8.1-13)
Dougherty then derived his solution technique as follows: Using eqs. 8.1-1 and 8.1-12:
P i -P$
(8.1-14)
= Q/Kn
Using eqs. 8.1-2, 8.1-4, 8.1-5, and 8.1-13:
(),(
P5- CPih =
Q2
C-1
7)
(8.1-15)
(8.1-16)
272
Adding eqs. 8.1-14 and 8.1-15, one obtains:
pi-
cp&= Q/K,+ Q*(C-')/('n?)
(8.1-17)
Substituting eq. 8.1-16 into eq. 8.1-17:
or (8.1-18) Introducing new terms a,, b,, and c: a , = C / E + (C-l)/(?B,) b,
= lJK,
and c
= p i - Cp,2
equation 8.1-18 becomes : a n Q 2+ b,Q - c = 0
(8.1-19)
This is a quadratic equation: Q-
[ -b,f
ii6,2+]/2an
Inasmuch as negative Q is not physically possible,
\i( I
Q=
+)2
c
+
bn a, --'an
(8 3.20)
Thus, Q can be easily determined. It is only necessary to assume an average pressure for pZ product evaluation. Inasmuch as the pZ product for gases is almost constant over small pressure ranges, any assumption of Pwh will not affect the result significantly. If required, the calculations can be performed twice to give a better accuracy. This procedure is illustrated in Example 8.1-1.
273
Example 8.I-1 Calculate the deliverability of a gas field for n = 1, 10, and 20 wells if the pipeline inlet pressure is 1250 psia. The following information is given: A = 4000 acres; Gg= 0.75; r, = 3 in.; T, = 220°F; PR= 4500 psia; T, = 80°F; k = 50 md; I, = 10,000 ft; h = 20 ft; I, = 5280 ft; d , = 4 in.; d , = 2.5 in.; friction factor f = [2 log d + 6.53]-*; skin factor S = 2.0; and turbulence coefficient = 0. Solution: Assume pwr= 2500 psia and a common gas gradient of 0.08 psi/ft. For a depth of 10,000 ft, the static pressure difference between wellhead and bottomhole is equal to:
A p = (l0,000)(0.08) = 800 psi. Due to flow, an additional pressure drop due to friction is introduced, which is assumed to be 200 psi. Thus:
A p = 800 + 200 = 1000 psi and
pwh=pwr- A p = 2500 - 1000 = 1500 psia. For Gg= 0.75, ppc= 668 psia, Tpc= 406"R, TR= 220°F = 680"R * ( TPr),= 1.675,
r, = 540"R * (Tpr),= 1.330, pR= 4500 psia * ( ppr)R = 6.737, pwr= 2500 psia pwh = 1500 psia
-
( pPr)wf = 3.743, ( ppr),h= 2.25,
pd = 1250 psia * ( ppr)d= 1.87.
Using Fig. 8-21,
Z(TR, jj,) z(TR,
= 0.947,
p w f ) =o*845,
Z ( q , pwh)= 0.690
274
and,
Z ( T , , p d ) = 0.730. Using Fig. 8-23, p(TR, p R ) ~ 0 . 0 2 4 8CP
and, p ( T R , p,,) =0.020 cP. Thus: (0.0248)( 0.947)
+ (0.020) (0.845) = 0.0202 CP,
2
(pz)av=
z,= 0.845 +2 0.690 = o.7675, T a T=+ TR L - 610°R, 2
s^= 0.0375Gglt - (0.0375) (0.75) (10,000) = 0.6007, ‘ a ‘a (0.7675) (610)
f,= (2 log d, + 6.53)-2= (2 log 2.5 + 6.53)-, f,= (2 log 4 + 6.53)-2 B,
=
B,,
= 0.0186,
= 0.0167,
(198 .6),d:
(198.6),(2.5)’ = 58.9761, GgTaZafJ, (0.75) (610)( 0.7675) (0.0186) (10,000)
= (10),(58.9761) = 5897.61,
B2, = (20),(58.9761)
= 23,590.43,
(198.6),(4)’ E = (198.6)’d: G,T,Z,&J, (0.75)(540)(0.71)(0.0167)(5280)
A,
=
kh (50)(20) 1422(p Z)av TR (1422) (0.0202) (680)
A,,
= lOh, = 0.5120,
h,,
= 20A, = 1.024,
and
=
= 0.0512,
1592.92,
275 = /(4000)(43560)/~ = 7447.3
re,, =
d re,2o= 4re,,, =
K,
=
ft,
m = 2355.04 ft, and = 1665.3 ft, A1
ln(0.472re,l/rw)
+S
0.5120 K1o = ln(0.472 x 2355/0.25) Kzo= ln(0.472 X
0.0512 ln(0.472 x 7447.3/0.25)
+ 2 = 4.43 x 10-3,
+ 2 = 0.04923,
1.024 1665.3/0.25)
+ 2 = 0.10186.
Next, the gas flow rate Q for each of the three cases can be determined as shown below: Parameter
Number of wells n
0.0244 225.734 17,400,953
n = 10 0.00138 20.313 17,400,953
27,102.6
112,532.45
120,338.19
- 4625.7
- 7359.78
- 4080.38
22,477
105,173
116,258
n=l
a, = e'/E b, = 1/K,
+ ( eg - l ) / ( S , )
c = p i - eJpt
= 20
0.001203 9.817 17,400,953
SAMPLE PROBLEMS
(1) Fracture fluid having specific gravity of 1.05 is to be injected into three wells simultaneously as shown below. Pipe section AB is 4 in. in diameter and 500 ft long. Sections BC, BD and BE are each 1000 ft long with I.D. of 2.5 in. Friction losses at each wellhead are equal to 50 psi. There are 2 flanged globe valves in each one of the sections BC, BD, and BE. The depth to the midpoint of perforations in each of the wells is 2500 ft. The tubing is 2.5 in. I.D., and casing is 4.5-in. I.D., with c/d = lop4. The storage is open to atmosphere (Fig. 8-32). For a fracture fluid rate of 50 bbl/min and a pressure of 3000 psig against the sandface, determine the horsepower of the pump required at the fracture-fluid storage outlet. Assume viscosity of fracture fluid = 100 CP and the fracture fluid is pumped through the tubing-casing annulus.
276
-a
Frocture Fluid Storoge
C
4in
4 in
ID
Pump
41: 1
Formation
Fig. 8-32. Sample problem 1.
(2) A 45"API oil with solution gas/oil ratio R,=400 scf/bbl is to be transported at a flow rate Q = 12,000 bbl/D, through a 65-mile long, 10-in. I.D. pipeline. Pipe relative roughness, ~ / = d soil temperature T, = 50°F; a, = 6.5 X lo6 ft2/sec; inlet oil temperature = 105°F; k , = 0.5 Btu/hr-ft-OF; k , = 1.5 Btu/hr-ftOF; kin= 0.05 Btu/hr-ft-OF; insulation thickness = 1 in.; pipe O.D. = 5.375 in. The pipe is buried 10 ft below the surface. Determine: (a) The time required for flow to reach steady state. (b) The steady-state pipeline outlet oil temperature. (3) Gas is to be transported through a 4.5-in. I.D. 5-mile pipeline at 60 MMscf/D. The inlet pressure is 1500 psia, gas gravity Gg= 0.80, average flowing temperature = 75"F, friction factor f=O.O15. For a supply pressure of 800 psia, find the horsepower of a 90%-efficientpump that must be placed 3 miles upstream from the exit end. REFERENCES Barry, E.G., 1971. Pumping non-Newtonian waxy crude oils. J. Insr. Per., 57(554) : 74-85. Bed, C., 1946. The viscosity of air, water, natural gas, crude oil and its associated gases at oil-field temperatures and pressures. Trans. Soc. Per. Eng. AZME, 165 :94-115. Beggs, H.D. and Brill, J.P., 1972,An Experimental Study of Two-phase Flow in Inclined Pipes. 47th Annu. Fall Meet., Soc. Pet. Eng. AIME, San Antonio, Texas, Oct. 8-11, SPE 4007,13pp. Beggs, H.D. and Robinson, J.R., 1975.Estimating the viscosity of crude oil systems, JFT Forum.J . Per. Tech., 27(9) :1140-1141. Brill, J.P. and Beggs, H.D., 1974.Two Phase Flow in Pipes. Textbook for courses at the Univ. of Tulsa. Brown, G.G., 1945.A series of enthalpy-entropy charts for natural gases. Trans. Soc. Pet. Eng. AIME, 160:65-76.
277 Brown, G.G., Katz, D.L., Oberfell, G.B. and Alden, R.C., 1948. Natural Gasoline and Volatile Hydrocarbons. N.G.A.A., Gas Processors Assoc., Tulsa, Okla. Brown, K.E. and Beggs, H.D., 1973. The Technology of Artificial Lijt Methodr, Vol. 1. Pennwell, Tulsa, Okla., 487 pp. Campbell, J.M., 1974. Gus Conditioning and Processing. Campbell Petroleum Series, Norman, Okla., pp. 152-185. Carr, N.L., Kobayashi, R. and Burrows, D.B., 1954. Viscosity of hydrocarbon gases under pressure. Trans. SOC.Pet. Eng. AIME, 201 : 264-272. Chernikin, V.I., 1958. Pumping of Viscous and Congealing Oils. Gostoptekhizdat, Moscow, (in Russian). Chew, J. and Connally Jr., C.A., 1959. A viscosity correlation for gas-saturated crude oils. Trans. Soc. Pet. Eng. AIME, 216 :23-25. Colebrook, C.F., 1938-1939. Turbulent flow in pipes, with particular reference to the transition region below the smooth and rough pipe laws. J . Inst. Ciu. Eng. Pap. 5204, 11 : 133-156. Craft, B.C., Holden, W.R. and Graves, Jr., E.D., 1962. Well Design, Drilling and Production. Prentice-Hall, Englewood Cliffs, N.J., 571 pp. Cragoe, C.S., 1929. Miscellaneous Publication No. 97. U S . Bureau of Standards. Crane-U.S.A., 1942. Technical Paper No. 409. Crane-U.S.A., New York, N.Y. Davenport, T.C. and Conti, V.J., 1971. Heat transfer problems encountered in the handling of waxy crude oils in large pipelines. J . Inst. Pet., 57(555) : 147-164. Dougherty, E.L., 1982. Aduanced Reseruoir Engineering. Lectures at the University of Southern California, Los Angeles, Cal. Edmister, W.C., 1947-1949. Hydrocarbon absorption and fractionation process design. Petrol. Eng., May 1947 through March 1949. Edmister, W.C., 1961. Applied Hydrocarbon Thermodynamics, Vol. I. Gulf, Houston, Tex., 312 pp. Frick, T.C., 1962. Petroleum Production Handbook. Vol. 11. McGraw-Hill, New York, N.Y., pp. 19-39. Hansen, W.P., 1960. Time-saving chart for pipe wall thickness selection. Petro/Chem. Eng., 32(6) : 1-15. Heinze, F., 1971. Hydratbildung. Lehrbogen 3/3 von der Bergakademie Freiberg. Hydraulic Institute, 1979. Engineering Data Book. Hydraulic Inst., Cleveland, Ohio, 203 pp. Institute of Gas Technology, 1972. Steady Flow in Gas Pipelines. Technical Report No. 10. Kato, H., Nishiwaki, N. and Hirata, M., 1968. On the turbulent heat transfer by free convection from a vertical plate. Int. J. Heat Mass Transfer, l l ( 7 ) : 1117-1125. Katz, D.L., Cornell, D., Kobayashi, R., Poettmann, F.H., Vary, J.A., Elenbaas, J.R. and Weinaug, C.F., 1959. Handbook of Natural Gar Engineering. McGraw-Hill, New York, N.Y., 802 pp. Lee, A.L., Gonzalez, M.H. and Eakin, B.E., 1966. The viscosity of natural gases. Trans. Soc. Pet. Eng. AIME, 237 : 997-1000. Ludwig, E.E., 1977. Applied Process Design for Chemical and Petrochemical Plants, Vol. 1. Gulf, Houston, Tex., 371 pp. Makowski, M.M. and Mochlinski, K., 1956. An evaluation of two rapid methods of assessing the thermal resistivity of soil. Proc. Inst. Electr. Eng., Oct. Marks, A., 1978. Handbook of Pipeline Engineering Computations. PennWell, Tulsa, Okla., 347 pp. Moody, L.F., 1944. Friction factors for pipe flow. Trans. Am. SOC.Mech. Eng., 66:671-684. Perry, R.H. and Chilton, C.H. (Editors), 1973. Chemical Engineers' Handbook. McGraw-Hill, New York, N.Y. 5th Ed. Petroleum Extension Service, 1953. Oil Pipeline Construction and Maintenance. Univ. Texas, Austin, Tex., 193 pp. Rohsenow, W.M. and Hartnett, J.P., 1973. Handbook of Heat Transfer. McGraw-Hill, New York, N.Y., pp. 3-121. Standing, M.B. and Katz, D.L., 1942. Density of natural gases. Trans. SOC. Pet. Eng. AIME, 146: 140-149. Szilas, A.P., 1975. Production and Transport of Oil and Gas (Developments in Petroleum Science, 3) Elsevier, Amsterdam, 630 pp. Thomas, R. and Turner, W.C., 1953. Insulation for heat and cold. Chem. Eng., 60(6):222. Wylie, E.B., Streeter, V.L. and Stoner, M.A., 1972. Unsteady Natural Gas Calculations in Complex Piping ' systems. In: 47th Annu. Fall Meet., Soc. Pet. Eng. AIME, San Antonio, Tex., Oct. 8-11, SPE 4004. Zaba, J. and Doherty, W.T., 1956. Practical Petroleum Engineers' Handbook. Gulf, Houston, Tex. 4th ed.
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
279
Chapter 9
DESIGN OF FLOWING WELL SYSTEMS
SANJAY KUMAR, KERN H. GUPPY and GEORGE V. CHILINGARIAN INTRODUCTION
The design of production facilities necessitates familiarity with and an understanding of the basic concepts of flow in the overall system consisting of the reservoir, subsurface equipment, and the surface flow configuration. It is imperative that good judgement be exercised in designing these facilities to match the oil or gas well’s production in order not to underdesign or overdesign the system. Figure 9-1 is a schematic representation of the overall flow configuration in a typical well as fluid flows frqm the reservoir to the surface separator. Each segment of the flow configuration can be separated and treated individually. Equations can be derived to predict pressure drops in each segment. In general, the entire system is separated into the following flow segments: (1) Reservoir or porous fluid flow. (2) Vertical or directional flow in tubing or casing.
rCHOKE
SEPARATOR
i .
;AS
+
-.,-.
i c-FLOW ~HROUGHO POROUS MEDIUM
Fig. 9-1. The overall production system. (After Brown and Beggs, 1977, p. 68; courtesy of PennWell Publishing Company.)
280
(3) Horizontal or inclined flow in surface flowlines. (4) Restricted or choke flow. The objective in the overall design is to minimize pressure drops in each portion of the system. Hence, the type of flow, whether single- or two-phase flow, can have a significant impact on the design criteria. RESERVOIR FLUID FLOW
The ability of a reservoir to produce is influenced by several factors, such as reservoir permeability, reservoir pressure, and the type of drive mechanism. To predict this ability to flow, it is important that the relationship between flow rate and pressure be described accurately. The pressures normally used are the flowing bottomhole pressure and the average reservoir pressure. For any given flow rate, the smaller the difference between these pressures, the more efficient is the ability of the reservoir to produce fluids. To compare different wells with different drive mechanisms quantitatively, a parameter called the productivity index, J , is used, which is defined as follows:
(9-1)
J=q/(jR-Pwf)
where q = flow rate, bbl/D; jR = average reservoir pressure, psia; and pwf= flowing bottomhole pressure at the wellbore, psia. Productivity index J , commonly expressed as PI, can be based on total fluid production, or on individual oil, water, or gas production rate, as illustrated in Example 9-1. Example 9-1 Given: jR = 3000 psia, pwf= 2500 psia, q, and h = 20 ft.
= 200
bbl/D (bpd), water cut = 258,
Find: (1) J based on oil production, (2) J based upon total liquid production, and (3) Specific J for (1) and (2) above. Solution: Water cut = 0.25 = q,/( q, + qo). Hence, qo = 4,- q, = 3q, = 600 bbl/D and jR- pwf= 3000 - 2500 = 500 psi. 0.25 600 (1) J = qo/( jR - p W r )= -= 1.2 bbl/D/psi. 500 (2) = (40 + q w ) / ( P R - P w f ) = 6oo5002oo = 1.6 bbl/D/psi. (3) Specific J , based upon oil production, J,,, is equal to: +
J,,
=
J h
- = 1.2/20
= 0.06 bbl/D/psi-ft.
281
00
q+
Fig. 9-2. Typical inflow performance curves. (Modified after Brown and Beggs, 1977, p. 1; courtesy of PennWell Publishing Company.)
Specific Jst, based upon total liquid production, is equal to: 1.6 J,, = - = 0.08 bbl/D/psi-ft. 20 To properly design the 'correct production components, it is very important to predict the flowing bottomhole pressure for any given flow rate. It has been found that the drive mechanism in the reservoir has the greatest influence on this relationship, called the Inflow Performance Relationship ( I P R ) . It is shown in Fig. 9-2 for water-drive, gas-cap-drive and solution-gas-drive mechanisms. A quantitative measure of the I P R is the productivity index, the inverse slope of the IPR curve. For the water-drive mechanism, J is constant. For the gas-cap and solution-gas-cap drives, J is not constant and varies with flow rate as follows:
Figure 9-3 illustrates the J characteristics for the three different types of reservoirs. It should also be noted that a combination of drive mechanisms can exist in many reservoirs. In a newly discovered reservoir, the reservoir pressure is above the bubble point. Below the bubble point, gas and oil segregate forming a two-phase oil and gas mixture. Inasmuch as the IPR curve ranges from the maximum pressure (average reservoir pressure) to the minimum pressure of zero, the system exhibits a combination of linear J above the bubble point and a non-linear solution-gas-drive J below the bubble point. It is important to remember that the I P R curve represents the relationship between the flowing bottomhole pressure and flow rate at a given reservoir pressure. Thus, as the reservoir pressure changes, the IPR curve will become different. For the water-drive system, the slopes will be the same, but the actual flow rate and pressure values will be different. Whereas in the water-drive system, IPR can be accurately predicted by testing a well and assuming a linear relationship, in the case of the solution-gas-drive, IPR is more difficult to predict. In 1968, Vogel offered a technique for describing the IPR
282
'0
1.0 2.0 3.0 4.0 5.0 6.0 7.0 CUMULATIVE RECOVERY (MMSTBO)
Fig. 9-3. Relationship between productivity index, J , and recovery for different types of reservoirs. (In: Brown and Beggs, 1977, p. 2; courtesy of Shell Oil Company.)
curve for solution-gas-drive reservoirs. In this paper, using different P YT data from several reservoirs, he was able to dimensionally represent ZPR curve in a form shown in Fig. 9-4. It was also found that a nonlinear equation can be used to describe this relationship as follows: 2
qo/(qo)m,
(3 (3
= 1 - 0.2 r
- 0.8
-
(9-3)
where (qO)m, is estimated at pwr= 0. As in the case of water-drive system, testing a well to determine pwfcorresponding to a flow rate qo enables one to determine the IPR curve. In Vogel's equation, however, ( qo)ma'i is determined from substituting qo and pwfin eq. 9-3. Example 9-2 The following data were obtained from a well test: p R= 3000 psia, pwf= 2500 psia, qo = 500 STB/D. Find: (1) (40)m,, i.e.9 40 at P w r = O . (2) qo at pwf= 500 psia, using Vogel's method. (3) qo at pwf= 500 psia, assuming a constant J . Solution:
283
w
n
3 v)
v)
w n a
w
U
3 v) v)
w U
a
-0
0.40
0:20
PRODUCING RATE (q,/(qo)max),
0:60
0.80
l:oo
FRACTION OF MAXIMUM
Fig. 9-4. Inflow performance relationship for solution-gas-drive reservoirs. (After Vogel, 1968. fig. 5 . p. 85; courtesy of the Society of Petroleum Engineers of A.I.M.E.)
Using eq. 9-3:
--4o ( 40 )"ax
- 1 - 0.2(0.8333) - 0.8(0.8333)2 = 0.2778 40
500
= 1800 STB/D. . ' . ( q )max = 0.2778 = 0.2778 ~
Pwr 500 = = 0.1667 (2) r PR 3000
284 40
As before,-
= 1 - 0.2(0.1667)
- 0.8(0.1667)2= 0.9444.
(40 )mu
Thus: qo = 0.9444(q0),, (3) J = -
40
PR
-Pwf
=
= (0.9444)(1800) = 1700 STB/D.
500 (3000 - 2500)
= 1.O
bbl/D/psi.
Thus: qo = J( j R- p w r )= l(3000 - 500) = 2500 STB/D. One shortcoming of Vogel’s equation is the assumption that the well does not have a pressure drop around the wellbore due to the wellbore damage or to fractures. Standing (1971) extended Vogel’s work to account for the so-called “skin effect” in the vicinity of the wellbore. He defined a term known as the flow efficiency ( F E ) as follows:
where J, = actual productivity index, Ji = ideal productivity index, pkf = ideal bottomhole pressure, and pwr= actual bottomhole pressure ( = p i p- Ap,,,). Equation 9-4 can be redefined as folloisrs:
Equation 9-5 shows that if FE < 1, the well is damaged. If FE > 1, the well is stimulated. When FE = 1.0, Standing’s correlation becomes Vogel’s equation. Figure 9-5 shows Standing’s curves for various FE values. One parameter in eq. 9-5 that needs to be determined is Ap,,,. The simplest method used is to test the well and develop a Horner Plot in order to determine the skin effect, S . The Horner plot yields slope m , whch enables determination of Ap,,,, i.e., Ap,,, = 0.87 Sm. Details of this procedure are given in Chapter 10. Example 9-3 The following information is available from a well test: qo = 500 bbl/D, jR= 3000 psig, pWr= 2500 psig, and FE = 0.7. Find: qo at pwf= 1700 psig, when (a) FE = 0.7. (b) The well is reworked to yield FE = 1.0. (c) The well is fractured to yield FE = 1.3. Solution: Pwr - 2500 j R 3000
= 0.833
S8Z Fig. 9-5. IPR curves for damaged wells producing by solution-gas-drive. (After Standing, 1970, fig. 2, p. 1400; courtesy of the Society of Petroleum Engineers of A.I.M.E.)
From Fig. 9-5, at FE = 1.0, qo/(qo)kk-l = 0.19. Thus, (qo)cE1 = 500/0.19 bbl/D. For a pwr of 1700 psig:
= 2632
From Fig. 9-5: (a) For FE = 0.7, qo/(qo)Lzl = 0.47 Thus, qo = (0.47)(2632) = 1237 bbl/D. (b) For FE = 1.0, qo/(qo)k%=l = 0.65 Thus, qo = (0.65)(2632) = 1711 bbl/D. (c) For FE = 1.3, qo/(qo);L*l = 0.77 Hence, qo = (0.77)(2632) = 2026 bbl/D. The above example illustrates how to predict the flow rate at various values of pwr.Hence, as long as FE is known, an IPR curve can be developed for a particular well. In summary, methods have been shown in this section for predicting the relationship between flowing bottomhole pressure, pwf,and flow rate, qo. Once a relationship is determined, the IPR curve is drawn. In many cases, it may be required to design the tubing and flowline combination before the well is drilled. It
286
is useful to test an adjacent well in order to predict the inflow performance relationship for the new well. An accurate estimate must be made of the relationship between the flowing bottomhole pressure and flow rate for a given average reservoir pressure. VERTICAL FLOW
For oil reservoirs, the vertical flow in the tubing or casing requires accurate methods for predicting pressure drop from the bottom of the wellbore to the surface. Such calculations become very complicated when gas and oil are flowing together, e.g., as a result of flashing that may take place due to the large reductions in pressure as the fluid moves upward in the tubing. As in the case of single-phase flow in vertical columns, prediction of frictional loss in the case of two-phase flow requires estimating friction factors which are dependent on viscosity, density, and velocity of the fluids. In two-phase flow, viscosity and density are actually those of a mixture (liquid and gas). Determination of the properties of mixtures requires introduction of a new parameter called the liquid holdup factor, H L , defined as the volume fraction of liquid- in a vertical column. The holdup factor is usually determined from correlations based on experimental work. It depends on the flow pattern, gas and liquid velocities, and the pipe inclination. Frequently, it is taken as the no-slip holdup, A , which can be calculated directly from the flow rates (see Chapter 11). Vertical flow correlations
Various methods used for predicting pressure drops in vertical columns use different empirical correlations for determining H , and the friction factor for the two-phase mixtures. Following a pioneering paper by Poettmann and Carpenter (1952), considerable amount of research work has been done in this area. Most of these correlations differ only in (1) the way the liquid holdup is evaluated in the computation of density; (2) the handling of friction losses; and (3) the distinction made in flow regimes. Correlations presented by Hagedorn and Brown (1965) and Beggs and Brill (1973) are considered to be applicable over all velocity ranges of multiphase flow. Hagedorn and Brown used a 1500-ft deep experimental well to develop their correlation. Data was taken for liquids of varying viscosity using three different tubing sizes (1-2.5 in.). They used the general energy equation to obtain the equation for pressure loss in a two-phase system:
where A p
= pressure
drop in psi, through a vertical distance A h in ft; d = tubing
287
diameter, ft; w = mass flow rate, Ib,/D; V, = velocity of mixture, ft/sec; P, = pLHL &(l- H L ) = average mixture density, lb,/ft3; pm = density of the mixture at the reference point; g = gravitational acceleration = 32.2 ft/sec2; and g, = 32.2 lb, ft/lb, sec2. They represented the mixture viscosity by the relation proposed by Arrhenius. Thus, the Reynolds number, expressed in oilfield units for the two-phase flow becomes:
+
(9-7)
where pL= liquid viscosity, cP; and pg = gas viscosity, cP. Flow patterns were not considered. In the modified Hagedorn and Brown technique, however, the Griffith modification for bubble flow has been incorporated for use in the bubble flow regime. Another modification is in the use of mixture density. The larger of the two values, one calculated by using the Hagedorn and Brown holdup correlation and the other by assuming no-slip, is used. These modifications render this correlation applicable quite accurately over a wide range of flow conditions. Beggs and Brill (1973) conducted experiments on scaled-down versions of the real situation in the laboratory. They used 90-ft long pipe sections, 1 in. and 1.5 in. in diameter. The singular advantage offered by such a setup was that the same pipe could be manipulated at all angles, from horizontal to vertical. All other parameters remaining the same, the variation in flow characteristics under the influence of any one parameter could be studied. This correlation was developed primarily for directional or inclined flow. It is quite accurate for horizontal and vertical flow. The Duns and Ros (1963) method was developed through large-scale, carefully controlled laboratory data, suitably modified using field data. Their mist flow
Fig. 9-6. Vertical flow patterns. (After Duns and Ros, 1963; courtesy of Halliburton Services.)
N
m W
Fig. 9-7. Flow regime map. (After Duns and Ros. 1%3; courtesy ofWliburton Smites.)
289
correlation is the most widely accepted. An interesting aspect of this work is the introduction of the flow regimes (Fig. 9-6) and the flow regime map (Fig. 9-7). ~ NLv= liquid-velocity In Fig. 9-7, Ngv= gas-velocity number = Kg(~ , / g u ) ' / and number = V,,( ~ , / g a ) ' / ~ ; Kg= superficial gas velocity, ft/sec; V,, = superficial liquid velocity, ft/sec; pL = liquid density, lb,/ft3; and u = surface tension, lb,,,/sec2. The above-described correlations all require the use of complex programs and computers to accurately predict pressure drop, and details can be obtained from the original references. Working pressure traverse curves for vertical flow
To avoid the use of large programs and computers for individual wells, a more generalized approach has been made to predict pressure drops in vertical columns: traverse curves, which are plots of depth versus pressure for selected oil and gas properties at various gas/liquid ratios, are used. The most common traverse curves used are prepared by using correlations of Hagedorn and Brown and are presented in Figs. 9.1-1 through 9.1-16 in Appendix 9.1. These curves enable conversion of pressures into equivalent vertical lengths and vice versa. The technique of using the traverse curves can be described as follows: (1) Select the applicable curve for the given tubing size, production rate, and gas/liquid ratio. (2) Locate the known pressure on the pressure curves, go vertically down to the applicable gas/liquid ratio curve, and read off the depth on the vertical depth axis. (3) Correct this depth as follows: (a) Add the well depth to the depth value found in Step (2), if the known pressure was the surface pressure. (b) Subtract the well depth from the depth value found in Step (2), if the known pressure was the bottomhole pressure. (4) Read off the unknown pressure corresponding to the corrected depth. Example 9-4 serves to illustrate this procedure. Example 9-4 Given: qo = 800 bbl/D, G / O = 300 scf/bbl, z = 8000 ft, (linear), tubing size = 2.5 in. Find: the flowing wellhead pressure,
PR = 2800 psig, J = 1.0
Pwh.
Solution:
J = - qo P R
- Pwr
=1.0
Therefore, p w p= 2800 - 800 = 2000 psig. Using Fig. 9.1-9 for vertical flow, pwh at 1400 ft (= 9400 - 8000) is equal to 130 psig.
290 MULTIPHASE FLOW IN DIRECTIONAL WELLS
In the case of directional wells with deviations not exceeding 15-20' true vertical depth can be used along with the vertical multiphase flow correlation to ascertain the pressure traverse. This approximation, however, is invalid for deviations greater than 20°, because (1) a directional well has a greater length than a vertical well for the same depth, resulting in a greater frictional head loss, and (2) holdup differs and may be greater than that for vertical flow. Beggs and Brill (1973) introduced corrected holdup factors to account for directional flow. Their results, however, have not yet been tested sufficiently to be widely accepted. Ney (1968) presented two new solutions, whereas Fuentes (1968) extended his work. One of these solutions, which is presented here, combines the use of vertical flow and horizontal flow correlations of Hagedorn and Brown. First the pressure loss is calculated using only the true vertical depth in a vertical flow correlation. Then the frictional pressure drop due to the extra length of the tubing (i.e., total tubing length minus true vertical depth of tubing) is determined using a horizontal flow correlation. The sum of these two pressure losses is the total pressure loss for the deviated well. Ney (1968) and Fuentes (1968) have pointed out that this method works fairly well. Example 9-5 outlines this procedure. Example 9-5 In a directionally-drilled well, the true vertical depth is equal to 7000 ft; length of 2-in. tubing is equal to 9000 ft; Pwh = 100 psig; q = 1000 bbl/D (100% water); G / L = 800 scf/bbl. Determine the flowing bottomhole pressure, p w f .
Solution: Using the vertical flow correlation for a vertical depth of 7000 ft, p;, = 1760 psig. A trial and error procedure is required to determine p w f .As a first approximation, pwr= 1800 psig. Thus, the average pressure, j = ( pwh p w r ) / 2= (100 1800)/2 = 950 psig. On locating this average pressure ( jj = 950 psig) on the horizontal flow correlation chart in Fig. 9.1-19, and using additional length of 2000 ft (= 9000-7000), the downstream pressure is found to be 890 psig. The head loss due to friction in this extra 2000 ft of pipe, therefore, is equal to:
+
+
A p , = 950 - 890 = 60 psig Thus pwf= pLf + A p , = 1760 + 60 = 1820 psig. Second trial: Assuming pwf= 1820 psig, j = (100 + 1820)/2 = 960 psig. From the horizontal flow correlation, downstream pressure = 900 psig. Thus: A p , = 960-900 = 60 psig and pwf= 1760 + 60 = 1820 psig. Consequently, the second assumption was correct and pwr= 1820 psig.
291 HORIZONTAL FLOW IN SURFACE FLOWLINES
The main objective in designing flowlines is to choose a flowline size that will not cause significant back pressure on the well, restricting fluid flow from the well. Usually, the separator pressure is predetermined and it is necessary to determine the optimum wellhead pressure to produce at the allotted flow rate. Horizontal flow correlations As in the case of vertical flow, several correlations have been presented in the literature for determining two-phase pressure drop in the horizontal lines. Unlike the vertical flow, however, there is no elevation component. Only liquid holdup and friction loss parameters are necessary for characterizing horizontal flow. Lockhart and Martinelli (1949) were the first to present a correlation, which was determined from laboratory-scale data. They, however, neglected flow patterns and any acceleration. Thus, their method may result in large errors, especially in designing large-diameter pipes. Dukler et al. (1964) and Dukler (1969) collected laboratory and field data and used these to develop correlations for liquid holdup and friction factor. They studied two cases: (1) the case of no slip between phases and a homogeneous flow; and (2) the case where slip occurs, but it is assumed that the ratio of the velocity of each phase to the average velocity is constant. Flow patterns were not considered. Their friction factor correlation is one of the most accurate for horizontal flow. Eaton et al. (1967) developed correlations for friction factor and liquid holdup from extensive field studies under controlled conditions. Flow patterns were not considered. The liquid holdup correlation presented by them is very accurate and is frequently used along with Dukler’s friction factor correlation. Working pressure traverse curves for horizontal flow
The correlations cited are fairly complex and require the use of a computer to accurately calculate the pressure traverse. It is recommended that the correlations of Dukler’s Case I1 or the Eaton’s correlation be used if accurate predictions are needed. For reasonable results, the workmg pressure traverse curves prepared by Brown are sufficient. These curves are based on Eaton’s correlation and give satisfactory results except for low rates and low G / L ratios. Similar to the vertical flow curves, plots of pressure versus length of horizontal pipe have been prepared for various G / L ratios. It should be pointed out that these curves were prepared using water, but can be used interchangeably for oil, provided the free-gas/oil ratio is used for the G / L parameter. Horizontal flow pressure traverse curves are presented in Figs. 9.1-17 through 9.1-22. The steps involved in using them can be summarized as follows: (1) Select the curve for the given line size, flow rate, and gas/liquid ratio. (2) Enter the pressure axis using the known pressure and locate the length corresponding to this pressure on the correct G / L ratio curve.
292
(3) Correct this length for the pipeline length by: (a) Adding the pipeline length to the length in Step (2), if the known pressure is the outlet pressure, and (b) Subtracting the pipeline length from the length determined in Step (2), if the known pressure is the inlet pressure. (4) The unknown pressure is the pressure corresponding to this corrected length.
Example 9-6 A well is producing 800 bbl/D of oil with G/O=800 scf/bbl at a flowing wellhead pressure of 400 psig. Determine the separator pressure for a 2.5 in. ID, 9000-ft long line. Solution: Assume that at a pressure of 400 psig there is no gas in solution. Hence free-gas/oil ratio is 800 scf/bbl. Using Fig. 9.1-20 and the procedure described above: pSe, = 300 psig
INCLINED OR HILLY TERRAIN MULTIPHASE FLOW
Inclined flow implies flow through pipes that deviate from the horizontal, such as flow over hills, etc. Flanigan (1958) and Beggs and Brill (1973) presented some correlations. Flanigan’s method, however, is the only method available that can be applied to field problems without the use of complex computer programs. He calculated the effect of hills on pressure drop in pipelines by observing several field tests for various inclined flowlines at different flow rates, and concluded that most of the pressure drop occurred in the uphill section of the line. Flanigan defined two main pressure drop components that influence the two-phase flow in an inclined system and presented a method to determine each one of them: (1) Pressure drop due to friction, which is the predominant component in horizontal lines. (2) Pressure drop due to the liquid head, which is the predominant component in vertical and inclined flows. The sum of these two components determines the total pressure drop (Fig. 9-8). The uphill sections are treated as equivalent vertical columns containing an equivalent amount of liquid. Inasmuch as in two-phase flow the pipe is not completely filled with liquid, Flanigan introduced the term HF, which is the fraction of the total static pressure drop that exists as the elevation component. The pressure drop A p (in psi) due to elevation is determined by using the following equation: Ap=
PLHFXH
144
(9-8)
293
Gas flow rate
Fig. 9-8. Pressure drop components in two-phase flow. (After Flanigan, 1958; courtesy of the Oilund Gus Journal.)
where pL = liquid density, lb,/ft3; HF = elevation factor, dimensionless; and C H = the total uphill rises in the direction of flow, ft. The correlation between HF and the superficial gas velocity, V&, as determined by Flanigan, is as follows:
H,
=
1
Baker (1960) showed that for
H,
=
(9-9)
+
1 0.3264V,kOo6
Kg> 50, the applicable formula is:
0.00967( 1)1'2 v0.7 sg
(9-10)
where I = length of the flowline; and 31,194qgTF
"=
d'j(520)
(9-11)
Example 9-7 A flowline passes over 6 hills having the following vertical heights: 120 ft, 80 ft, 220 ft, 40 ft, 70 ft, and 180 ft. The flowline is 4 in. in diameter and 2000 ft long. qL = 6000 bbl/D (95% water); Gg= 0.7 (with respect to air); G, = 1.07; gravity of oil is 42" API; average pressure in line, j = 300 psia; and average temperature, 120°F. Find the pressure loss due to the hills if the gas/liquid ratio G/L = 200 scf/bbl.
r=
Solution: E : H = 1 2 0 + 8 0 + 2 2 0 + 4 0 + 7 0 + 1 8 0 = 7 1 0 ft. T= 120°F = 580"R, j = 300 psi, and Gg= 0.7.
294
Using Figs. 8-20 and 8-21 (Chapter 8), the compressibility factor, eq. 9-11: V
sg
qgz 7
= 31,194--
dzjj 520
=
(31194)(6000 X 200 X 10-6)(0.96)(580) (16) (300) (520)
z=0.96. From
= 8.35
ft/sec
Using eq. 9-9: H,
1
=
= 0.266,
+ (0.3264)(8.35)''w6 Go = (141.5)/(131.5 + OAPI) = (141.5)/(173.5) = 0.8156, G, = O.95Gw+ 0.05G0 = (0.95)(1.07) + (0.05)(0.8156) = 1.06, 1
and yL = (1.06)(62.4) lbf/ft3. Therefore, Aphills= (1.06)(62.4)(0.266)(710)/144
= 86.75
psi.
FLOW THROUGH CHOKES
All flowing wells utilize some kind of surface restriction, such as a choke, in order to regulate the flowing rate. Chokes serve many useful functions: (1) maintaining desirable flow rate; (2) maintaining sufficient back pressure to prevent sand entry; (3) protecting surface equipment; and (4) preventing gas or water coning. It is desirable to size a surface choke in a flowing well, so that flow through it is critical. Critical flow implies a flow where change in downstream pressure (such as separator pressure) does not affect the flow rate or the upstream pressure. This situation is obviously highly desirable in field operations. Critical flow is assumed to occur when the downstream pressure, p d , is approximately half of the upstream pressure, p u : Pu/Pd = 2
(9-12)
The generalized equation for critical two-phase flow through a choke is: (9-13) where qL = liquid flow rate, STB/D; pu = upstream pressure, psia; d i = inside diameter of choke, 6 4 t h ~in.; and R = producing gas/liquid ratio, scf/STB. Various investigators have proposed different values for a, b and c. Most commonly, however, Gilbert's (1954) correlation is used, where Q = 1.89, b = 10.0, and c = 0.546: qL = ( p u d ' . 8 9 ) / ( 1 0 R 0 . 5 4 6 )
(9-14)
295
Example 9-8 A reservoir having J = 1.0 and jTR = 2400 psig, is producing through 2.5-in. tubing, 5000-ft deep at a rate of qo = 1000 bbl/D with G / L = 600 scf/bbl. This well produces a large amount of sand when the oil production rate is above 1000 bbl/D; therefore, it is required to install a choke (“choke the well back”). Inasmuch as hydrate problems have made it impossible to install a surface choke, a bottomhole choke must be designed. It is proposed that the choke be installed at a depth of 4000 ft, i.e., 1000 ft above the bottom of the tubing. Determine: (a) The choke size required. (b) The flowing wellhead pressure, assuming that the flow through the choke is critical with p, = 2pd. Solution: (a) pWr= jTR - qo/J = 2400 - 1000/1 = 1400 psig. Using Fig. 9.1-10, the pressure at 1000 ft above the bottom of tubing is equal to p, = 1175 psig. The system is shown in Fig. 9-9. Using eq. 9-14: d!.89= l o q , ~ 0 . 5 4 6- (10)(1000)(600)0’546
1175
PU
= 279.8
in 6 4 t h ~in.
Thus, di = (279.8)1/1,89= 19.71 = 20/64 inches. (b) pd = i(1175) = 588 psig. Using the vertical flow correlation of Fig. 9.1-10, Pwh= 100 psig.
1175 p r i g
P,
= 1400psig
Fig. 9-9. Diagram of bottomhole choke for solving Example 9-8.
296
THE OVERALL PRODUCTION SYSTEM
Figure 9-1 illustrates the overall interconnected system. The inflow performance (l),vertical flow performance (2), surface flowlines (3), and chokes (4) correspond,
respectively, to (1) flow in the reservoir (porous medium), (2) subsurface flow up the tubing to the wellhead, (3) flow in surface lines, and (4) flow through the choke. Figure 9-10 shows the graphical representation of this system. Typically, (1) the pressure loss in the porous medium ( A p , =jR-pwf) is equal to 10-50% of the total loss; (2) the pressure loss in the vertical tubing string, A p , , is equal to 30-80% of the total loss; and (3) the pressure loss in the surface facilities, A p , , is equal to 5-30% of the total loss. The pressure versus flow rate plot shown in Fig. 9-11 exemplifies a plot used by an engineer in designing the production facilities for a given well. The procedure can be briefly described as follows (Brown and Beggs, 1977): (1) Plot the inflow performance curve. (2) Knowing the depth of the well, G/L,tubing diameter, etc., determine the values of the wellhead pressure corresponding to different flow rates and then plot. (3) Plot the surface choke performance curve for different flow rates. Sometimes, as shown in Fig. 9-10, a single curve is drawn corresponding to the flow rate desired.
tubing string
=
( P w f - PWh )
&@
----
s u r f a c e facil i t i e s
A PZ
I I
1
\
Fig. 9-10. Relationship between pressure and flow rate. (After Juch, 1967; courtesy of PennWell Publ. CO.)
297
0
I loo0
1
I
I
1
2000
3000
4000
SO00
WOO
TOTAL LIQUID FLOWRATE, qL, bbVD
Fig. 9-11. Tubing and flowline analysis for Example 9-9.
(4) The vertical line at the desired rate gives the values of pwf,4 , pwh,and choke size required for optimum performance. Example 9-9 below outlines the procedure used in the selection of correct combination of tubing and flowline sizes.
Example 9-9 A well is ready to be completed with several different tubing and flowline combinations. Determine the possible combinations, given the following information: J = 10.0, WOR = 1.0, length of tubing = 8000 ft, G/O = lo00 sc€/bbl, flowline length = 4000 ft, average tubing temperature = 150°F, average flowline temperature = 120°F, and Gg= 0.65. The well should not produce above 2000 bbl/D of liquid because of sand problems. The reservoir pressure is 3000 psig and the separator must be operated at 100 psig. The tubing and flowline sizes available are: (1) Tubing: 2 in. and 3 in. ID. (2) Flowline: 2 in., 2.5 in., 3 in., and 4 in. ID.
Solution: WOR (water/oil ratio) is 1 and the G/O ratio is 1000 scf/bbl. Hence: G/L ratio = 500 scf/bbl. J=10=
40
3000 - pwr
298
Therefore: 40 pwf= 3000 - 10
qL or pwf= 3000 - 20
(9-15)
where qL = qo + q, = 2q0 as given. Now, assuming various values for q L , one can obtain pwrusing eq. 9-15 above. Then the vertical correlation is used to determine the flowing wellhead pressure, pwr.Also using a separator pressure of 100 psig, one can determine the wellhead pressure, pwh,from the horizontal correlation; the flowline length is 4000 ft. The results obtained on following the above-described procedure are as follows: Vertical correlation qL
G/L
1000 1500 2000 3000
500 500 500 500 500
4Ooo
Pwr
pwhfor tubing size (ID) (in.)
(Psi@
2
3
2900 2850 2800 2700 2600
640 480 240 -
740 640 560 440
Horizontal correlation qL
600 1000 1500 2000 3000 4000 5000
G/J-
500 500 500 500 500 500 500
pwh for flowline size (ID) (in.)
2
2.5
3
220 350 540 720 -
160 200 300 400 600 -
250 370 520 -
-
4 -
160 190 240 280
The above data is plotted in Fig. 9-11. The intersections below q L = 2000 bbl/D give the possible combinations as follows: Tubing size (in.) 2 2 2 3
Flowline size (in.) 2 2.5 3 2
Total flow rate, q L (bbl/D) 1375 1750 2000 1850
Oil flow rate, q,, (bbl/D) 687.5 875 1000 925
299
SAMPLE PROBLEMS
(1) In a solution-gas-drive reservoir,
PR = 3500 psi,
FE = 1, pwf= 2800 psi, and
qo = 750 bbl/D.
(a) Determine the maximum oil production rate at a reservoir pressure of 3500 psi. (b) Find qo when pwrdrops to 1800 psi. (2) An oil well gave the following pressure response on January 2, 1982: Rate, STB/D: pwf, psis:
500 2600
1000 2040
1450 1500
The jR was 3000 psia on January 2, 1982. On January 2, 1983, a new test was run and it was found that jR = 2550 psia, and that qo = 600 bbl/D for a pwfof 1620 psia. With respect to the new conditions: (a) Determine ( qo),,,= (Hint: Plot qo versus drawdown on a log-log paper). (b) Determine qo for a drawdown of 1500 psia using the graph drawn in part (a). (c) Repeat (a) and (b) using Vogel's technique. Assume FE = 1. (3) In a 2.5411. ID tubing, a well is making 200 bbl/D of water having specific gravity of 1.08; G / L = 500. The specific gravity of gas is 0.65 and the average temperature is 120°F. Assume no slippage and that no gas goes into solution (because only water is being produced). At a pressure of 500 psia, find: (a) The no-slip holdup, A. (b) The (i) gas, (ii) liquid, and (iii) mixture velocities. (4) Given: 2.5-in. flowline; 3-in tubing; separator pressure = 200 psig; wellhead pressure = 650 psig; flowline length = 10,000 f t ; tubing length (depth) = 4000 ft; G / L = 1500 scf/bbl; and 100% water. Determine: (a) the flow rate possible in the flowline (assume all water) and (b) the flowing bottomhole pressure. (5) Given: depth = 8000 ft, jR= 2500 psig, G/O = 350, pwh= 120 psig (100% oil production). A safety valve was installed in this well at 3000 ft from the surface. A flow test conducted later showed that J = 5.0 (assume linear relationship). If the tubing is 4 in. in diameter and qo at the time of the test was 3000 bbl/D, is the valve partially closed or not? If so, what is the pressure drop across the valve? Assume qo is constant before and after valve installation (this is a simplifying assumption and is not necessarily true). (6) Given: length of pipeline 1 = 5000 ft, elevation of hills = 700 ft, upstream pressure pup= 400 psi, downstream pressure pdn= 150 psi, qw = 3000 bbl/D (all water), G, = 1.10, G / L = 800 scf/bbl, T = 120"F, and Gg= 0.65. Design the necessary flowline over the hills. (7) A production system is overdesigned in such a way that the separator pressure of 100 psig cannot be maintained. The system consists of a 2.5-in. tubing having a length of 5000 ft. The flowline is 2.5 in. in diameter and 4000 f t long. The well produces 100%oil having 35" API gravity with GOR = 500 scf/bbl. The reservoir has an active water drive and consists of three separate layers having
300
permeabilities of 310 md, 80 md, and 100 md. The net pay of these layers is 20 ft, 30 ft, and 80 ft, respectively. Reservoir pressure = 3000 psig, wellbore diameter = 6 in., B, = 1.22, and po = 8 cP. A 40-acre spacing is used. It is proposed to install a surface choke. Find (a) the production rate for critical flow across choke and (b) the necessary choke size. (8) An oilwell stops flowing due to a low reservoir pressure. Tubing size = 2 in., water cut = 0.5, wellhead pressure = 160 psig, G/O = 200 scf/bbl, and tubing length = 8000 ft. It is required to produce 400 bbl/D of oil. An orifice is located in the tubing at a depth of 5000 ft (from the surface). It is possible to inject gas through the casing at high pressures. Liner size = 8 in., spacing = 60-acre, net pay = 95 ft, k, = 10 md, k, = 200 md, B, = 1.04, B, = 1.15, p, = 1.0 cP, p, = 8 cP, and reservoir pressure = 2700 psig (maintained by a water drive). What is the required gas flow rate through the casing? APPENDIX 9.1-HAGEDORN AND BROWN (1965) PRESSURE TRAVERSE CURVES. CORRELATION AMONG PRESSURE, LENGTH OF PIPE, GAS/LIQUID RATIO, AND VERTICAL OR HORIZONTAL FLOWING PRESSURE GRADIENT. (FROM TUBING SIZE SELECTION, BY HALLIBURTON ENERGY INSTITUTE, 1976; COURTESY OF HALLIBURTON 'SERVICES, DUNCAN, OKLA.).
See Fig. 9.1-1 through Fig. 9.1-22 (pp. 301-322).
301
0
1
2
3
4
8
4
B
PRESSUREIn 100 pslg 12 16
H
1-
r5
i
!!i 6
7
0
. .
Q
8
8
%
9 . .i
..
~
I
10
~
itt, VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2 in: ID
600 B/D
G& specific Gravity Average Flowing Temp.
35"API
Fig. 9.1-1. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 14tP F
302
0
4
8
PRESSURE in 100 psig 12 16
20
24
28
1
2
3
4
5
B
e 5
I
i Y
-1
I
I
T
I
VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2 in. ID 1000 B/D 35"API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-2. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
303
0
4
PRESSURE in 100 psig 12 16
20
% / $%%
24
28
%%
VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2 in. ID 1200 BID 35" API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-3. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
%2
304
1
2
3
4
Y
e-
--
c5 I
PRESSURE in 100 sig 12 18
4
0
.
I
43
20
24
28
1 1
6
.. I
7
8
9
10
VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2 in. ID 1500 Bf D 35" API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-4. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
305
0
4
8
PRESSURE in 100 psig 12 16
20
24
28
i
3
4
I
-i
r 5
E
Pw &
6
0
7
*? 6
8
B 9
% 10
%&+.g%%%%
%
%
VERTICAL FLOWING PRESSUREGRADIENTS (50% OIL-%% WATER) Tubing Size Producing Rate Oil API Gravity
2 in. ID
600 BID 35"API
Water Specihc Gravity Gas Spedfic Gravity Average Flowing Temp.
Fig. 9.1-5. (Reprinted with permission of Halliburton Company; all rights reserved.)
1.074 0.65
140°F
306
0
4
PRESSURE in 100 psig
1
2
3
4
H
i.-
c5 I
k2
6
7
8
9
10
xg$%%%% % VERTICAL FLOWING PRESSUREGRADIENTS (50% 01L-50% WATER) Tubing Size Producing Rate Oil API Gravity
2 in. ID 800 BID 35"API
Water Specific Gravity Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-6. (Reprinted with permission of Halliburton Company; all rights reserved.)
1.074 0.65 140" F
307
0
4
8
PRESSURE in 100 psig 12 16
20
24
28
~-u-v-v-
VERTICAL FLOWING PRESSUREGRADIENTS (50% OIL-50% WATER) Tubing Sire Producing Rate Oil API Gravity
2 in. ID 1000BID 35" API
Water Specific Gravity Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-7. (Reprinted with permission of Halliburton Company; all r i a t s reserved.)
1.074 0.65 140" F
a
0
1
2
3
4
Y
@-
c5
i!
6
7
8
9
10
-%vwe" " "
v
v
"
VERTICAL FLOWING PRESSUREGRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID 600 BID 35"API
Gas Specific Gravity Average Flowing Temp.
Fig. Fig.9.1-8. (Reprinted with permission of Halliburton Company: all rights reserved.)
0.65 140" F
309
0
4
8
PRESSURE in 100 pslg 12 16
20
28
24
1
2
3
4
P
ia 4
E5
Y
6
7
8
9
10
%&$+g%% % %
%
%
%
VERTICAL FLOWING PRESSUREGRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID 800 B/D 35"API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-9. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.85 140" F
310
0
4
8
PRESSURE in 100 psig 12 16
20
24
28
1
2
3
4
-H
li5
c5 I
6
7
8
1
9
10
I VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID 1000 B/D 35"API
Gas Specific Gravity Average FlowingTemp.
Fig. 9.1-10. (Reprinted with permission of Halliburton Company: all rights reserved.)
0.65 140°F
311
>$%3g%Qo
%
%
%
VERTICAL FLOWING PRESSUREGRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID 1500 BID 35"API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-11.(Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
312
0
2
B 4
8
PRESSURE In 100 prig 12 16
20
24
28
VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size 2.5 in. ID Gas Specific Gravity 1200 BID Average Flowing Temp. Producing Rate Oil API Gravity 35" API Fig. 9.1-12. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
313
0
4
8
PRESSURE in 100 psig 12 16
24
20
28
, .
I
.
.- -t 1
I
1
2
3
4
H
B
r 5 I
Ii
3
6
7
8
9
10
%%%% % % VERTICAL FLOWING PRESSURE GRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID 2OOO B/D 35" API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-13.(Reprinted with permission of Halliburton Company; all rights resewed.)
0.65 140" F
314
0
4
8
PRESSURE in 100 pslg 12 16
20
24
28
1
2
3
4
I
1
.E 5 I
b
z
6
7
8
9
10
?$gg$% %% %
'ib
% I
VERTICAL FLOWING PRESSUREGRADIENTS (50% OIL--509'0 WATER) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID
600 B/D 35" API
Water Specific Gravity Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-14. (Reprinted with permission of Halliburton Company; all rights reserved.)
1.074 0.65 140" F
315
PRESSURE In 100 p i g 0
4
8
12
16
20
24
28
.
'-t
-4-
1
t--
. ..
-
r
2
3
4
5
i
c5
I
b Y
6
7
8
9
10
VERTICAL FLOWINGPRESSURE GRADIENTS (50% 011--50% WATER) Tubing Size Producing Rate Oil API Gravity
2.5 in. ID
800 BID 3.5"API
Water Specific Gravity Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-15. (Reprinted with permission of Halliburton Company; all rights reserved.)
1.074 0.65 1400 F
a
0
1
2
3
4
6
7
8
9
10
VERTICAL FLOWING PRESSUREGRADIENTS (ALL OIL) Tubing Size Producing Rate Oil API Gravity
3 in. ID 600 B/D 35"API
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-16. (Reprinted with permission of Halliburton Company; all rights reserved.)
0.65 140" F
317
0
2
4
PRESSURE In 100 prig 6 8
10
12
14
2
4
6
a
Y
s
-r l a
k!I
12
14
16
18
20
Flowline Size 2in. ID Gas k f i c Gravity Producing Rate 600 BID Average Flowing Temp. Water SpecificGravity 1.07 Fig. 9.1-17. (Reprinted with permission of Halliburton Company; all rights reserved.)
.065 120°F
318
0
2
4
PRESSURE in 100 psig 6 8
10
12
Fig. 9.1-18. (Reprinted with permission of Halliburton Company; all rights reserved.)
14
319
2
0
4
PRESSURE In 100 pslg 6 8
10
12
14
2
4
6
8
H
r
8-
c 10
4 w
4
12
14
16
. -._. .
.~.
.
..
*
.
>
^
18 .
-
,
-
,
.
.
,..
.
20
HORIZONTALFLOWING PRESSURE GRADIENT [ALL WATER) Flowline Size Producing Rate Water Specific Gravity
2in. ID loo0 B/D 1.07
Gas Specific Gravity Average Flowing Temp.
Fig. 9.1-19. (Reprinted with permission of Halliburton Company; all rights reserved.)
.065 120" F
320
%%%%%%% $
%
%
HORIZONTALFLOWING PRESSUREGRADIENT (ALL WATER)
Flowline Size Producing Rate Water Specific Gravity
2.5 in. ID 800 BID 1.07
% %
Gas Specific Gravity Average flowing Temp.
Fig. 9.1-20. (Reprinted with permission of Hdliburton Company;all rights reserved.)
.065 120" F
"
U
"
U
"
= . a
-0
HORIZONTAL FLOWING PRESSUREGRADIENT (ALL WATER) Flowline Size ProducingRate Water Speclflc Gravity
2.5 in. ID 1500 BID 1.07
Gas Speciri Gravity Average Flowing Temp.
Fig. 9.1-21. (Reprinted with permission of Halliburton Company; all rights reserved.)
.m5 140" F
322
0
4
2
PRESSURE in 100 paig 6 8
10
12
14
2
4
6
8
I! '-c
10
E
0
3 12
14
16
18
20
% % % % % HORIZONTAL FLOWING PRESSURE GRADIENT (ALL WATER) FlowlineSize Producing Rate Water Specific Gravity
2.5 in. ID 2000 BID
Gas Specific Gravity Average Flowing Temp.
1.07
Fig. 9.1-22. (Reprinted with permission of Halliburton Company; all rights reserved.)
,065
140°F
323 APPENDIX 9.11-INTRODUCTION
TO CHOKES
The purpose of a choke is to provide precise control of wellhead flow rates in surface production applications involving oil, gas, and enhanced recovery. A choke is a restriction in a flowline that causes a pressure drop or reduces the rate of flow through an orifice. Chokes are capable of causing large pressure drops. For example, gas can enter a choke at 5000 psi and exit at 2000 psi or less. The use of the choke as a control device has found many applications in the petroleum industry. Typically, in a flowing well, the choke is used to maintain a back pressure in the reservoir while allowing an optimum flow of gas or oil. Such control is often necessary to ensure cost effective production over the life of the well. There are two types of chokes that are commonly available, fixed and adjustable. Figure 9.11-1 shows a cross-sectional view of a fixed or positive choke. The pressure drop of the choke is determined by the flow of the medium through the internal diameter of a fixed orifice, often called a bean. The fixed bean choke is generally used where the flow conditions do not change over a period of time, because the changing of the bean requires a shutdown of the flow through the choke. Adjustable chokes are used where there is an anticipated need to change the flow rate periodically. There are' several types of adjustable chokes with each design offering several features. One of the varieties of adjustable chokes is the needle and seat type as shown in Fig. 9.11-2. The pressure drop of this choke design occurs as the flow is restricted through the area between the seat and the needle portion of the stem. The size or area of the opening is increased as the needle is moved farther away from the seat. This allows a change in the flow rate without shutting in the well.
I Fig. 9.11-1. Positive choke. (Courtesy of S.I.I. Willis, Long Beach, Calif., and Mr. Rick Floyd.)
324
Fig. 9.11-2. Needle and seat of a choke. (Courtesy of S.I.I. Willis, Long Beach, Calif.. and Mr. Pav Grewal.)
Fig. 9.11-3. Multiple orifice valve. (Courtesy of S.I.I. Willis, Long Beach, Calif., and Mr. Matthew L. Philippe.)
325
CLOSED
PARTIALLY OPEN
FULLY OPEN
Fig. 9.11-4. Disc plate-multiple orifice choke. (Courtesy of S.I.I. Willis, Long Beach, California.)
Another type of adjustable choke is the Multiple Orifice Valve (MOV), as shown in Fig. 9.11-3. This design uses two flat discs (Fig. 9.11-4) to control the flow. There are two holes in each disc such that as one disc turns in relation to the other, the area of the opening vanes. Fixed and adjustable chokes are used in a variety of applications with surface production equipment. When chokes are used for oil production, the major difference is the absence of a heater. Chokes are also used to control the rate of flow in enhanced oil recovery applications where fluids and gases are injected into the reservoir. Injection is used to maintain reservoir pressure and an economic rate of production. This process can be applied to flowing and pumping wells to improve the rate of recovery. Because chokes must operate in a wide variety of corrosive and harsh environments, components need to be constructed of materials designed to provide maximum performance, such as stainless steel and tungsten carbide. Typically, chokes can be customized to meet the requirements of any specific application. Trim material and sizes, body materials and seal materials can all be selected to provide a cost effective approach to controlling the rate of production.
326
In addition to tailoring a choke to perform in specific environments, an adjustable choke can be fitted with an actuator for remote operation. Actuators can be powered by electrical, hydraulic, or pneumatic systems and are normally used when the application involves frequent changes in production rates. The use of actuated chokes is increasing as computers are installed to manage production more efficien tly. REFERENCES Baker, O., 1960. Designing pipelines for simultaneous flow of oil and gas. Pipeline Eng., Handbook Section, Feb.: 67-80. Beggs, H.D. and Brill, J.P., 1973. A study of two-phase flow in inclined pipes. J. Pet. Tech., 25(5): 607-617. Brown, K.E. and Beggs, H.D., 1977. The Technology of Artificial Lift Methoh, Vol. 1. PennWell, Tulsa, Okla., 487 pp. Dukler, A.E., 1969. Gas-Liquid Flow in Pipelines. Vol. I, “Research Results”. Am. Gas Assoc., Am. Pet. Inst., May. Dukler, A.E., Wicks, M. and Cleveland, R.G., 1964. Frictional pressure drop in two-phase flow: a comparison of existing correlations for pressure loss and holdup, B-an approach through similarity analysis. Am. Inst. Chem. Eng. J., lO(1): 38-51. Duns Jr., H. and Ros, N.C.J., 1963. Vertical flow of gas and liquid mixtures in wells. Proc. Sixth World Per. Congr., June 1963, Sect. 11, Pap. 22-PD6. Earlougher Jr., R.C., 1977. Advances in Well Test Analysis. Monograph Vol. 5 Henry L. Doherty Series. Soc. Pet. Eng. A.I.M.E., Dallas, Tex., 264 pp. Eaton, B.A., Andrews, D.E., Knowles, C.R., Silberberg, I.H. and Brown, K.E., 1967. The prediction of flow patterns, liquid holdup and pressure losses occumng during continuous two-phase flow in horizontal pipelines. Trans. SOC.Per. Eng. A.I.M.E., 240: 815-828. Flanigan, O., 1958. Effect of uphill flow on pressure drop in design of two-phase gathering systems. Oil Gas J., 56(10): 132-141. Fuentes, A.J., 1968. A Study of the Mulfiphase Flow Phenomena in the Direcfional Well. Thesis. Univ. Tulsa, Tulsa, Okla. Gilbert, W.E., 1954. Flowing and gas-lift well performance. Drill. Prod. Pracr., A.P.I., p. 143. Hagedorn, A.R. and Brown, K.E., 1965. Experimental study of pressure gradients occumng during continuous two-phase flow in small-diameter vertical conduits. J. Pet. Tech., 17(4): 475-484. Halliburton Energy Institute, 1976. Tubing Sire Selection. Halliburton Services, Duncan, Okla., 57 pp. Juch, A.H., 1967. Natural Flow and Gas Lift. Oil Production Methods Course at the Zulia Univ., Maracaibo, March-June. Lockhart, R.W. and Martinelli, R.C., 1949. Proposed correlation of data for isothermal two-phase, two-component flow in pipes. Chem. Eng. Progr., 45(1): 39-48. Ney, C., 1968. A Laboratory Investigation of Holdup and Pressure Loss in Directional Multiphase Flow. Thesis. Univ. Tulsa, Tulsa, Okla. Poettmann, F.H. and Carpenter, P.G., 1952. The rnultiphase flow of gas, oil and water through vertical flow strings with applications to the design of gas lift installations. DrilJ. Prod. Pract., A.P.Z., p. 257. Standing, M.B., 1970. Inflow performance relationships for damaged wells producing by solution gas drive. JFT Forum. J. Per. Tech., 22(11): 1399-1400. Standing, M.B., 1971. Concerning the calculation of inflow performance of wells producing from solution gas drive reservoirs. J. Per. Tech., 23(9): 1141-1142. Vogel, J.V., 1968. Inflow performance relationship for solution gas drive wells. J. Pet. Tech., 20(1): 83-93. Wylie, M.R.T., Gregory, A.R. and Gardner, L.W., 1956. Elastic wave velocities in homogeneous and porous media. Geophysics, 21(1): 41-70.
327
Chapter 10
WELL TESTING SANJAY KUMAR AND GEORGE V. CHILINGARIAN INTRODUCTION
Well testing and analysis is a well established procedure and there are various tests available for the evaluation of a formation. The interpretation of these tests is a sophisticated and highly advanced subject. Fundamental understanding of the procedures for the various types of tests, mathematical interpretations, and knowledge of limitations of tests are of utmost importance. The various tests on a flowing oil well can be summarized as follows: (1) Drillstem test (2) Drawdown test (3) Buildup test (4) Isochronal test and modified isochronal test (for gas wells) (5) Flow-after-flow tests (6) Special tests (a) Interference tests for horizontal anisotropy (b) Pulse tests (c) Vertical interference tests (d) Injection and falloff tests. DRILLSTEM TESTING
Drillstem test (DST) is the most significant test available for the qualitative and quantitative evaluation of a formation, because it simulates the conditions of a completed well (see Lynch, 1962). It can be used (1)to determine the static reservoir pressure; (2) to determine formation parameters like average permeability, and extent of permeability damage around the wellbore due to drilling; and (3) to obtain representative samples of the formation fluids. A drillstem test is usually run when drilling has reached total depth and formation of interest has been evaluated by samples, logs, and cores. This test essentially consists of lowering a packer and a length of perforated tail pipe (coupled to the end of the drillpipe) with valve to the level of the formation (see Fig. 10-1). Setting the packer against the borehole wall seals off the test interval from the drilling fluid column above. The valve is then opened, reducing the pressure in the zone opposite the formation. Thus, the formation fluids can flow
328
Fig. 10-1.A conventional straddle drillstem test tool assembly. (Courtesy of Lynes Inc., Houston, Tex.)
329
into the borehole and up through the drillpipe. In effect, this is a temporary well completion and the produced fluids represent the well production upon final completion. By the addition of specialized devices to the test tool string, the test has been improved with time and is capable of giving more information. Additional devices in use include pressure recorders, the shut-in valve, and the choke. Other auxiliary components are the disk valve, the safety joints, the hydraulic jar and the surface control head. The arrangement of various tools for drillstem testing are presented in Fig. 10-1. (See Lynch, 1962, for greater details.)
Component parts of a conventional drillstem tester Anchor shoe The anchor shoe supports the weight of the drillstem and the mud column. It is generally made from heavy drillcollar stock and thus has a greater wall thickness. Perforated anchor pipe The perforated anchor pipe supports the weight of the drillstem and mud column. It aids in setting the packer by holding the bottom part stationary, while weight is applied to the upper part of the packer. The anchor pipe also serves as a flow passage and allows drilling fluid communication through the perforations between the open hole and the inside of the drillpipe. It is made from the same material as the anchor shoe. The anchor shoe, attached to the bottom of the perforated anchor pipe, should rest on the actual hole bottom. It should not rest on loose debris in the hole, because failure of the anchor shoe (or pipe) will cause the packer to slip, resulting in test failure. The anchor pipe can also be used when the test is run far above the hole bottom, as is the case in a straddle DST. It is also possible in such cases to use a set of dogs (or slips) below the lower packer. These slips provide enough support to set the packers. An equalizing line is used to connect the mud columns above and below the packers and equalize the hydrostatic pressure in these two zones. Pressure recorders The bottomhole pressure recorders are sophisticated versions of the bourdon type or the spring piston type of pressure gauges. The variation of pressure with time may be digitally recorded at the surface, or on cylindrical charts. The recorders are usually placed inside a recorder carrier to protect them in the wellbore. In order to safeguard against failure and for comparison purposes, it is necessary to use at least two pressure recorders. Generally, the following are used: (1) Below-straddle recorder. This recorder, placed below the bottom packer, can indicate whether the packer has effectively sealed the bottom section of the wellbore from the formation being evaluated. If the bottom packer holds, this recorder will show hydrostatic pressure of the mud column above it. If the bottom packer does
330
Fig. 10-2. A conventional packer assembly. (Courtesy of Lynes Inc.. Houston, Tex.)
331
not hold, this recorder will indicate almost the same pressure as the recorders placed in the potential production zone. (2) Inside and outside recorders in the producing interval. The outside recorders are exposed directly to the formation pressure. Usually two such recorders are used in the test interval to safeguard against failure and also for checking and comparing the results. Pressures reported by the outside recorders are also compared to those recorded by the inside recorder, which is placed above (Fig. 10-1). When the formation fluids enter the perforations and flow up inside the test tool assembly (through the packers, safety joint, hydraulic jars, and past the inside recorder on their way up to the drillstring), the inside recorder is exposed to the formation pressure minus the hydrostatic fluid head between the formation and the inside recorder. A comparison between the pressures recorded by the inside and outside recorders can thus reveal plugging or flow obstruction. Packers Packers serve as sealing devices. A conventional packer consists of a 20-60 in. (or more) long rubber element mounted on a steel mandrel. The rubber is vulcanized to two steel heads that are free to move relative to each other. The packer can be set by allowing a part of the drillpipe weight to act on the upper head, keeping the bottom head supported in a fixed position. This causes the rubber to be compressed and squeezed against the borehole wall. A good testing procedure is to use a rubber size within about 1 in. (25 mm) of the size of the borehole which has to be sealed. Many variations of this basic design have been introduced over the years. One such design is shown in Fig. 10-2. In this design, right-hand rotation of the drillstem activates a downhole pump, which utilizes the annular fluid to inflate all the packers in the test assembly simultaneously. Typically, the packer is inflated to approximately 1700 psi above the hydrostatic head of the drilling fluid. (See Fig. 10-3.) Equalizing valve The purpose of the equalizing valve, which is normally open, is to allow drilling
OUTER COVER STEEL BRAID
INNER BLADDER
Fig. 10-3.Cross-section of an expanded packer. (Courtesy of Lynes Inc., Houston, Tex.)
332
fluid to bypass the packer through the inside of the drillpipe. It is closed only during the test when the tester valve is opened. By opening the equalizing valve, the pressure above and below the packer is equalized. This valve also relieves the pressuring action when the packer is run into the hole and the swabbing action when it is pulled out. Tester valve Tester valve is the main valve essential to the drillstem test. This valve, which is normally closed, controls the flow of the formation fluids into the anchor pipe and then into the drillpipe. It is opened by supplying the weight of the drillpipe string, as in the case of setting the packer. Thus, a mechanical or hydraulic delay system is required to delay the opening of the valve until the packer has been set. At the completion of the test, an upward pull on the drillpipe relieves the weight on this valve, closing it automatically. Choke The choke is a small flow restriction placed near the main valve to control fluid flow rate from the test zone. The pressure changes are thus made more. gradual, which protects the packer and other elements from the pressure shock caused by suddenly opening the tester valve. The choke also maintains a back pressure against the formation face, which has a sand control effect. An accurate average production rate can be maintained during the test period by choosing the proper size of choke. Commonly used sizes range from 3/16 to 3/8 in. Shut-in valve Shut-in valve is designed to close by the rotation of the pipe so that the shut-in pressure at the end of the test can be recorded. It is located above the main tester valve and is sometimes combined with the circulating valve in a single unit. Circulating valve The purpose of the circulating valve is to allow drilling fluid circulation in order to remove the combustible test fluids from the drillpipe before withdrawing it from the hole. This valve connects the drillpipe and the annulus a short distance above the tester. Drilling fluid pumped down the annulus can then return through the drillpipe, carrying the test fluids ahead of it. A combination of circulating valve and shut-in valve is in common use. This makes it possible to shut the shut-in valve and simultaneously open the circulating valve simply by rotating the drillpipe through a specified number of turns. Other components Other minor components of the DST tool assembly include: Safety joint-to enable retrieval of the top section of the tool if it becomes stuck in the hole.
333
Hydraulic jars-to impart sharp impact blows to the drillstring if the test string and drillstem become stuck in the hole. Hydraulic valve-to keep the drillstem dry while the tester is lowered into the wellbore. Fluid sampler-to trap a sample of the formation fluid under flowing reservoir conditions at the end of the last flow period. Drillstem test procedure
Inasmuch as the drillstem test is a very hazardous operation, it is necessary to observe the utmost care and caution (see Gatlin, 1960, p. 253). Before beginning the test, the borehole and drilling fluid must be in good condition. In addition, drilling fluid should be circulated to remove all cuttings and the circulation should be continued until the test is started. The drilling fluid density must be carefully measured in order to check the pressure that is indicated by the pressure recorders. The choice of tools is dictated by the well size and depth, depth interval of test zone, test duration, drilling fluid density, number and type of packers and pressure recorders, and choke size. The assembled length of the tool assembly must be determined for accurate setting in the hole. The packer should be set along a hard, consolidated section of the formation, because shales and unconsolidated sands may collapse (flow under) under the pressures involved, leading to packer failure. A good packer seal against the borehole wall is essential to the success of drillstem test. Blowout preventers should be checked, as well as the standby equipment, which is installed for use in case of test failure. During the drillstem test, the tools should be lowered into the hole at a safe speed, usually 75% of the normal value. Level of drilling fluid in the annulus should be carefully monitored while lowering the tool string. Lost circulation, as seen by the sharp drop in the drilling fluid level, could arise from the fracturing of a weak formation from the pressure-surges created during the running-in process. Leakage in the drillpipe, which causes the air to issue from the end of the drillpipe, would also register itself in a similar manner. A slow drop in drilling fluid level indicates a leaking packer or fluid loss to adjacent zones. In the latter case, fluid loss will continue even after closing the shut-in valve. When the bottom is reached, the control head is connected, the packer is set, and the main test valve is opened. A high-pressure hose is connected to the control head, the other end immersed in a bucket of water. As soon as the test valve is opened, the blow of air from the drillpipe can be detected. The absence of a blow indicates a malfunctioning tester, a non-productive formation, or a plugged tool. As mentioned before, an appreciable pressure shock created on opening the main test valve is reduced to some extent by the choke. An additional back pressure against the formation is provided by a “water cushion”. This involves filling of the drillpipe with water to a desired level, so that on opening the main test valve a back pressure proportional to the hydrostatic pressure of this water column is exerted. It
334
may, however, have undesirable effects. In low-pressure wells, a water cushion may inhibit production from the formation. If the recovery from the well is small, the produced fluids may mingle with the water. High-pressure nitrogen gas, however, can be used to avoid both of these complications. Upon opening the main test valve and allowing the well to flow, an appreciable length of time will pass before the fluids reach the surface. The blow of air from the drillpipe, however, will indicate that the fluids from the formation are entering the drillpipe. The shut-in valve is closed when the desired test period is over. The pressure in the wellbore opposite the formation, which falls due to the production, will then buildup. If the shut-in time is sufficiently long, the pressure will approach the static reservoir pressure. In general, the shut-in period should be kept at least as long as
TIME
AK-1 CHART
.
TIME
Fig. 10-4. Typical drillstem test chart. (Courtesy of Lynes Inc., Houston, Tex.) Kuster K-3 chart records left to right, whereas Kuster AK-1 records right to left.
335
the production period. The pressure buildup with respect to time, as recorded by the bottomhole pressure recorders, can be used for valuable quantitative evaluation of the formation, e.g., permeability, extent of damage, reservoir pressure, presence of faults, and reservoir limits. This aspect is discussed in the section on buildup and drawdown tests. A five-position valve, run directly above the tester valve, is the device employed for the double shut-in pressure test. Initially, the main test valve is opened and the five-position valve is set in the open position. It is closed again by a few clockwise turns of the drillpipe, after an initial flow period of 5-10 min, to record the initial shut-in pressure. This initial shut-in period varies from 20 min to 1 hr. Additional rotation of the drillpipe reopens the valve for the usual flow period. At the end of the test, this valve is again closed upon a few more turns of the drillpipe. The circulation valve is then opened, after starting the mud pumps and establishing mud circulation.
INFLATE
@
INFLATE
Fig. 10-5. (a) DST chart showing packer seat failure-lost seat after tool opening. (b) DST chart showing packer seat failure-communication from annulus to interval. Indicates a hydrostatic leak during the ISI; this type of surging during the ISI cannot be caused by the formation pressure. (Courtesy of Lynes Inc., Houston, Tex., and Bill Clark.) ISI = initial shut-in.
336
Fig. 10-6. Pressure chart indicating slight plugging during the final flow period. Shut-in pressures suggest very low permeability. PF = 193 kPa; IS1 = 5460 kPa; IF = 586 kPa; FF = 579 kPa; FSI = 5536 kPa.
The test fluids are diverted into a portable tank through a suitable metering device. If the production is small, it may be measured in the drillpipe itself. The recovery is equal to the difference between the drillpipe volume and the volume of the mud that was added to the annulus to bring the test fluids to the surface. Because of the danger of a blowout and fire, the tools are pulled out at the end of the test very carefully. Any sign of an upward flow of drilling fluid indicates possibility of a blowout and necessitates emergency operations. As the drillpipe is drawn up onto the surface, the oil-bearing sections are capped to avoid spills. The drillstem test enables determination of the recovery and the composition of the formation fluids. Considerable amount of qualitative information is also easily available from the drillstem test pressure charts, as shown below. Qualitative drillstem test interpretation
A typical pressure chart from a drillstem test is shown in Fig. 10-4. The diagonal line up to A is the record of the hydrostatic drilling fluid pressure during the
@?I
Jm. RESTIIC'ION SURFACE
yp
Fig. 10-7. DST chart indicating surface restrictions. Buildup in pressures during VO is due to the fact that orifice size was too small;as orifice size was increased, bottomhole pressure decreased, whereas flow pressures at surface increased. The 6.35-mm choke was changed first to 12.7 mm and then to 19.05 mm. PF = 4915 kPa; IS1 = 11,858 kPa; IF = 4460 kPa; FF = 6856 kPa; FSI = 11,872 kPa. (Courtesy of Lynes Inc., Houston, Tex.) PF = preflow, IF = initial flow, FF = final flow, FSI = final shut-in.
331
Typical Damage Curve
I7
'
HIGH DIFFERENTIAL SHARP BETWEEN SHUT-IN /ANDFLOW \ RISE
Fig. 10-8. A typical DST chart in the case of wellbore damage. (Courtesy of Lynes-Inc.. Houston, Tex.)
lowering of the tools. The rough, jagged, or fuzzy appearance is attributed to the surges associated with running-in the tools. Excessive vibrations, however, indicate bad hole conditions. (See Lynes, 1980, for greater details.) At point A , the tools are at the bottom of the hole. Pressure at this point is the initial hydrostatic pressure ( I H P ) . Opening the main test valve leads to a sharp pressure drop below the packer to point B. As the fluids are produced, the pressure increases along line BC, the shape of which depends upon the properties of the formation and fluids. The point C represents the end of the flow period, whereas line CD shows the pressure buildup on shut-in. If the shut-in period is sufficiently long, the pressure at point D will approach the static reservoir pressure. Upon completion of the first shut-in, the tool is opened for the second and, possibly, final flow period. As tool is opened, the pressure drops to point E and then increases towards point F, similar to the initial flow period (line B C ) . The final shut-in is recorded as line FG (Fig. 10-4). As the packer is pulled loose on reaching point G , the recorder indicates an increase from the final shut-in pressure to the hydrostatic drilling fluid pressure at point H . This final pressure, FHP, must be equal to I H P . Figures 10-5 through 10-12 show pressure charts that have been obtained under various test conditions. From these and other charts it is possible to determine the
'
The pressure may increase, decrease, or remain constant depending upon the type and volume of fluid(s) entering the wellbore.
338 1, Trace does not return to baseline.
Possible Cause a) Over-relaxation of Bourdon tube. b) Temp, expansion of fill fluid. c) Gas leaking into bellows.
2. Trace returns below baseline.
Possible Remedial Action a) Prepressure before each use. b) Utilize filter assembly. c) Replace bellows (check length).
Possible Remedial Action
Possible Cause a) Bourdon tube leaks resulting in shortened bellows. b) Recorder housing leaked allowing pressure into gauge.
3. Stairstepping on chart. Possible Cause a) Friction due to dirt or improper adjustment.
4. Trace does not start or stop on baseline.
Possible Cause
a) Replace Bourdon tube (check bellows length). b) Replace "0'-rings on all housing connections.
Possible Remedial Action a) Check and clean lead screw and bearing. b) Check chart holder bearing. c) Check stylus tension.
Possible Remedial Action a) Make sure all connections are snug before drawing baseline.
a) Gauge was tightened after basiline was drawn.
5. Wavy baseline. Possible Cause a) Dirty lead screw. b) Improper stylus tension. c) Gauge too cold when baselinedrawn.
a) Clean lead screw. b) Adjust stylus tension. c) Warm gauge to above freezing
6. Clock ran away.
Possible Remedial Action
Possible Cause Many possible causes.
7.Clock stopped. Possible Causes Many possible causes.
Possible Remedial Action
Clocks should be repaired by trained personnel.
L
Possible Remedial Action Clocks should be repaired by trained personnel.
Fig. 10-9.DST charts indicating pressure gauge malfunctions. (Courtesy of Lynes Inc., Houston, Tex.)
339
Fig. 10-10. A DST run with water cushion inside the drillpipe. Point A indicates where the " trip-in'' was halted to add the cushion. Point B is the point where tool was opened for the first flow period; the flow pressure dropped to a point equal to the weight of cushion (hydrostatic head of water cushion) which had been added. The actual amount of water (fresh) cushion was 1700 m. Assuming a gradient of 10.42 kPa/m for fresh water, the hydrostatic head should be equal to 10.42X 1700 = 17,714 kPa. Inasmuch as the downhole gauge recorded a pressure of 17,744 kPa. the tools operated correctly. (Courtesy of Lynes Inc., Houston, Tex.)
High Rate Gas Test
Fig. 10-11. Pressure chart from a high-rate gas well. Depth = 1155 m: pressures: I S f P = 7927 kPa. F S f P = 7875 kPa, fFP = 422 kPa, and FFP = 319 kPa: times: 10. 60. 60 and 120 min: recovery: 169 m3/D gas, steady throughout final flow: 9 m-drilling fluid. (Courtesy of Lynes Inc., Houston, Tex.)
340
Fig. 10-12. Pressure chart from a well with slugging fluid. On testing a gas zone on or near the gas-water contact, there is a periodic pressure buildup as the gas chases water to the surface and then a pressure drop as the fluid is unloaded. Depth = 1166-1191 m; pressures (kPa): PF = 7314.1, IS1 = 9713.6, IF = 7342, FF = 8521.0, FSI = 9713.6; times: P F = 3. ISI = 75, FF = 90, and FSI = 90; blow: S I P on PF-GTS in 1 min, VO-weak to strong blow surging with heavy water spray; recovery: 3 m of sulfurous condensate; estimated 252,000 m3/D; high permeability. (Courtesy of Lynes Inc., Houston, Tex.)
recovery and whether or not the test was valid. In the case of test failure, the reason for failure can be determined. The quantitative aspects of DST interpretation are identical to the drawdown and buildup test calculations presented in the next section.
BUILDUP AND DRAWDOWN TEST FUNDAMENTALS
In order to develop a mathematical model of what is actually happening in the reservoir when a well is tested, many simplifying assumptions have to be made. A simple model is illustrated in Fig. 10-13. A vertical well of radius r, intercepts a homogeneous horizontal formation having constant thickness h , infinite extent, constant and uniform porosity 9, permeability k, fluid saturations So, S,, and Sg, and total compressibility ct. The gross formation fluid viscosity is p. It is further assumed that the formation properties are not time dependent.
Fig. 10-13. A simple model of the reservoir.
341
Taking a mass balance across the infinitesimal volume element (see Fig. 10-13): Mass-in - Mass-out = Accumulation or At[(pq)r-
( P q ) r + A r l =2Trh@At(pt+At-p,)
(10-1)
where q = volumetric flow rate of the fluid, p = density of the fluid, and At = infinitesimal time period. Dividing both sides of eq. 10-1 by ( A t A r ) and taking the limit as A t + 0 and A r + 0, yields:
a ( p q ) = 2 ~ r h +aPar
at
( 10-2)
Equation 10-2 is the continuity equation for radial flow. The Darcy's law for radial flow can be presented as follows: (10-3) Substituting eq. 10-3 in eq. 10-2, one obtains the following relation: (10-4)
The compressibility of a fluid, c f , is defined as the change in fluid volume per unit of total volume, per unit change in pressure: (10-5)
where V is the volume and T is the temperature. This equation may also be written as follows:
(10-6) Assuming incompressible rock, cf represents the total compressibility and is assumed to be constant. Thus: (10-7)
Inasmuch as:
(10-8)
342
then: (10-9) Combining eqs. 10-4 and 10-9:
wIa p
i a
(10-10)
or: (10-11) Equation 10-11 is the diffusivity equation in terms of density. Inasmuch as according to eq. 10-8: aP
=P
aP
C f Z
then:
aZp = a ar2
( ”) ar p c f Z
a P aP
= c --
f a r ar
+ pc,-a 2P = C f ( pc, )( ar2
S)+
a2p
PCIS
(10-12)
In addition:
( 10-13) Substituting eqs. 10-8, 10-12, and 10-13 into eq. 10-11:
or : (10-14) In order to obtain a linear equation, an additional assumption is required: either cf is very small or the pressure gradient a p / & is small enough (even around the wellbore) for cf ( a p / 3 r ) 2 to be negligible. Then eq. 10-14 becomes: (10-15)
343
This is the familiar diffusivity equation in terms of pressure. The term k / + p c f is known as the hydraulic diffusivity, q. For pressure-dependent porosity (significant rock compressibility, c,) an approximate result is similarly obtained with the fluid compressibility cf replaced by the total compressibility, c, (c, = c, + cf). In multiphase flow, cf = coSo+ cwSw+ cgSg. Generally, the rock compressibility is not negligible and it is, therefore, appropriate to write the diffusivity equation as follows: (10-16) Solution to the diffusivity equation for infinite reservoirs
If the reservoir is initially at an equilibrium uniform pressure, p i , the well starts flowing at a constant rate q at time t = 0, and the wellbore radius is negligible, the “line source solution” first derived by Kelvin (1904) can be used. and boundary (B.C.) conditions to be used are: The initial (Z.C.) (1) I.C.: p ( r , 0) = p i ( r > 0). Thus, the pressure at all values of r at t = 0 is p i . (2) B.C.: It p ( r , t ) = p i , t > O r-+ w (3) B.C.: For constant flow rate at the wellbore: It r4O
[r (?)I-
=--
” =constant 277kh
The “line source solution” for the diffusivity equation using these conditions is: (10-17) where p is the pressure, psia; q is the flow rate, bbl/D; p is the viscosity, cP; B is the formation volume factor, bbl/STB; k is the permeability, md; h is the formation thickness, ft; c, is the total compressibility, psi-’; r is the radius, ft; t is the time, hr; and $I is the porosity, fraction. These sets of units are commonly referred to as oilfield units or practical units. T o facilitate understanding and avoid confusion of different units, dimensionless variables have been introduced:
(10-18)
rD = r/rw
in particular: rDe = re/rw tD =
0.0002637kt
(in practical units)
(10-19)
344
and : p, - p = 141.2-p, 4BP
(in practical units)
kh
(10-20)
In terms of these dimensionless variables, the solution (eq. 10-17) becomes: p, = - i E i (
2)
(10-21)
where Ei is the exponential integral defined as: Ei(-x)
=
e-u -1 -du x
(10-22)
u
which is shown in Fig. 10-14. The Ei solution is valid for t,/r; 3 25, or when r , 3 20 and tD/rh 2 0.5. For low values of I , , and t,/r;, the Ei solution does not hold because of the assumptions made in the derivation. Van Everdingen and Hurst (1949) presented the wellbore solution and Mueller and Witherspoon (1965) gave the solution for intermediate values of r. These solutions are shown in Fig. 10-14 for comparison. For t D / r ; > 100, - Ei( - r6/4tD) [h(t,/rh) -t 0.809071 and eq. 10-21 becomes: PD =
3 [In( tD/rA) + 0.809071
(10-23)
The error in using eq. 10-23 is only about 1%for tD/rh > 55 and about 2% for t,/r; > 5. Thus, for practical purposes, the log approximation is satisfactory. As a matter of convenience, the natural logarithm in eq. 10-23 can be converted to the base-10 logarithm, malung eq. 10-23 more compact: (10-24) On converting p, to actual pressure drop using eq. 10-20: Pi-P(r, t)
=
162.6qBp kh
(10-25)
Pseudosteady state fjow Inasmuch as no real reservoir is truly infinite in extent, after the initial infiniteacting period, reservoirs will eventually either exhibit steady state or pseudosteady state behavior.
10-2 10-2
10-1
I00
101
/ r D2
t
D Fig. 10-14. Dimensionless pressure for a single well in an infinite system with no wellbore storage and no skin. (After Mueller and Witherspoon, 1965, pp. 471-474; courtesy of the Society of Petroleum Engineers of AIME.)
102
I 03
346
we1 I bore Distance from wellbore
+
Fig. 10-15. Pressure distribution in pseudosteady state flow.
In steady state flow, pressure at every point in the system does not vary with time. For radial steady state flow: 4=
0.00708kh ( p , - p , ) BPln(r,/rw)
which is simply the Darcy's law. In pseudosteady state flow, the pressure change with time, d p /d t, is constant throughout the reservoir. Figure 10-15 illustrates the above and also clarifies the nomenclature for the pressures p , p , and pi. The dimensionless pressure during pseudosteady state flow was given by Ramey and Cobb (1971):
(10-26) where A =drainage area, ft2; tDA = t D r:/A; and c, = shape factor, which is a characteristic of the system shape and the well location. Values for C,, which were given by Brons and Miller (1961), Dietz (1965), and others, are presented in Table 10-1. Radius of drainage and stabilization time
Stabilization time is defined as the time corresponding to the beginning of the pseudosteady state flow in the reservoir. It can be determined easily, using Table 10-1. For example, for a well in the center of a square, according to Table 10-1, the pseudosteady state flow begins at t D , = 0.1:
341
Solving for stabilization time, t,: (10-27) For a well at the center of most symmetrically shaped areas, (tDA)pss= 0.1 (Table 10-1) and eq. 10-27 can be used. The radius of drainage, also known as radius of investigation, depends largely on the criterion used to determine it. Some engineers define it as the radius at which q is 1% of the flow rate at the well, whereas others define it as the radius, r , at which ( p i - p , ) is 1, 2 or 5% of ( p i - pWr). In most definitions, the drainage radius defines a circular boundary with a pseudosteady state pressure distribution from the well to this boundary. The drainage radius, rd, is equal to: r d = 0.029
/
(10-28)
tc:
as given by Van Poollen (1964), Gibson and Campbell (1970) and Kazemi (1970). As the well is allowed to flow, the drainage radius, r,,, increases. Eventually, reservoir boundaries or drainage regions of adjacent wells will be encountered and rd will no longer increase further. Thus, eq. 10-28 applies only until the initiation of pseudosteady state as given by eq. 10-27 (or the ( t D A ) p svalues s in Table 10-1). Drawdown test
In drawdown test, the well is allowed ideally to flow at a constant rate and the bottomhole pressure is recorded at regular time intervals or continuously. The greatest difficulty is in maintaining the constant flow rate. Fortunately, multiple rate testing techniques are available today to analyze a variable flow rate test. A well, which has been shut-in for a long time, is suddenly flowed at a given rate q. In the constant flow rate analysis, it is assumed that the flow rate versus time is a step function, i.e., the flow rate instantaneously jumps from zero for the shut-in condition to q when the well is opened to flow. As the flow continues, the bottomhole pressure starts declining. These pressure decline characteristics convey valuable information about the reservoir and the condition of the well (storage, damage, etc.). The bottomhole flowing pressure, pwr,can be recorded either at the flowing well itself or at an observation well located at some distance r . The latter case is known as an interference test. When pressure is recorded at the flowing well, r = r, and thus rD = 1. The recorded pressure p ( r D = 1, t ) is equal to p W r ( t ) ,the wellbore flowing pressure. The following equation describes the relationship for such a situation, when t , 2 5:
4
p D( rD = 1, t ) = [In( t D )
+ 0.809071
where rD = 1 at the wellbore.
(10-23)
348
TABLE 10-1 Shape factors for various closed single-well drainage areas (after Earlougher, 1977, table C.1, pp. 203-204).
(Q
31 62
3 4530
-I 3224
01
0 06
0.10
31 6
3 4532
-1.3220
0.I
0 06
0 I0
27 6
3 3178
-I 2544
02
0 07
0.09
27 I
3 2995
- I 2452
0.2
0.07
0 09
21 9
3 0865
-I 1387
0.4
0 12
o.oe
0 098
-2 3227
*I 5659
0.9
0 60
0.015
30 8828
3 4302
- I 3106
0.1
0 05
0.09
I2 9851
2 5638
-08714
0.7
0 25
0 05
'2
I 5070
-03430
0.6
0 30
0.025
3 3351
I2045
-0 igr7
0.7
0 25
0.01
21 8 x 9
3 0836
- I I373
0.3
OJ5
0 025
I0 6374
2 3830
- 0 7870
0.4
0 I5
0 025
4 5141
I5072
-0 3491
1.5
0 50
0 06
2 0769
0 7309
+00391
1.7
0 50
002
3 1573
I 1497
-0 1703
0.4
0 I5
0 005
v3
El
EB H
B
l
349 TABLE 10-1(continued) CP
E€€Ell
I n CA
0.5813
-05425
*O 67%
2 .o
0.60
0.02
0.1 I09
-2. I991
+1.5041
3.0
0.60
0.005
5.3790
I6825
- 0 4367
0.8
0.30
0.0I
2.68%
0.9894
-00902
0.8
0.30
0.01
0.2310
-1.4619
t
I. I355
4.0
2.00
0.03
0.1I55
-2.1585
*I4838
(.O
2.00
0.01
2.3606
0.8589
-0.0249
I.o
0.40
0.025
2
Ep Ep
USE [ x , / x f ) *
,a
IN PUCE OF A/r:
FOR FRACTURE0 SYSTEMS
0.9761
-0.0835
0.175
0.08
CANNOT US€
0.7104
t0.0493
0.175
0.09
CANNOT USE
0 6924
*00583
0.175
0.09
CANNOT USE
1.6620
0.5080
t 0.1505
0.I75
0.09
CANNOT USE
1.3127
0.2721
+02685
0.175
0.09
CANNOT USE
-0.2374
t0.5232
0.175
0.09
CANNOT USE
2.95
-1.01
-
-
-
322
-1.20
-
-
-
I
1
350
There is an additional pressure drop due to the wellbore damage. This can be accounted for by using a skin factor, S , in the right hand side of the above equation (Hurst, 1953; Van Everdingen, 1953) as follows:
+
p D = [In( t D ) 0.809071 + S
4
=
[In( t D )+ 0.80907 + 2S]
(10-29)
Combining eqs. 10-19, 10-20 and 10-29 and rearranging, the familiar form of the drawdown equation may be obtained as shown below (Matthews and Russel, 1967): PWf =Pi - 162'6qBp kh [log t
+ log(
&)
- 3.2275
+ 0.86859s
1
(10-30)
Inasmuch as k , cp, p , c , , r: and S are constant for a given case, a plot of pwr versus log t will theoretically yield a straight line, with a slope, m , which is equal to: m
= - 162.6qpB/kh
(10-31)
This straight-line relationship is as follows: Pwf
=
log + Plhr
where: k
- 3.2275
+ 0.86859s
1
(10-32)
Thus, the formation permeability can be determined by plotting pwr versus log t and using eq. 10-31:
k = - 162.6qBp/mh
(10-33)
The skin factor, S , can be determined by using eq. 10-32:
S
= 1.151
+ 3.2275
(10-34)
where plhrcan be determined from the extrapolation of the semi-log straight line to a flow time of 1 hour (note that log 1 = 0). Multiple-rate drawdown testing
As pointed out earlier, it is usually impossible or impractical to maintain a constant flow rate long enough for a complete drawdown test. Multiple-rate test analysis techniques are required for such cases, which may range from totally uncontrolled variable rates to a series of step changes in rates.
351
Provided eq. 10-23 applies, the following analysis can be used (Earlougher, 1977): (10-35)
a straight line with slope: m ’ = 162.6Bp kh
(10-36)
and intercept: k
- 3.2275 + 0.86859s
(10-37)
is obtained. The qN signifies the last rate at any time r. The permeability can be obtained from the following equation: k=
162.6Bp m’h
(10-38)
and the skin factor is determined by using the following equation: (10-39)
Example 10-1 (1) Compute the transmissivity ( k o h / p o ) from the following pressure data:
0.1 0.3 1.0 3.0 5.0 10.0 50.0 100.0
1569 1560 1550 1542 1539 1533 1521 1516
352
Giuen: Flow rate, qo = 100 bbl/D, B, = 1.2 RB/STB, r, psi--', and C+ = 0.20. cP, c, = 200 x
=9 i
in. = 0.41 ft, p o = 2
Solution: Slope, m , can be determined from the pwrversus t plot (Fig. 10-16): m
= - 17
psi/cycle
Using eq. 10-33:
k = - 162.6qBp/mh Thus transmissivity is equal to:
kh/p
= - 162.6qB/m = - (162.6)(100)(1.2)/( - 17) = 1148
md-ft/cP
(2) If h = 40 ft, as determined from the logs, compute (a) the reservoir permeability. Also, (b) check the validity of the semi-log approximation. (a) k = (kh/p)oil(po/h) = (1148)(2)/40 = 57.4 md. (b) Semi-log approximation is valid for t , 2 100.
Thus, almost all the available data are in the range of validity of the semi-log approximation method used.
t
1520
1510-
Fig. 10-16.Semi-log plot for Example 10-1.
9
-
353
Pressure buildup test
In a buildup test, the well is shut-in after a producing period and the downhole pressure is recorded as a function of time. Downhole pressure is also recorded immediately before shut-in ( A t = 0). It is important to stabilize the well at a constant flow rate, q, before shut-in. The simple analysis technique presented here also assumes that the well has produced long enough to establish pseudosteady state flow conditions before shut-in, i.e., producing time, t,, is greater than the time for pseudosteady state, tpS:. If the flow rate during the production period was not constant, as is usually the case, the general practice is to use an “equivalent producing time”, t p e , determined by dividing N, (the total number of STB produced) by q (the last flow rate before shut-in):
(10-40) where N, = STB produced; q = flow rate just before shut-in, STB/day; and t,, = equivalent producing time, hr. A shut-in well can be viewed as flowing at a rate q - q ( A t = 0). What this means is that a well originally producing at a rate q for a time t and then shut-in at At = 0 (and kept in this condition for a length of time A t ) is equivalent to a well flowing at a rate q, superimposed by a flow rate of - q for time A t . If p,, is the shut-in bottomhole pressure, then using the principle of superposition:
or: p i -pWs = (162.6qBp/kh) log([ + A t ) + (162.6 ( - q ) B p / k h ) l o g ( A t ) . Thus:
(10-41) Equation 10-41 indicates that if the recorded shut-in pressure is plotted versus log(([ A t ) / A t ) , a straight line having a slope -m would result, where m is equal to:
+
m = 162.6qBp/kh
( 10-42)
This plot proposed by Horner (1951), is known as the Horner plot. The straight-line portion of the Horner plot can be extrapolated to a shut-in time, A t ,
354
ii
S h u t -in
I
\-attp
I
T1rne.t
Fig. 10-17. Idealized rate and pressure history for a pressure buildup test. (After Earlougher, 1977, p. 45, fig. 5.1; courtesy of the Society of Petroleum Engineers of AIME).
equal to 1 hr. The pressure at this shut-in time, as indicated by the extrapolated line, is called plhr.It can be used to determine the skin factor as follows:
+ 3.2275
S = 1.151
(10-43)
+
The straight-line portion of the Homer plot may also be extrapolated to ( t p A t ) / A t = 1 to obtain the initial reservoir pressure p i for an infinite system. Inasmuch as no real system is truly infinite, the extrapolated pressure is known as p * and is used as a calculation parameter to estimate the average reservoir pressure.
1100
I-L
1000
I000
800
1 n n 0
600 500
-
0
I
I
I
I
2
3
4
Time, months Fig. 10-18. Production schedule ( q o versus time) used in Example 10-2.
355
TABLE 10-11 Buildup test data used in Example 10-2. At
PWS
(hr)
(Psi4
0 3 3.5 3.82 4.20 5.25 10.52 21.10 42.40 85.71 221.0 466.0 600 840 1400 2100
2200 2350 2370 2375 2400 2410 2440 2470 2500 2530 2570 2600 2610 2620 2630 2635
(tp
+At)/Ar -
1401 1201 1100.5 1001 801 400.2 200.1 100.1 50.0 20.0 10.0 8.0 6.0 4.0 3.0
The most widely used techniques are those presented by Miller et al. (1950), Matthews et al. (1954) and Dietz (1965). Example 10-2 An exploratory well was placed on production according to the schedule presented in Fig. 10-18. Then a long buildup test was conducted and the data shown in Table 10-11were recorded. If 9 = 0.18, r,,,= 0.25 ft, h = 125 ft, p o = 10 cP, B, = 1.25 RB/STB, and ct = 20 X psi-', determine: (1) Reservoir effective permeability to oil. (2) Skin factor and the pressure drop due to skin effect. Solution : Using eq. 10-40: t , = Np/qlmt= (24 x 30)(1000 1100 800 600)/600 = 4200 hr. The p,, is plotted versus log[(t, + A t ) / A t ] in Fig. 10-19 (Horner plot). Slope m can be determined from this plot: m = 162.6q,pB0/k,h = 100 psi/cycle (1) k , = (162.6)(600)(10)(1.25)/(100)(125) = 98 md (2) Using eq. 10-43, the skin factor, S , can be calculated as follows: S = 1.151[(p l h r- p W r ) / m- log(k/+pctr2) + 3.231 = 1.151[(2340 - 2200)/100 - 10g(98/0.18 X 10 X 20 X X (0.25)2)+ 3.231 = -3.46 This answer indicates that the well has a higher permeability near the wellbore.
+
+
+
356
2200 104
103
10 2
101
I00
(%) Fig. 10-19. The Homer plot, pws versus log[( 1,
+ A r ) / A 11. for Example 10-2.
The pressure drop due to skin, Ap,, can be calculated using the following equation:
Apski,= (141.2qpB/kh)S
= 0.87 m S =
(0.87)(100)( -3.46)
=
-301 psi
Buildup following a long producing time Wells that have been producing for a long time may also be subject to a buildup test. In such wells, the buildup time At will be very short, even negligible, as compared to the producing time t , . Miller et al. (1950) suggested an approximate technique that saves considerable amount of time. For t, >> At, log[(t, + At)/At] = log( t,/At). Thus eq. 10-41 becomes: p,, = p i - 162.6(qBp/kh) log(t,) + 162.6(qBp/kh) log(At) or: pWs= constant + 162.6(qBp/kh) log(At). Thus, instead of plotting p,, versus log[(t, At)/At] as in the Horner plot, a much simpler plot of p,, versus A t can be used. It must be remembered, however, that this technique is only valid for shut-in times much shorter than the producing time.
+
Equivalent producing time The following equation gives the simplest approximation for equivalent producing time, as suggested by Horner (1951): t,,
= 24Np/9
(10-44)
This simplification, which corrects for small variations in the flow rate, is not applicable if sudden rate changes occur a very short time before shut-in. In other words, it is applicable only in the case of stabilized rate changes.
357
Odeh and Selig (1963) proposed a more general technique. They suggested that t: and q* should be used in the Horner's plot as follows:
t;
=2
and:
1
N
i= 1
t, -
(10-45)
(10-46) This method is very often used to analyze drillstem tests with no production at the surface. The flow rate at the rockface, which may be estimated from the pressure variations, does not stabilize in this case. The Horner plot can still be used to analyze buildup data, provided t: and q* are used instead of t and -4, as suggested by Odeh and Selig. Drawdown and buildup tests in gas wells Buildup and drawdown tests in gas wells may be analyzed using the same relationships as for liquid oil. One simply needs to convert the gas rates in STB/D and use the gas formation volume factor, Bg, instead of B, in the foregoing equations. Wattenbarger and Ramey (1968) have shown that this method is reasonably correct at high pressures (above 3000 psi). Gas viscosity and density vary significantly with pressure and, therefore, the assumptions used in deriving the diffusivity equation (eq. 10-16) are not satisfied. The most rigorous approach applicable to all pressure ranges is that of the real gas pseudo-pressure, m( p ) , defined by Al-Hussainy et al. (1966) as follows:
( 10-47) The pressure dependent p and Z are combined together with pressure and integrated over the interval pb to p , where pb is any arbitrary base pressure and p is the pressure of interest. Inasmuch as in pressure transient analysis one is only concerned with the difference in pressures, the following relationship can be used: .t(Pwr)= 2JPW'(P / P Z ) d P Pb
m ( Pi) =
P/~z)~P Pb
358
and:
This relationship shows that the base pressure pb is automatically cancelled out and does not enter into the calculation procedure. The pseudo-pressure, m ( p ) , is used instead of the pressure, p , for gases as follows:
where q is the flow rate in Mscf/day, T is the temperature in OR, and D I q I is the additional rate dependent skin effect due to non-Darcy flow around the wellbore. ] log f Examination of eq. 10-48 indicates that a plot of [ m ( p i ) - m ( p W f ) versus would be a straight line having a slope b, from which kh can be computed:
b = 1.151 x 50,300( p s c / q c ) (q T / k h )
(10-49)
The total skin effect, [ S + D I q I ] , can be evaluated from the following equation: S + D l q l =1.151([m(p,,,) - m ( p , , ) ] / b - l ~ g [ k / + ( p ~ ~ ) ~ r ~+3.2275) Z] (10-50) At low pressures, p Z is essentially constant for gases, whereas at high pressures it is directly proportional to pressure. Thus, m ( p ) is proportional to p at high pressures and is proportional to p 2 at low pressures. Equation 10-48, therefore, can be simplified:
at high pressures ( p > 3000 psia) and:
at low pressures ( p .c 2000 psia). Equation 10-48 is recommended in the intermediate pressure range: 2000 < p < 3000. At high pressures, eq. 10-51 is similar to eq. 10-30 for liquid drawdown. At low pressures, a plot of ( p ; - p $ ) versus log t should yield the familiar radial flow straight line, having a slope b': b'
= 1.151 X
50,300( Zipgi)(p s c / q c ) (q T / k h )
The skin factor can be determined by using (p:,, -p,$)/b' m( p w f ) ] / bin eq. 10-50.
(10-53) instead of [m(plhr)-
359
The non-Darcy skin coefficient D may be measured in a variable rate test by plotting observed skin effect, S’, against the rate, q. Inasmuch as S’ is equal to S D I q I, the slope of the straight line through the data point is equal to D.When such an analysis is not possible, D can be approximated as follows:
+
(10-54) where Gg is gas gravity with respect to air (air = 1). Similarly, in the case of buildup tests for gas wells when the following relationship is true: m( pws)= m( p * ) - 1637(qT/kh) log[ ( 1 ,
+At)/At]
(10-55)
an analysis analogous to that for oil wells can be used. At low pressures, when m ( p ) is proportional to p 2 , one can plot p 2 versus log[(t, + A t ) / A t ] . On the other hand, at high pressures, one can plot p versus log[(t, + A t ) / A t ] , because m( p ) is proportional to p . The remaining analysis procedure remains as before. Gas well testing In addition to regular drawdown and buildup tests, the natural gas industry has developed other tests based on the principles of drawdown and buildup. These tests, which are essentially an elaborate combination of drawdown and/or buildup, are intended to simplify the analysis, reduce the testing time, and gather more information. The two basic types of multipoint tests are (Fetkovich, 1973): (1) “Flow-after-flow” test, which consists of producing the well at successively increasing (normal sequence type) or successively decreasing (reverse sequence type) flow rates, with no shut-in between the flows. (2) Isochronal test and the modified isochronal test, which involve a shut-in period between the flow periods. These tests are based on the pseudosteady state form of eq. 10-52:
Substituting 14.7 psia for psc and 520”R for T,, the above equation becomes:
p~-p~r=1422(pgZ)i(Tq/kh)[ln(0.606re/rw) + S + Dlql]
(10-56)
Instead of using the turbulence factor D,turbulence is accounted for by putting an exponential, n , on the pressure drop term:
q = 7.03 X
kh( p,‘ - p $ ) ” / [
( pgZ)iT(ln(0.606re/rw) + S ) ]
(10-57)
360
Taking the logarithms of both sides of eq. 10-57: log q = log
c + n log( p,‘ -p:J
(10-58)
where: C = 7.03 X
k h / [ ( pgZ),T(ln(O.606r,/r,,,)
+S)]
(10-59)
Thus, if log q is plotted versus log( p,‘ - p;,), a straight line is obtained having slope n and intercept log C . Flow -after -flow tests The “flow-after-flow” test consists of a series of increasing flow rates (normal sequence) or decreasing flow rates (reverse sequence) on an originally shut-in well.
PWf,
___c
TIME
Fig. 10-20. Normal sequence for flow-after-flow test. (After Fetkovich, 1975; courtesy of the Society of Petroleum Engineers of AIME.)
361
Shut-in periods are not allowed between the flows. Shut-in, however, may occur if it is necessary to change the orifice plate for a new desired flow rate. Stable flow rates are indicative of a good test. For practical purposes, constant bottomhole pressure and rate of flow for a period of at least 15 min define stable conditions. The procedure for the flow-after-flow test has been described by the U.S. Bureau of Mines as follows: (1) Obtain the static reservoir pressure p , from the pressure gauge. (2) Flow the well for a period of 3 to 4 hours, which is sufficient to achieve constant flow rate and flowing pressure. The rate is fixed by a choke of appropriate size. Record pressure versus time. (3) Repeat the above for different flow rates (at least 4). (4) Plot flow rate q versus ( p,' - p;, ) on log-log paper. A straight line should be obtained.
Pi
Fig. 10-21. Reverse sequence for flow-after-flow test. (After Fetkovich. 1975: courtesy of the Society of Petroleum Engineers of AIME.)
362
(5) The absolute open flow potential is then obtained by reading q at pwr (i.e., at p,') (see Figs. 10-20 and 10-21).
=0
Isochronal test The isochronal test offers the only method of obtaining reliable performance curves (Fetkovich, 1975). The name of the test stems from the fact that the flow periods are of equal duration. If all the flow periods were not of the same duration, only the data for flow periods of equal duration are plotted to obtain the correct value of the slope n (see Fig. 10-24). It is necessary to plot the rates and pressures at a particular time and not the average values. The isochronal test is based on the principle that the drainage radius is independent of the flow rate and is a function of dimensionless time only. This test can be conducted on a constant rate or on a constant pressure basis, and a constant rate is not essential to obtain valid results. The latter is required only for purposes of superposition techniques to reduce testing time.
~ o r r n a i i y T, = T,=T,=T,
Normally
( T s Need not be equal ) T -
A TI # AT,# AT,
Fig. 10-22. Normal isochronal test. (After Fetkovich. 1975: courtesy of the Society of Petroleum Engineers of AIME.)
363
-
PWf2
Fig. 10-23. Modified isochronal test. (After Fetkovich. 1975; courtesy of the Society of Petroleum Engineers of AIME.)
Fetkovich (1975) recommended the use of bottomhole pressure gauges instead of surface recorders, which are subject to response lag and the effects of friction and flow temperatures. Surface pressures should be recorded with a dead-weight tester, together with the flowing temperature. Early time data is critical for this analysis, because the variations are more pronounced at this stage (see Figs. 10-22 and 10-23). Thus, the frequency of recording pressure and flow rate must be sufficient. It is also very important to clean the well prior to conducting the test. Plotting of the test data and analysis of the drawdown, buildup, and back-pressure curves during the test in the field are essential for obtaining valid results (Brown and Beggs, 1977). The procedure for the isochronal testing of a well can be outlined as follows (Brown and Beggs, 1977): (1) Obtain p , as described above. (2) Open the well to a specified choke size. Although a flow period of one hour is preferred, periods as short as 10 min can sometimes be used. Record the flow rate and pressure at specified time intervals. (3) Shut-in the well until the static pressure returns to the original value, p,.
364
2-
-
1
1
I
1
I I I 1 1
I
1
1
1
I
1
1 I l l
Q, ~ c f da t 14.65 p s i a Fig. 10-24.( p : - &) versus q plot for the isochronal test. (After Fetkovich, 1975: courtesy of the Society of Petroleum Engineers of AIME.)
(4) Repeat the procedure with different choke sizes for different flow rates. These tests can be taken on the same or different days, months or, even, years. ( 5 ) Plot (p,' - p $ ) versus q as described before (see Fig. 10-24). A series of parallel lines having the same slope n are obtained for the different time intervals. ( 6 ) In order to find stabilized conditions of pressure and flow rate, one flow test can be conducted for a long period of time (up to 15 days). Modified isochronal tests The modified isochronal test was developed for low-permeability reservoirs in order to reduce testing time. In this test the shut-in periods are kept equal to the flow periods. Even after relatively short flow periods it may take several days to obtain a final buildup pressure.
365
The difference between the pressure obtained on shutting-in the well for a short period of time and the stabilized pressure obtained on shutting-in the well for a long period of time (greater than or equal to the stabilization time) is determined for one flow rate. This difference is then extrapolated for different flow rates. This method has never been justified either theoretically or by field comparisons with true isochronal tests (Fetkovich, 1975). It has been assumed that the pressure is a function of the log of time, p = f(ln t ) . In reality, the relationship varies from p = f(ln t ) to p = f ( h ) for linear flow in fractured reservoirs (most low-permeability reservoirs have to be stimulated in order to be commercial). In order not to sacrifice the accuracy of test results, Fetkovich (1975) did not recommend the use of modified isochronal test or any other method based on superposition techniques for shortening test times for low-permeability wells. To reduce test time, it is better to use the two-flow rate method of Carter et al. (1963). Special tests Interference tests Interference tests are also known as multiwell or multipoint tests because they require the use of at least two wells, with the producing (or injecting) well being called the active well. Pressure is recorded at one or more observation wells. The pressure response characteristic at the observation well(s) gives the average formation properties between the active and the observation wells. Interference testing is considered to be superior to the single-well testing previously discussed. If the distance between the active well and the observation well is much closer than the distance to the nearest boundary (or to another active well) in the system, the pressure response at the observation well can be described by the logarithmic approximation to the Ei solution. Thus if ( t D / r i ) > 100:
The analysis techniques are the same as before (eqs. 10-42 and 10-43) with r, replaced by r, which is the distance between the active and the observation well. In interference testing, the reservoir porosity-compressibility product, which is frequently important to know, can be determined by using the following equation: +c, = k/r2p antilog[( pi -plh,)/m
- 3.22751
(10-61)
This is simply a rearrangement of the plhr relationship (see eq. 10-32), assuming-no skin effect: p l h r = p i + m [ l o g ( k / + p c , r 2 ) - 3.22751
366
Interference tests for permeability anisotropy Inasmuch as perfectly homogeneous reservoirs are rarely encountered, permeability anisotropy (i.e., the variation of permeability with direction) is exhibited by many reservoirs. This has an important effect on the efficiency of fluid injection and oil recovery. A well-defined permeability anisotropy can be of great value in the location selection for development wells and in the design of enhanced oil recovery projects. Based on the work by Papadopulos (1965), Ramey (1975) presented the technique described below. In isotropic formations, pressure gradient is proportional to flow rate:
v p = akq' where 4' is the flow vector. In anisotropic formations, this is not the case. Using any x-y coordinate system, the pressure gradient becomes:
There is one particular system of coordinates where the permeability matrix reduces to:
where k , , is the direction of the major and k , , is the direction of the minor permeability axes (Fig. 10-25). There are three unknowns: k,,, k,,,, and k.".",or k x x , k , , and 0 (the angle between the arbitrary coordinate system x-y and the major-minor permeability axes X - Y ) . The following basic equations, which allow type curve matching on the Ei function, have been presented by Ramey (1975):
=
- fEi[ ( -@pct/0.00105t)( k,,y2
t
k x x = 0 4 k,,
+ k,x2
+ k Y Y )+ J( k,, - k J 2
+4
- 2 k , , x y ) / ( k,,k,, - k:,)] (10-62) 4
= k,,,
(10-63)
(10-64)
e = arctan[ ( k X x- kx,)/k,,,]
(10-65)
367 Observation Well at(x.y)
Y \
Major permeability ,/ axis
\
\
,/
\ \
1
\ kmin\\ Active well
/ /
/ /
/
/
/
f
Well pattern coord i nates
\ \
\
Minor
\ \ permeability
axis
\
\
Fig. 10-25. Nomenclature for anisotropic permeability system. (After Ramey, 1975, p. 12; courtesy of the Society of Petroleum Engineers of AIME.)
If the cpct product is known, data from two observation wells are needed to determine the permeability matrix. Otherwise, data from three observation wells are needed. A field example was reviewed by Ramey (1975). On the type curve, pressure match yielded the value of [( k,, - k,y.,)2 4k:,], and the two time matches gave the values of k,, and k,. Then principal permeabilities and their orientation may be determined using eqs. 10-63, 10-64, and 10-65.
+
Pulse testing Pulse test is an interference test where the active well is alternately produced and shut-in for short periods of time. Its main advantage is the reduction in testing time.
Fig. 10-26.Schematic pulse-test rate and pressure history showing definition of time lag ( t L ) and pulse-response amplitude ( A p ) . (After Earlougher, 1977, p. 112, fig. 9-14; courtesy of the Society of Petroleum Engineers of AIME.)
368
Flow p e r f o r a t i o n s
-1
Casing packer
-
Pressure gauge Observation perforations
Fig. 10-27. Vertical interference and pulse test nomenclature. (After Earlougher, 1977, p. 135, fig. 10-25; courtesy of the Society of Petroleum Engineers of AIME.) h = formation thickness.
Kamal and Brigham (1975) presented a very simple technique for analyzing pulse tests. They generated type curves that are easy to use in hand calculations. Many other techniques require the use of computers. Figure 10-26 illustrates a pulse-test rate and pressure history, and the standard nomenclature used in characterizing the test. Vertical interference tests Vertical interference test was first proposed by Prats (1970) to investigate vertical permeability. Earlougher (1980) reviewed two methods derived from Prats' technique to analyze vertical well tests taking into account wellbore storage effects. Falade and Brigham (1974) proposed a simpler method which applies only to situations with negligible wellbore storage. Figure 10-27 shows how the top part of the formation is produced while pressure is recorded in the lower part of the layer. The upper part is produced at a constant rate (simple vertical interference) or production may be scheduled as pulses (vertical pulse test). Injection and fall-off tests A water injection test in a water zone is often analyzed as a production test using a negative value of flow rate (and a fall-off test is analyzed as a pressure buildup test). The basic underlying assumption in doing this is that mobility of the injection fluid is the same as that of the reservoir fluid. If cold water is injected into a hot water interval, water viscosities are not equal in the two zones created (hot and cold) If water is injected into a transition zone, or an oil zone, both relative permeabilities and viscosities are different in the two zones created, causing a change in mobility ratios and necessitating a more specific analysis.
'.
The two zones are located on the two sides of the interface between the injected fluid and the reservoir fluid.
369
,
Well test analysis in the presence of a gas phase The various methods presented here for the interpretation of well tests are based on the assumption that the reservoir contains a single fluid having a constant and small compressibility and constant viscosity. These methods have been used (Perrine, 1956) to interpret well tests performed on reservoirs containing both oil and gas by introducing the effective total properties of the multiphase system (corresponding to the single-phase properties of the different fluids). This empirical approach was examined by Weller (1966) and Earlougher et a]. (1967), who demonstrated that Perrine’s approach is valid in the presence of a very small gas saturation. Analysis becomes less accurate, however, as gas saturation increases. Interpretation of two-phase well tests enables estimation of single-phase permeability rather than the effective permeability to each of the flowing phases. This is done by incorporating changes in reservoir fluid properties and relative permeability due to the presence of gas saturation in the reservoir. SAMPLE QUESTIONS AND PROBLEMS
( 1 ) Show a log-log plot of pressure data for a test where the wellbore storage coefficient goes through a continuous change: (a) increasing, and (b) decreasing. ( 2 ) Consider a general partial differential equation:
Discuss the solution to this equation for n = 0, n blanks below: Case
Flow regime
n=O n=l n=2
-
= 1,
and n
= 2.
Fill in the
Relationship between p , and t , (infinite system)
( 3 ) Determine the pressure drop at well A given the following information: c$ = 0.2, k = 300 md, Bo= 1.35, c, = lo-’ psi-’, and h = 60 ft. The flow rates and,flow times for wells B, C , D, and E are as follows:
p o = 5 cP,
Well
Distance from well A, ft
Rate, STB/day
Duration of flow, days
B C
2210 2037.08 901.39 1562.05
250 150 300 350
20 5 2
D E
Well A at the center is not producing or injecting.
10
370
(4) Design a pulse test for injection into a watered-out reservoir. Estimated conditions are as follows: p = 1000 psi, B = 1.01, =0.2, ct = psi-', T = l0O0F, h = 50 ft, A q = 500 bbl/day, r = 660 ft, k, at So,= 300 md, p , = 0.8 CP at 100 O F , and A p , = 1 psi. (5) The m( p ) versus p for a gas reservoir has been computed and is given below. How would you plot pressure buildup or drawdown data for this reservoir if the pressure range is from 1600 psia to 3000 psia? Also, for a pressure range of 500 to 1000 psia?
+
562.5 X lo3
750 1000 1500
50 X lo6 86 X lo6 192 X lo6
1000 x103 225 x104
2000 2500 3000
330 X lo6 468 x lo6 606 x lo6
400 x104 625 x104 900 x104
(6) A variable injection rate test produced the following data: Time
Rate before change, bbl/day
Bottomhole pressure before change, psia
8:OO a.m. 8:lO a.m. 8:20 a.m. 8:30 a.m. 8:40 a.m.
1320 1240 1145 1042 994
375 324 262 193 157
Given: psi-', cf = 3.5 x l o p 6 psi-', r, h = 50 ft, c, = 4 x $I = 0.15. Find formation permeability and skin factor.
= 0.5
ft, p ,
= 0.75
cP, and
REFERENCES Al-Hussainy, R. and Ramey Jr., H.J., 1966. Application of real gas flow theory to well testing and deliverability forecasting. J . Pet. Tech., 18(5): 637-642. Al-Hussainy, R., Ramey Jr., H.J. and Crawford, P.B., 1966. The flow of real gases through porous media. J . Pet. Tech., 18(5): 624-636.
371 Brons, F. and Miller, W.C., 1961. A simple method for correcting spot pressure readings. J. Pet. Tech., 13(8): 803-805. Brown, K.E. and Begs, H.D., 1977. The Technologv of Artificial Lifr Methods, Vol. I. Pennwell, Tulsa, Okla., 487 pp. Carter, R.D., Miller, S.C. and Riley, H.G., 1963. Determination of stabilized gas well performance from short flow tests. J. Pet. Tech., 15(6): 651-658. Dietz, D.N., 1965. Determination of average reservoir pressure from build-up surveys. J . Pet. Tech., 17(8): 955-959. Dolan, J.P., Einarsen, C.A. and Hill, G.A., 1957. Special application of drillstem test pressure data. Trans. Soc. Pet. Eng. AJME, 210: 318-324. Earlougher Jr., R.C., 1977. Advances in Well Test Analysis. Monogr. Henry L. Doherty Ser. Vol. 5. Soc. Pet. Eng. AIME, Dallas, Tex., 264 pp. Earlougher Jr., R.C., 1980. Analysis and design methods for vertical well testing. J. Pet. Tech., 32(3): 505-514. Earlougher Jr., R.C., Miller, F.G. and Mueller, T.D., 1967. Pressure buildup behavior in a two-well gas-oil system. Soc. Pet. Eng. J . , 7(2): 195-204. Falade, G.K. and Brigham, W.E., 1974. The unulvsis of single well pulse tests in a finite-acting slab reservoir. 49th Annu. Fall Meet., SOC.Pet. Eng. AIME, Houston, Tex., Oct. 6-9, 1974, SPE 5055B, 17 PP. Fetkovich, M.J., 1973. The isochronal testing of oil wells. 48th Annu. Fall Meet., SOC.Pet. Eng. AIME, Las Vegas, Nev., Sept. 30-Oct. 3, 1973, SPE 4529, 15 pp. Fetkovich, M.J., 1975. Multipoint Testing of Gas Wells. SOC. Pet. Eng. AIME kd-continent Sect., Continuing Education Course .Well Test Analysis, March 17. S.P.E., Richardson, Tex. Gatlin, C.. 1960. Petroleum Engineering, Drilling und Well Completions. Prentice-Hall, Englewood Cliffs, N.J., 341 pp, Gibson, J.A. and Campbell Jr., A.T., 1970. Calculating the distance to a discontinui(v from D.S.T. dutu. 45th Annu. Fall Meet., SOC.Pet. Eng. AIME, Houston, Tex., Oct. 4-7, 1970, SPE 3016, 6 pp. Homer, D.R., 1951. Pressure build-up in wells. Proc., Third World Pet. Congr., The Hague, Sec. 11, pp. 503-523. Also 1967, in: Pressure Analysis Methods, Reprint Ser., 9. SOC.Pet. Eng. AIME, Dallas, Tex.. pp. 25-43. Hurst, W., 1953. Establishment of the skin effect and its impediment to fluid flow into a wellbore. Pet. Eng., 25(10): B-6-B-16. Kamal, M. and Brigham, W.E., 1975. Pulse-testing response for unequal pulse and shut-in periods. Trans. Soc. Pet. Eng. A I M E , 259: 399-410. Kazemi, H., 1970. Pressure buildup in reservoir limit testing of stratified systems. J. Pet. Tech., 22(4): 503-51 1. Kelvin, Lord Sir W.T., 1841-1904. Collected Mathematiml and Physical Papers. Cambridge Univ. Press, Cambridge. Lynch, E.J., 1962. Formation Evalution. Harper and Row, New York, N.Y., 422 pp. Lynes Inc., 1980. Testing and Chart Interpretation. Lynes Inc., Houston, Tex.. Matthews, C.S. and Russel, D.G., 1967. Pressure Buildup and Flow Tests in Wells. Monogr. Henry L. Doherty Ser., Vol. 1. SOC.Pet. Eng. AIME, Dallas. Tex., 167 pp. Matthews, C.S., Brons, F. and Hazebroek, P., 1954. A method for determination of average pressure in a bounded reservoir. Trans. SOC.Pet. Eng. A I M E , 201: 182-191. Miller, C.C., Dyes, A.B. and Hutchinson Jr., C.A., 1950. The estimation of permeability and reservoir pressure from bottom hole pressure build-up characteristics. Trans. SOC. Pet. Eng. A I M E , 189: 91-104. Also 1967, in: Pressure Analysis Methods, Reprint Ser., 9. SOC. Pet. Eng. AIME, Dallas, Tex., pp. 11-24. Mueller, T.D. and Witherspoon, P.A., 1965. Pressure interference effects within reservoirs and aquifers. J . Pet. Tech., 17(4): 471-474. Muskat, M., 1937. Use of data on the build-up of bottom hole pressures. Trans. SOC.Pet. Eng. A I M E , 123: 44-48.
372 Odeh, A.S. and Selig, F., 1963. Pressure build-up analysis, variable-rate case. J. Pet. Tech., 15(7): 790-794. Also in: Trans. SOC.Pet. Eng. AIME, 228: 790-794. Also 1967, in: Pressure Analysis Methodr, Reprint Ser., 9. SOC. Pet. Eng. AIME, Dallas, Tex., pp. 131-135. Papadopulos, I.S., 1965. Nonsteady flow to a well in an infinite anisotropic aquifer. Proc. Dubrounik Symp. on Hydrology of Fractured Rocks, 1965. Int. Assoc. Sci. Hydrol., 1: 21-31. Penine, R.L., 1956. Analysis of pressure buildup curves. Drill. Prod. Pract. API: 482-509. Prats, M., 1970. A method for determining the net vertical permeability near a well from in-situ measurements. J. Pet. Tech., 22(5): 637-643. Ramey Jr., H.J., 1975. Interference analysis for anisotropic formations-a case history. J. Pet. Tech., 27(10): 1290-1298. Ramey Jr., H.J. and Cobb, W.M., 1971.A general buildup theory for a well in a closed drainage area. J. Pet. Tech., 23(12): 1493-1505. Schlumberger Well Surveying Corporation, 1959. Schlumberger Formation Tester, Supplement No. 1. Van Everdingen, A.F., 1953.The skin effect and its influence on the productive capacity of a well. Trans. Soc. Pet. Eng. AIME, 198: 171-176. Also 1967,in: Pressure Analysis Methods, Reprint Ser., 9. SOC. Pet. Eng. AIME, Dallas, Tex., pp. 45-50. Van Everdingen, A.F. and Hurst, W., 1949. The application of the Laplace transformation to flow problems in reservoirs. Trans. SOC.Pet. Eng. AIME, 186: 305-324. Van Poollen, H.K., 1961. Status of drill-stem testing techniques and analysis. J. Pet. Tech., 13(4): 333-339. Van Poollen, H.K., 1964. Radius-of-drainage and stabilization-time equations. Oil Gas J., 62(37): 138-146. Wattenbarger, R.A. and Ramey Jr., H.J., 1968. Gas well testing with turbulence, damage and wellbore storage. J. Pet. Tech., 20(8): 877-887. Weller, W.T., 1966. Reservoir performance during two-phase flow. J. Pet. Tech., 18(2): 240-246.
373
Chapter I I
PRODUCTION LOGGING SANJAY KUMAR and GEORGE V. CHILINGARIAN INTRODUCTION
The subsurface measurements in production (or injection) wells which yield information on the nature and movement of fluids within the well are referred to as production logging. They are run after the production casing string has been cemented and the well placed on production. Production logs are used for the following purposes (Allen and Roberts, 1978): (1) Evaluation of completion efficiency for production as well as injection wells. (a) Flow profile. (b) Productivity (or injectivity) index of each zone. (2) Mechanical condition of the well. (a) Casing, tubing, or packer leaks. (b) Corrosion damage. (3) Detection of anomalous fluid movements. (a) Thief zones. (b) Channeling behind casing due to poor cement job. (c) Plugged perforations. (d) Encroachment, breakthrough, coning, etc. (4)Reservoir management. (a) Initial fluid saturation in each zone. (b) Changes in these saturations due to production and/or extraneous fluid movement. (c) Reservoir depletion pattern. (d) Flow profile. ( 5 ) Design and evaluation of stimulation treatment. Precise and reliable answers to these questions require careful design and application of production logging techniques. Inasmuch as each one of the production logging devices has its own limitations, multiple logs are run and data is evaluated. LOGGING DEVICES
The most widely used through-tubing production logging tools are: (1) High-resolution thermometer. (2) Gradiomanometer.
374
cc L - Electronic cartridge
Temperature-rensit ive.. resistor
Fig. 11-1. High-resolution thermometer. (After Allen and Roberts, 1978, p. 12, fig. 2-2; courtesy of Oil and Gas Consultants International, Inc., Tulsa, Okla.)
Fullbore spinner flowmeter. Continuous flowmeter. Inflatable packer flowmeter. Caliper. Manometer. Radioactive tracer survey. Bottomhole pressure device. High-resolution thermometer High-resolution thermometer is used to record the subsurface temperature profile in a temperature-versus-depth readout. The sensing element (metallic filament) constitutes the fourth arm of a sensitive bridge circuit. This, in turn, controls the frequency of an oscillator in the downhole electronic cartridge (Fig. 11-1). Absolute temperatures are usually recorded. In addition, there are provisions for recording the differential temperature ( A T ) curve that compares temperatures at two points, a short distance apart, on a more sensitive temperature scale. The resolution is about 0.04"F and the range is 0-350°F. Gradiomanometer The gradiomanometer records the pressure gradient. The two sensing bellows (Fig. 11-2) measure the difference in pressure over the provided spacing of 2 ft. This pressure difference is the sum of the hydrostatic, friction, and kinetic heads between the two points. At usual flow velocities, friction is negligible and the flow velocity remains unchanged over the small vertical interval of 2 ft. The kinetic effect is also nil. Thus, the pressure difference, as reported by the gradiomanometer, represents
315 -ELECTRONIC
CARTRIDGE
--TRANSDUCER
t
- -UPPER --
SPAC ING 2 FEET
-SLOTTED -FLOATING
SENSING BELLOWS HOUSING CONNECTING TUBE
LOWER SENSING BELLOWS
EXPANSION BELLOWS
Fig. 11.2. Gradiomanometer.(Courtesy of Schlumberger.)
the average fluid density difference only. This feature renders the tool most effective for identifying gas entry and computing water holdup and, subsequently, locating the standing water levels. From dynamic (flowing well) gradiomanometer measurements, the apparent fluid density pt is established at various depths. Then, after the production logs have been run under dynamic conditions, the well is shut-in. After a short period of time, the oil and water will segregate. The gradiomanometer is then run (under these static conditions) again to determine the oil and water densities po and pw, respectively. The water holdup' yw can be calculated at various depths from the following equation: PI - Po
Yw =
Pw - Po
(11-1)
where p, = density of produced fluid (water + oil). The slippage velocity us = uo - uw, where uo = velocity of the oil stream, and u, = velocity of the water stream. The total fluid volumetric flow rate, q t , is determined from a flowmeter measurement. If A is the cross-sectional area of flow, then: (11-2)
(11-3) and
40 + 4 w = 4t
(11-4)
' Water holdup is defined as the fraction of water in the total volume of fluid at a particular level.
316
E LECTRON It CARTRIDGE
YATERCUl METER
-D E N S l l l E l I R __ PACKER SPRING
FLUID ENTRANCE
PORT
Fig. 11.3. Inflatable combination tool. (Courtesy of Schlumberger.)
Using these relationships, (11-5)
and 4w
'Yw[4t
- 4 ( l -Yw>l
(11-6)
The water cut, i.e, the fraction of water in the total flow stream moving past a particular level in the wellbore, is equal to: 4W water cut = -
(11-7)
41
Thus, the zone contributing water can be identified. Inflatable packer flowmeter
Inflatable packer flowmeter, which is used for recording injection and production profiles, is a spinner type velocimeter. The packer is inflated by pumping well fluids into it (see Fig. 11-3). The inflated packer forces all flow through the metering section consisting of the spinner flowmeter, vibrating densimeter, and a capacitive water-cut meter. The measurements have to be made step by step because the packer must be inflated and deflated for each measurement.
311
The spinner response is essentially linear with the volumetric flow rate and the viscosity effect is minimal, even in the case of gas. The range of the tool is about 10-1900 bbl/day (7-in. casing). At higher rates, the pressure drop across the packer and the altered flow profile due to the restriction may force the tool up the hole. The densimeter determines fluid density by measuring the resonance frequency of the hollow cylinder containing radial blades as shown in Fig. 11-3, through which the fluid stream is directed. Thus, densimeter differentiates between oil, water, and gas. The water-cut meter, on the other hand, differentiates between water and hydrocarbons through the measurement of the dielectric constant (80 for water and 2-6 for hydrocarbons). In case of formation of emulsions, only the continuous emulsion phase is identified by this meter. Although it is one of the most precise flow-measuring and fluid-differentiating tools available, operational complications limit its use. Continuous flowmeter
The continuous flowmeter is also a spinner-type velocimeter. Unlike the packer flowmeter, however, it records a continuous flow profile versus depth. The lower limit of the flow rate below which the tool will not operate is as follows (after Schlumberger, 1980): Casing and tubing size (in.)
Monophasic flow (bbl/day; viscosity = 1 cP)
5 7 9;
100 150
300
This tool is most effective for wells with high production rates and/or smalldiameter casing and single-phase flow conditions. The spinner speed is significantly affected by the fluid viscosity. Decrease in fluid viscosity causes increase in spinner speed, leading to recording of a higher rate at the surface. It is, therefore, important to establish the relationship between the downhole spinner speed and fluid viscosity. Figure 11-4 illustrates the construction of a continuous flowmeter. In addition to the requisite calibration, a correction factor has to be used to account for the flow profile. (The velocity in a circular conduit is maximum at the center and decreases to zero at the walls.) Experience shows that the average velocity is about 83% of that measured at the center. Thus, provided the tool is precisely centered in the well, a correction factor of 0.83 can be used to obtain the average velocity (after Schlumberger, 1980). Fullbore spinner flowmeter
The fullbore spinner flowmeter is very similar to the continuous flowmeter. It consists of a collapsible blade spinner velocimeter that can be lowered through a 24
378
-
CABLE
CONDUCTOR
-MAGNET /PICKUP
COIL
WELL CASING
Fig. 11.4. Continuous flowmeter. (Courtesy of Schlumberger.)
in. (or larger) tubing to measure flow rates in the casing below, above the following minimum flow rates (after Schlumberger, 1980): Casing size (in)
Monophasic flow (bbl/day; viscosity = 1 cP)
5 7 95
20
30 60
Below the tubing, the spinner blades open (Fig. 11-5) and are exposed to a large cross-section of the casing. This results in higher sensitivity, better response at lower flow rates, and reduced viscosity effects on spinner speed. The downhole calibration of the fullbore spinner flowmeter is similar to the continuous flowmeter. Caliper A through-tubing caliper is presented in Fig. 11-6. The caliper essentially measures the borehole diameter and is particularly important for determining openhole
319
Fig. 11.5. Full-bore spinner flowmeter: left-tool closed for running through tubing: right-tool open for logging flow in casing. (Courtesy of Schlumberger.)
Fig. 11-6. Through-tubingcaliper. (Courtesy of Schlumberger.)
380
flow profiles. It is also used inside casing for evaluating damage and scale of paraffin deposits that can seriously affect the flow-rate logs. The range of the instrument is 2-12 in., accuracy is k0.2 in., and the resolution is 0.1 in. (see Schlumberger, 1980). Manometer
The manometer consists of a Bourdon tube that drives a potentiometer. The accuracy of pressure measurements by the manometer (about 2%) is limited by the accuracy of the potentiometer. The usual pressure ranges are 0-5 psi and 0-10,000 psi (see Schlumberger, 1980). Radioactive tracer surveys
There are several different types of radioactive tracer surveys suited for specific applications in fluid-flow determinations. On using a suitable radioactive isotope and a well-planned tool combination and logging program, tracer surveys yield the best possible records of (1)quantitative fluid movement in water injection wells, and (2) the flow behind the pipe. Tracers, however, are not very effective for defining multiphase flow. The parameters to be evaluated are the radiation intensity, the half-life, bottomhole temperature, and compatibility with the wellbore fluid. As an example, radioiodine (13’1) in water solution is used for water injection well surveys, because this isotope is water-miscible and has a very short half-life of 8.1 days. Similarly, ethyl iodide or methyl iodide are used in surveying gas injection wells, because both are liquids containing 13’1iodine isotope having a half-life of 8.1 days. Radioactive (RA) tracer survey methods are listed below: ( I ) Velocity-shot method In the velocity-shot technique, the velocity of the RA shot ejected into the flow stream is measured, by recording the time necessary for the tracer to reach the downstream gamma ray detector(s) from the injector. It is preferable to use two detectors instead of one in order to establish a more accurate ejection time. The tool is stationary and the log is a function of time. Figure 11-7 shows a two-detector velocity shot. If h is distance between the gamma ray detectors, A is the cross-sectional area, and t is the time recorded, then the flow rate, 4, is equal to:
hA
4=7-
(11-8)
In a cased hole, A may be assumed to be constant. This method is most suitable for water- or gas-injection wells. It is not recommended for producing wells, because of the complications created by water holdup and slippage velocity, and the problem of producing RA fluid at the top.
381 VELOCITY SHOT
Fig. 11-7. Radioactive-tracer survey: velocity-shot analysis. (Schlumberger, 1973, p. 21, fig. 2-13; courtesy of Schlumberger.)
(2) Timed runs (controlled-time method) In controlled-time method, a large amount of the RA tracer is ejected at the bottom of the tubing. Subsequently, logging runs are made at definite time intervals, with a gamma-ray tool recording the position of the slug. The times of the ejection and of each run are carefully noted. Figure 11-8 presents a Schlumberger example of a timed-run RA tracer survey. The RA slug (points a, c, e , and h ) moves down the casing as shown. After entering the perforations opposite sand No. 3, a part of the RA slug (points f, j , n, and u ) channels up the casing annulus to sand No. 4. After entering sand No. 2, part of the RA slug (points i and p ) channels down the casing annulus to sand No. 1. Fluid appears to be entering sand No. 3, because of the stationary readings at points i, m , and q. Finally, some RA material is also trapped in a turbulence pattern just below the tubing as shown by points b, d , g, and k . Quantitative measurements in the casing annulus are impossible because of variations in the cross-sectional area of the flow path.
382 GAMMA RAY SURVEVS AT TIMED I N T E R V A L S
WELL SKETCH tI
tz
tr
t4
ts
Fig. 11-8. Radioactive-tracer survey: timed-runs analysis. (Schlumberger, 1973, p. 22, fig. 2-14; Courtesy of Schlumberger.)
The timed-run method only qualitatively detects the flow of fluids up or down the hole, in either the casing or the annulus. The use of this method is thus limited to the detection of any undesirable movement of injected fluid in the casing annulus.
383
Fig. 11-9. Radioactive-tracer survey: differential injection technique. (Schlumberger, 1973, p. 23, fig. 2-15; courtesy of Schlumberger.) SR = slightly radioactive; 1 = 100 bbl/D; 2 = 200 bbl/D; 3 = 300 bbl/D; 4 = 400 bbl/D; TL = total.
(3) Differential injection method Differential injection method is a special technique developed for openhole completions where the hole size is unknown. In this method, the tubing is run to the bottom of the openhole section and the water is then injected through both the tubing and the tubing-casing annulus (Fig. 11-9). Keeping the total flow rate constant, the individual rates in these two flow paths are varied. The water in either one is made slightly radioactive to distinguish one flow path from the other. For each combination of flow rates in the tubing and the tubing-casing annulus, the two waters will form an interface in the well, determined by the rates at which the water is being taken by the formations. These interfaces are detected by the gamma ray tool and the complete injection profile is then plotted as shown in Fig. 11-9. Miscellaneous tools
In very low-flow-rate wells, the inflatable combination tool (ICT) is used. This tool combines a packer flowmeter, a densimeter, and a water-cut meter. The production combination tool (PCT), which is used in high-flow-rate wells, consists of: (1) continuous flowmeter, (2) fullbore spinner, (3) gradiomanometer, (4) manometer, and ( 5 ) thermometer. In limited areas, the fluid sampler is used for getting a depth-controlled and pressure-sealed fluid sample for PVT analysis. A casing collar locator is routinely used together with all production-logging tools for providing positive depth correlations.
384
Fig. 11-10, Spinner in monophasic flow. (Courtesy of Schlumberger.) INTERPRETATION OF FLOWMETER LOGS
Monophasic flow
In the spinner type flowmeter logs, the spinner responds to the velocity of the fluid. As pointed out earlier, inasmuch as the tool is usually centered in the wellbore, the spinner responds to the velocity at the center of the well (Fig. 11-10), which is therefore, is different from the maximum. The measured fluid velocity, u,,,, average fluid velocity. A correction factor, c, (Fig. 11-11)can be used to account for this as follows: uavg.
c=-
(11-9)
urn,,,.
1.0
09
Lz
08
P0 2
07
z Q
MONOPHASIC FLOW
06 LL
5 0 0.5
lo3
lo4
lo5
106
REYNOLDS NUMBER N R ~
Fig. 11-11. Relationship between the correction factor, c, and Reynolds number, N R c , in monophasic flow. (Courtesy of Schlumberger.)
385
CL
"t
FLUID
VELOCITY,
V
Fig. 11-12. Spinner response characteristics. (Courtesy of Schlumberger.)
Extensive field experimental studies show that c = 0.83 f 0.5% at usual flow regimes. Figure 11-11,however, gives better results. Figure 11-12 shows the spinner response (rotations per' second) to fluid velocity for different theoretical cases (Schlumberger, 1980): (1) Ideal response: frictionless spinner in ideal (non-viscous) fluid. (2) With mechanical friction: real spinner in ideal fluid. (3) With both mechanical and viscous effects: this describes the real case in the presence of spinner friction and viscous fluid. For the last case, the intercept of the asymptotic response, i.e., the threshold velocity, u,, depends on the density and viscosity of the fluid. The velocity, u,, increases with increasing viscosity, p, and decreases with increasing density, p . Inasmuch as viscosity of the wellbore fluid at bottomhole conditions is unknown to some extent, U , is also unknown, which leads to tool inaccuracy at low flow rates. The in-situ calibration technique, however, overcomes this problem. In-situ spinner calibration The spinner responds to the relative velocity between the fluid and the tool. Its rotation speed, therefore, increases when it moves against ihe flow, and decreases when it moves with the flow, as shown in Fig. 11-13. When the tool moves with increasing velocity along the direction of flow, the rotation speed of the spinner continues to decrease. Eventually, a point is reached where the fluid and tool velocities are equal and the spinner does not turn. If tool speed is incrr ised further, the spinner will start turning again, but in the opposite direction. The in-situ calibration is based on the above-described principles. The spinner response is recorded over the whole of the producing interval (zone by zone if the producing interval is not continuous) by moving the tool both along as well as against the direction of flow of the fluid. The spinner speed is plotted versus cable
386
v,
Cable speed 0)
Flowmeter, in moving fluid
lllllli V,
A
(Fluid v e l o c i t y )
Frequency of rototion
b ) F l o w m e t e r response versus relative v e l o c i t y
V,
ROTATION
-
VO
v o + Vc
Relative velocity
-
Fig. 11-13. In-situ calibration of a spinner flowmeter. (Courtesy of Schlumberger.)
speed (i.e., tool velocity). Upon establishing the response line of the spinner, ut and the fluid velocity, uo, can be determined as shown in Fig. 11-13. Polyphasic flow
Polyphasic flow is defined as the concurrent flow of immiscible fluids. In oil field operation, this involves the flow of the following combination of phases: (1) oil and gas, (2) water and gas, (3) oil and water, and (4) oil, water, and gas.
387
For quantitative evaluations of the flow rates of each phase, several measurements are needed. Some important concepts are discussed below. It is important to note that polyphasic flow is always turbulent. (See Schlumberger, 1980.) Flow types Liquid-liquid flow types Diphasic liquid-liquid flow types include bubble flow and emulsion flow. In the case of bubble flow (Fig. 11-14), the continuous phase is either the light or the heavy one, both phases moving at different velocities. Emulsion, on the other hand, implies the intimate mixing of two phases (on a macroscopic scale) to form a single very viscous phase. Gas-liquid flow types Gas-liquid flow is usually a bubble type flow, i.e., gas bubbles dispersed in liquid. In addition, slug flow and mist flow are also possible. Slug flow, which is the alternate flow type of gas and liquid, is very heterogeneous and thus difficult to measure. In the case of mist, flow, tiny droplets (mist) of liquid move with the gas, the two having the same velocity. Emulsion and mist flows can be considered as approximating monophasic flow and can be easily handled by the logging tools. Bubble flow can be quantitatively evaluated by use of correlation charts. Slug flow, on the other hand, is difficult, if not impossible, to measure and the logging results can only be qualitatively interpreted.
0 0 0
0
0
0
0
Bubble flow
Slug
Emulsion
Mist
Fig. 11-14. Flow patterns in polyphasic flow. (Courtesy of Schlumberger.) Bubble flow: liquid-liquid or gas bubbles in liquid; emulsion: liquid-liquid; slug flow: gas-liquid; mist flow: tiny droplets of liquid move with the same velocity as gas.
388
Triphasic flow types An increased amount of complexity is introduced by triphasic flow and production logging becomes more qualitative than quantitative. All the flow patterns described previously are possible in this type of flow. In some cases, it is practical to consider the liquid phases to be an emulsion, e.g., when the gas forms slugs. Flow parameters
At least three flow parameters are usually required: (1) total flow rate, (2) slippage velocity, and (3) phase concentrations. Alternatively, it is necessary to know: (1)oil velocity, (2) water velocity, and (3) phase concentrations. To date, however, production logging tools only provide the total flow rate (flow-meter) and the average density (gradiomanometer). It is necessary, therefore, to resort to experimental correlations. Inasmuch as bubble flow is representative of polyphasic flow in many respects, some features of this type of flow are discussed below. These concepts can be easily extended to other flow types. If uo=velocity of the oil bubbles and uw = water velocity, then the slippage velocity us = uo - u,. Also, if qo = volumetric oil flow rate and qw = volumetric water flow rate, then the total flow rate qt = qo + qw and the water cut = q w / q t . The production logging tools (flowmeter and gradiomanometer) determine the total flow rate, q t , and the water holdup, y,. The water cut and water holdup are related, as previously described (eq. 11-6), through the slippage velocity as follows (Schlumberger, 1980, p. 41):
(11-6) This may be misleading, however, and presents the main problem in production logging. Inasmuch as the slippage velocity cannot be measured, experimental correlations have to be used to relate the measured holdup with the unknown slippage velocity. The influence of slippage velocity is due to the product u,(l - y , ) in the above equation. The slippage velocity is relatively viscosity-independent and depends mainly on the buoyancy forces arising from the density difference between the two phases. The slippage velocity, us, and holdup, yw, vary inversely, i.e., one decreases when the other increases, when oil bubbles move in water. This is obvious from Fig. 11-15, which shows a typical correlation between the slippage velocity and (1) the density difference and (2) water holdup. These charts, which are established in the laboratory, should be valid for downhole conditions and can be used for oil-water bubble flow. For gas-liquid flow, the slippage velocity for large gas bubbles is almost always close to 60 ft/min, and from field experience, a good approximation can be made.
389
0' 0.0
I
0.1
I
I
I
I
0.2
0.3
0.4
0.5
Denslty difference, g/cc
Fig. 11-15. Slippage velocity versus density difference between water and oil at various water holdups. (Courtesy of Schlumberger.)
The above-described discussion is true for bubble flow type only. Moreover, if the well is inclined, the buoyancy forces tend to segregate the fluids. The lighter oil flows through the upper part of the tubing. Thus, the flow is strongly perturbed and it has to be dealt with differently. Spinner response in diphasic flow
In diphasic flow, usually the two fluids move with different velocities. Obviously, it is important to know which one of these two velocities govern the spinner response. Laboratory studies have shown that the spinner speed versus cable speed relationship for diphasic flow is the same as for monophasic flow (Fig. 11-16). It was observed that: DIPHASIC F L O W OF OIL AND WATER
I / FREOUENCY OF ROTATION
, qOr65o Bop0 Q,= 500 B W P D pipe LO.= 5.9"
WITH
AGAINST
FLOW
i 50
5c
100
FLOW
-
CABLE SPEED ( F P M
1
Fig. 11-16. In-situ calibration of the spinner flowmeter in diphasic flow. (Courtesy of Schlumberger.)
390
(1) The threshold velocity u, depends on the water holdup. (2) The uf (Fig. 11-16) is related to the total flow rate as follows (Schlumberger, 1980, p. 54):
where A = cross-sectional area of the pipe and c = correction factor for flow profile ( = 0.83 as before). (3) As far as the spinner response is concerned, the diphasic flow behaves as a monophasic flow, but viscosity is different from that of either of the flowing fluids. Inasmuch as the water holdup varies with depth in a well, diphasic flow spinner calibration should be carried out over the whole producing depth interval. INTERPRETATION OF THE GRADIOMANOMETER
The gradiomanometer measures the pressure difference across the 2-ft spacing of the two sensing bellows, scaled in the terms of density (g/cm3). As mentioned earlier, the gradiomanometer reading por is the sum of the average fluid density, pf, and the friction and kinetic effects (Schlumberger, 1980, p. 49): PGr = Pf (1
+K +F )
(11-10>
where K = kinetic term and F = friction term. In inclined wells,
where 8 = angle of inclination of the well with respect to the vertical axis. The friction term includes: (1) pipe wall friction (can be accounted for by standard charts); (2) friction at the tool-fluid interface, which can be estimated by moving the tool at different speeds (similar to flowmeter calibration described earlier). The kinetic term appears in the logs as noise. This is significant and can be observed when the fluid velocities across the upper and lower parts of the gradiomanometer are different from each other. The average fluid density pf is related to the holdup y, as follows (in the case of oil-water diphasic flow) (Schlumberger, 1980, p. 49): Pf = P W Y W + P o 0
-Yw)
(11-12)
or Yw =
~
Pf - Po - Po
Pw
This is the relationship always used in gradiomanometer measurements.
(11-1)
391 TEMPERATURE SURVEYS
Temperature surveys are carried out using the high resolution thermometer under static (shut-in well) as well as dynamic (flowing well) conditions. The subsurface temperatures can be recorded downhole in the wellbore or at the surface. Subsurface recording requires only a wireline to lower the instrument to the desired depth, whereas surface recording necessitates an additional conductor cable to transmit the data from the measuring device to the recorder at the surface. Usually tools for spinner surveys or radioactive tracer surveys incorporate a temperature measuring device. Temperature variations downhole are recorded versus depth or time, by the recorder at the surface. A temperature-measuring instrument can also be permanently installed at the bottom of the tubing in the wellbore. As pointed out by Bogart and Woodruff (1969, p. 309), in order to run temperature surveys with wireline equipment, it is necessary to enter the well through a lubricator. Fluid loss (water, oil, or gas) through the lubricator could invalidate the results of the temperature survey. The temperature profile will be distorted and displaced upward if water is lost through the lubricator in a water-injection well which has been shut in to obtain equilibrium conditions (Fig. 11-17). Erroneous conclusions from interpretation of temperature surveys are obtained if the fluid loss through the lubricator is overlooked. On examining Fig. 11-17, the obvious and immediate
Fig. 11-17. Temperature survey showing the effect of fluid loss through the lubricator. (After Bogart and Woodruff, 1969, p. 310, fig. 4; courtesy of American Elsevier Publishing Co.,Inc., New York.)
392
conclusions would have been that water was moving upward behind the casing and behind the blank section of pipe in the liner. Calibration of the temperature-measuring element is necessary when correct absolute temperatures are required. Differences in absolute temperatures recorded during different temperature surveys may be attributable to utilizing different temperature elements which were not calibrated just prior to the surveys. Normal procedure with wireline equipment is to start at the bottom of the well (or at the bottom of the interval) and to run the temperature survey coming out of the well. Exact location of the instrument in the wellbore is known unless there is considerable stretch of the wire. In this case the operator should note the depth and the force needed to pull the instrument. Going down the wellbore, the instrument can hang up; therefore, its exact location will be unknown (Bogart and Woodruff, 1969, p. 309). In the case of the bomb type of instrument, a time period from one to several minutes is required for the temperature element to reach equilibrium with the environment. A period of ten minutes may not be sufficient to reach temperature equilibrium at the bottom of the wellbore (see Fig. 11-18; Bogart and Woodruff, 1969, p. 309). Temperature surveys can be carried out under either dynamic or static conditions, as described below.
STOP-TYPE SURVEY STARTEDAT BOTTOM xx OF EACH INTERVAL
X
RUN NO.2
'!
X
ia, u. X-
.x-x.
ZURVES DUE TO -ACK OF TEMP. CQUILIBRIUM
RUN No.3
'xx, X
a
X
\.
RUN N O . ~
\-
X-X-~_
RUN No. 2
\
.\ "
\
'X *
X
X X
'1
DEPTH x.x-fx
RUN No. 1: TOO FEW STOPS "2 AND LACK OF EQUILIBRIUM CONDITIONS DUE TO WIDE VARIATION IN TEMPERATURE
Fig. 11-18. Temperature surveys showing the importance of reaching equilibrium with the surrounding environment. (After Bogart and Woodruff, 1969, p. 312, figs. 5 , 6; courtesy of American Elsevier Publishing Co., Inc., New York.)
393
Dynamic conditions
The temperature profile in a producing well depends on the production flow rate, the geothermal gradient, system geometry, formation and fluid thermal properties, size and depth of producing interval, and the producing time. In an injection well, injection time and flow rate govern the profile, other factors being the same as before. An additional factor here is the temperature of injected fluid. Static conditions
Where large volumes of fluids have been injected over long periods of time, the temperature anomalies persist for extended periods of time. Best results are obtained by running the temperature log after the well has been shut in for at least twelve hours. The rate at which the temperature along the wellbore returns to thermal equilibrium is a function of: (1) rate at which fluids were injected; (2) injection time; (3) temperature of injected fluid or, alternatively, the temperature difference between injected fluid and the formation; (4) system geometry; ( 5 ) thermal capacity and thermal conductivity of beds; and ( 6 ) temperature variation during the injection. For all the examples presented here, the geothermal profile is assumed to increase linearly with depth. The geothermal gradient is normally obtained from a temperature survey in an idle well that has been shut in for some time. For best results, the wellbore should be full of static fluid. Obviously, well conditions prior to, and during, the survey determine the usefulness of the data. Presented below are some specific applications of temperature surveys. ( I ) Lost circulation Lost circulation may be defined as the loss of fluid to a formation during drilling. The degree of fluid loss may vary from gradual seepage to complete loss; in the latter case only a few feet of hydrostatic head can be maintained in the hole. Spinner surveys have been used, but the tool may become jammed by the fibrous or flaky materials commonly added to muds to restore circulation. Temperature surveys complement radioactive tracer surveys. As indicated in Fig. 11-19, the wellbore has been cooled by the mud. The two points of deflection are the results of fluid loss at each interval. Below the last zone of lost circulation, the temperature increases toward the normal geothermal gradient (see Bogart and Woodruff, 1969, p. 311).
(2) Cementing Cementing operation is the process of displacing cement slurry down the casing, tubing, or drillpipe to a predetermined point. The slurry is formed by mixing water with cement or with cement blended with other materials. Squeeze cementing and cementing the casing or liner are the usual purposes of cementing operations.
394
\
LOST CIRCULATION
-*
ZONES O F CIRCULATION
Fig. 11-19. Temperature surveys showing zones of lost circulation (cooled by the mud). (After Bogart and Woodruff, 1969, p. 312, fig. 7; courtesy of American Elsevier Publishing Co., Inc., New York.)
1
CEMENTING OPERATIONS
I
\-
TOP OF CEMENT
I
CHANGE IF HOLE SIZE
Fig. 11-20. Temperature survey during cementation operation, showing height to which the cement rose behind the pipe. (After Bogart and Woodruff, 1969, p. 312, fig. 8; courtesy of American Elsevier Publishing Co., Inc., New York.)
395
The cement bond log, radioactive tracer survey, and temperature survey are used to detect the top of the cement column and to measure the effectiveness of the bond between casing and cement column. The amount of heat evolved by the setting of cement placed behind the casing is sufficient to be measured inside the casing. The temperature increase enables the determination of the height to which the cement rose behind the pipe (Fig. 11-20). The amount of heat given off is a function of the volume of cement, and changes in this volume are caused by changes in hole size or by channeling of the cement (see Bogart and Woodruff, 1969, p. 311). (3) Fracturing Hydraulic fracturing can increase the productivity of a well and may increase the ultimate oil recovery. Fracturing is most effective in tight formations or when a damaged zone exists uound the wellbore. The effectiveness of a fracturing operation can be determined by taking a series of temperature surveys after the fracturing operation. Radioactive tracers are also useful in this regard. According to Bogart and Woodruff (1969, p. 311), during hydraulic fracturing, there is a considerable heat, transfer between the injected fluid and the formation
\ INTERVALS TAKING FLUID
FRACTURING WITH HOT FLUID
INTERVAL TAKING FLUID
Fig. 11-21. Temperature anomalies after fracturing: (a) fracturing fluid colder than the formation and (b) fracturing fluid at a higher temperature than the formation. (After Bogart and Woodruff, 1969, p. 314, figs. 9 and 10; courtesy of American Elsevier Publishing Co., Inc., New York.)
396
surrounding the borehole. The temperature change is greater for the formation which breaks down and accepts fluid than for adjacent formations which do not accept fluid. A quantitative interpretation can be made from the amplitude of the anomaly. Temperature anomalies obtained with fracturing fluids colder than the formation are indicated by Fig. 11-2l.a, whereas anomalies from fracturing fluids with a temperature above the temperature of the formation are indicated by Fig. 11-21.b. (4) Production
Temperature surveys will complement radioactive tracer surveys in determining production profiles in oil wells. The points of deflection (Fig. 11-22.a) result from fluid entry into the wellbore. At the bottom of the well, the temperature curve deviates from the geothermal gradient when fluid (water and/or oil) enters the wellbore and moves up. This fluid is warmer than the fluid in the intervals it is flowing past and is cooled by fluid entering the wellbore at a lesser depth (Bogart and Woodruff, 1969, p. 313). In a gas well, the pressure in the wellbore is less than the pressure in the reservoir and the gas cools as it expands due to the change in pressure. The degree of cooling is dependent on the composition. of the gas and the extent of the expansion. A
1
OIL PRODUCING WELL
I I
I\
I GEOTHERMAL\ I GRADIENT 1
I
I
I I I I I
I I
I
d
\
G Fig. 11-22. Temperature surveys in (a) an oil-producing well and (b) a gas-producing well. (After Bogart and Woodruff, 1969, p. 314, figs. 11 and 12; courtesy of American Elsevier Publishing Co., Inc., New York.)
397
deviation from the geothermal gradient indicates that gas is entering the wellbore (Fig. 11-22.b). A quantitative interpretation can be obtained from the amplitude of the anomaly (Bogart and Woodruff, 1969, p. 313). Casing leaks, tubing leaks, and flow behind the pipe from one zone to another can also be detected by temperature surveys. (5) Fluid injection
Fluid injection occurs during pressure maintenance, secondary recovery, or tertiary recovery. The purpose is to increase oil recovery by some means of artificial stimulation. In gas-injection wells, temperature surveys are used in combination with spinner surveys or radioactive tracer surveys to determine the injection profile during injection operations. The injected gas will be at a temperature higher than the temperature of the formation (Fig. 11-23.a). The amount of gas injected will influence the cooling of the gas. The increase of the temperature profile toward the geothermal gradient indicates that gas is not flowing past the interval. In a water-injection well, the temperature curve will be almost a straight line during injection operations owing to the cooling effect of the cold water (Fig. 11-23.b). The deviation to,ward the geothermal gradient indicates that a small
I
I
I
I
I I I I I I I
I
I
A
Fig. 11-23. Temperature survey in injection wells. (a) Gas-injection well. (b) Water-injection well. (After Bogart and Woodruff, 1969, p. 316, figs. 13 and 14; courtesy of American Elsevier Publishing Co., Inc., New York.)
398
L WATER
\
I
8 HOUR AND 30 HOUR SHUT-IN TIME PRODUCED
.. I
11-
FLU,D MOVING DOWN
FLUID INJECTED
Fig. 11-24. Temperature survey under 'static conditions in a water-injection well (8-hr and 30-hr shut-in time). (After Bogart and Woodruff, 1969, p. 316, fig. 16; courtesy of American Elsevier Publishing Co., Inc., New York.)
amount or no water is passing through the interval. This temperature curve indicates the injection profile at the time the survey was run and does not take into account past performance. A spinner survey or radioactive tracer survey indicates where the water is leaving the wellbore in a water-injection well. The distribution of the water into the formation can be determined from a temperature survey in a shut-in water-injection well. Usually two surveys are run approximately 8-12 hrs and 24-30 hrs after the water-injection well is shut in. The temperature curve will move toward the geothermal gradient in the intervals not taking water at the present time or in the past. An anomaly might be detected during a temperature survey in a water-injection well that is shut in (Fig. 11-24). Water from a sand with high pressure enters the wellbore and moves down toward the low-pressure sand. Fluid leaves the wellbore opposite the sand with the low pressure. A leak of the casing shoe might be suspected if a spinner survey or radioactive tracer survey indicates a large amount of water leaving the wellbore at the casing shoe in a water-injection well. Only a temperature survey in a shut-in water-injection well will definitely show when water is flowing past the casing shoe and behind the pipe into another sand formation (Fig. 11-25). Usually two surveys are run approximately 8-12 hrs and 24-30 hrs after the water-injection well was shut in. The advance of the flood front in a waterflood project can be followed by running temperature surveys in a producing well that is under static conditions (Fig. 11-26).
399
I
\
WATER INJECTION
\
\
\ \\\ /
SHUT-IN
I I
I I
J \
f
TIME
LEAK O F CASING SHOE
-----
Fig. 11-25. Temperature survey showing leak around casing shoe at 12-hr and 30-hr shut-in times in a water injection well. (After Bogart and Woodruff, 1969, p. 318, fig. 17; courtesy of American Elsevier Publishing Co., Inc., New York.)
1-7SECOND SURVEY SIX MONTHS AFTER FIRST SURVEY
-
BLANK PIPE
PRODUCING WELL UNDER STATIC CONDITIONS
Fig. 11-26. Temperature survey in a producing well under static conditions, for quantitative interpretation of the flood front advance. (After Bogart and Woodruff, 1969, p. 318, fig. 18; courtesy of American Elsevier Publishing Co.,Inc., New York.)
400
1 DAY SHUT-IN TIME AFTER STEAM INJECTION
X
,1
I
I I I
I I I
I I
I
I I-
___j
6 DAY SHUT-IN
I
Fig. 11-27. Temperature survey during “soaking” and production periods at one day and 6 days shut-in times after steam injection. (After Bogart and Woodruff, 1969, p. 318, fig. 19; courtesy of American Elsevier Publishing Co., Inc., New York.)
The sand to be investigated has to be behind a blank pipe. A quantitative interpretation of the flood front advance can be made from the amount of water injected and from temperature surveys at different time intervals. Temperature surveys and sometimes spinner surveys are used to determine profiles of steam injection in producing wells during the “huff-and-puff” method of steam injection. A quantitative interpretation can be obtained from temperature surveys at different time intervals. Usually steam is injected in a producing well for a period of five to twenty days; then steam injection is halted and the well is allowed to “soak”. During this soaking period, the pressure around the wellbore decreases and heat is transferred farther into the reservoir. After the soaking period, which lasts from one day up to ten days, the well is put back on production at a rate sufficiently high to take advantage of the increased production capacity resulting from increased oil mobility. Temperature surveys normally are obtained during the soaking period and during the production period (Fig. 11-27). Immediately after the steam injection is halted, the temperature profile becomes almost a straight line. During the production period, the temperature profiles will indicate the intervals which have been heated by large quantities of steam or hot water during the injection period (Bogart and Woodruff, 1969, p. 317).
401 APPENDIX 11.1-PRODUCTION
Example lI.I-I-PCT
LOGGING (FIELD EXAMPLES) (Courtesy of Schlumberger Inc.)
survey in a gas well
The production continuous tool (PCT) was run during a multipoint test in this high gas producer. The test was run on 48/64 in., 36/64 in., and 28/64 in. chokes. A typical sequence of PCT surveys for each flow rate includes (see Fig. 11.1-1): (1) A record of the bottomhole pressure from the moment the well is opened until it is stabilized. In this example, the manometer log shows that the well is practically stabilized in less than ten minutes. The bottomhole flowing pressure is also necessary for conversion of downhole flow rates to surface production rates. (2) Several flowmeter surveys at various cable speeds. This provides an “in-situ” calibration of the flowmeter (seee Fig. 11.1-2). From this calibration and the knowledge of the pipe inside diameter, the downhole flow rate can be computed at each depth. (3) A gradiomanometer survey. In this example, this survey is necessary to evaluate the flow rate between the two lowest sets of perforations (depth = 400 ft), where gas is bubbling through water. Under these conditions, the flowmeter is not reliable. (4) A temperature survey. It is necessary to know the flowing temperature and pressure to convert the measured downhole flow rates to surface production rate. Such logs provide a rough estimation of the producing capability of each producing horizon. As an illustration, the figure shows the determination of the open flow potential for that well deduced from log measurements: 55 MMcf/D. The result obtained from conventional testing, including a test on 1-in. choke, was 50 MMcf/D (the PCT was not run during that test; see Fig. 11.1-3). Example I I . I-2- Completion evaluation: initial flow profile
The well was drilled through a carbonate reservoir, as can be seen in Fig. 11.1-4 where the computer processed interpretation of the open-hole logs is reproduced. A 7-in., 26-lb casing was set, cemented and one 90-ft long interval was perforated in the oil zone as shown in Fig. 11.1-4. The well was acidized and then produced. Shortly after, PCT (production continuous tool) and TDT (thermal decay time) surveys were performed while the well was producing. The gradiomanometer, flowmeter and TDT logs are shown in Fig. 11.1-4. An interpretation of the logs indicates that: (a) Despite the apparent homogeneity of the reservoir, most of the production comes from the upper five feet of perforations (3830-3835 ft) (see flowmeter log in Fig. 11.1-4). The acid effect apparent in the TDT log proves that acid was injected mostly in the upper part of the perforated interval. (b) Between 3835 ft and 3920 ft, oil bubbles through water: the flowmeter
402 -
~
CONTINUOUS peed in I
FLOWMETER
SRADIOMANOMETI
spinner velocity ( r.p.s .)
( gm /cc )
MANOMETER
('F)
,2
1
~~
THERMOMETEF
11.7
at depth 510
12:
iurfoce Iroduction r a t e : !4.4 MMcf/D
22.7 MMcf / D
0
-!iz2 I
)
I
opened well a t 03.00
;\ closed well a t 05.15
3,4MMcf/D
I
Logs run w;th we flowing on 48/64' choke
-
Fig. 11.1-1. Production continuous tool (PCT) survey in a flowing well with 2-in. choke. Example No. 11.1-1. (Courtesy of Schlumberger, Well Surveying Corp.)
frequency increases slowly on the way up, while the water holdup is around 40% (downhole oil density = 0.45 g/cm3). Conclusion The completion is not very efficient. Only five feet of perforated interval out of ninety contribute significantly to the production of oil.
403
spinner 100
100
I
200
300 400 500 cable speed ( I t Irnin)
Fig. 11.1-2. In-situ calibration of the flowmeter. Example No. 11.1-1.(Courtesy of Schlumberger.)
This set of logs (CPI, PCT and TDT) form a sound reference for the control of a re-stimulation, or for the analysis of future reservoir and well problems. Example 11.I - 3 - Evaluation of completion: monitoring of acidizing
Figure 11.1-5 shows the result of acidizing in an oil well. Before the acidizing, the well was producing 2660 bbl/D, and 8250 bbl/D (measured at the surface) after acidization. A flowmeter was run before and after acidizing at the same logging squired (psi*)
drawdown
--- -----------
Pwf2-patrn2 10). ----
los.
I I
1 I
open flow potential
lo5
Fig. 11.1-3. Determination of the openflow potential of the well. pwr = flowing pressure measured by manometer. Example No. 11.1-1. (Courtesy of Schlumberger.)
404 TOT-K
,
30 RATIO
G
F
s/sm3
RPS
Dl
PO
41
490(
500( PRODUCTION
3900 bbl/D /OF
OIL
(NO WATER I
Fig. 11.1-4.Flowmeter, gradiomanometer, and TDT logs for Example No. 11.1-2. (Courtesy of Schlumberger, modified.) FM CH = formation characteristics, average core density, g/cm3; TDT = thermal decay time; neutron capture cross-section; G = gradiomanometer; F = flowmeter; S, = water saturation, I; OHL = openhole log, April 1974.
speed (70 ft/min). After redrawing the two logs on the same paper, a scale in bbl/D can be fitted so that the full flow is 2660 bbl/D for the first run and 8250 bbl/D for the second run. Then the contribution of each producing zone can be read directly, before and after acidizing. Zero flow cannot be taken directly from the logs in Fig. 11.1-5. Below interval A , there is a fluid of different nature, probably completion water, having a lower viscosity. Run No. 2 reads zero at the very bottom of interval A . Run No. 1 reads below the zero line at the bottom. The shape of the log and this low reading are typical of the response in oil bubbling up through a standing column of water. Example 11.I-4- Evaluation of completion: monitoring the perforating of a two-stage completion in a gas well
Figures 11.1-6 and 11.1-7 are PCT logs on a well which has been completed with 7-in. casing, filled with water, and flushed by perforating zone A (trigger zone) at the
405
Fig. 11.1-5, Flowmeter logs before and after acidizing. Example No. 11.1-3. (Courtesy of Schlumberger.) Flowmeter reading is in rps.
base of the gas reservoir. Completion No. 1 was intended to open zones B and C by perforating in gas. A series of production logs were then recorded at two different flow rates, 240,000 m3/D and 1,000,000 m3/D, as well as with the well shut in. Final completion No. 2 was achieved by opening zone D,known beforehand to be the zone of the highest permeability. Another series of production logs were performed with the well shut in and at flow rates of 500,000 and 1,400,000 m3/D (Fig. 11.1-7). Comparison between logs recorded at the high flow rates for completions No. 1 and No. 2 makes it obvious that zones A and B are not going to contribute much to the total production because a pressure drawdown of 49 kg/cm2 (700 psi) is necessary to bring them alive (Fig. 11.1-9), whereas the final pressure drawdown is of the order of 7 kg/cm2 (100 psi). This result suggests that another type of completion could increase the well deliverability or at least that a stimulation of zones A , B , and C is advisable before opening zone D (of course, the stimulation efficiency should be determined with another series of PL logs). From the data obtained from the logs (see Table 11.I-I), flow rates from in-situ calibration of the flowmeter (Figs. 11.14 and 11.1-9), and pressure drawdown from the manometer readings, it has been possible to determine with fair accuracy the openflow potential for each stage of completion (Fig. 11.1-10).
406 F
G
RPS
T
F
OC
2 3 0
T
0
RPS 100 2c
OC
I
2 62.6 66
'
! I
I
i
, II
!
f
+ I
:
!.-
MANOMETER
Y)
m
f
'DG
B
-m
MANOMETER
\
s 0
Y)
WELL FLOWING 1000000 M3/d
GAS
!EE
WELL FLOWING 1400000 M 3 / d GAS
Fig. 11.1-6. (Left) Production continuous tool (PCT)-A logs run after completion No. 1. Example No. 11.1-4. (Courtesy of Schlumberger.) I = tension device; 2 = speed; 3 = CCL (caliper log); F = flowmeter; G = gradiomanometer; T = thermometer.
Fig. 11.1-7. (Right) Production continuous tool (PCT)-A logs run after completion No. 2. Example No. 11.1-4. (Courtesy of Schlumberger.)
TABLE 11.1-1 Flow rates, flowing pressure, drawdown and openflow potential for Example 11.1-4 Flow rate, q (m3/D)
Flowing pressure (kg/cm2 )
Pressure drawdown
Completion No. 1
0 240,000 1,000,000
219 214 170
0 5 49
Completion No. 2
0 500,000 1,400,000
219 217.5 212
0 1.5 7
Open flow potential (m'm 1,800,000
8,200,000
407
Fig. 11.143. In-situ calibration of flowmeter at several depths and different flow rates following completion No. 1. Example No. 11.1-4. (Courtesy of Schlumberger.)
W
so
100
CABLE SPEED (rn/min)
Fig. 11.1-9. In-situ calibration of flowmeter following completion No. 2. (Courtesy of Schlumberger.) OPEN FLOW POTENTIAL COMPLETION NO. 2
-
OPEN FLOW POTENTIAL COMPLETION NO I-
PRODUCTION RATE
OIM~/~.~)
Fig. 11.1-10.Relationship between ( p z - p : ) and production rate for determining openflow potentials. Example No. 11.1-4. (Courtesy of Schlumberger.) pi = average reservoir pressure and p r = flowing pressure.
408
(Courtesy of Schlumberger.)
Example 11.I-5- Diagnosis of a problem: oil well producing water A well produces 6900 bbl/D of oil with a 7.5% water cut; the completion is open hole: 5i-in. bit. The PCT was run in order to determine the origin of water (only the bottom part of the logs is reproduced here; see Fig. 11.1-11). Gradiomanometer indicates that at point 1 only water is flowing. The flowmeter shows a flow rate of 540 bbl/D, which is close to the surface flow rate of 520 bbl/D (= 6900 X 0.075). All the water, therefore, is produced from the bottom part of the well. A cement plug set at a depth of 5600 ft eliminated the water completely.
Example 1l.I-6-Diagnosis of a well problem: gas channeling behind liner A well is completed with a 9$-in. casing, a 7-in. liner, a 5-in. liner, and a 4 i - h . open hole. The PCT log could not reach the open hole because of obstructions in the open hole. The well produces oil and gas, with no water. With the well flowing, the gradiomanometer shows an important friction effect in the small liner (see Fig. 11.1-12). The log was run down at 50 ft/min. The readings are as follows: (a) In the 5-in. liner: Density = 0.78 g/cc (b) In the 7-in. liner: Density = 0.7 g/cc (c) In the 9$-in. casing: Density = 0.64 g/cc
409
Fig. 11.1-12. Production continuous tool (PCT) for Example No. 11.1-6Schlumberger.)
flowing well. (Courtesy of
According to the flowmeter calibration in the 5-in. liner, the flow rate is 12,000 bbl/D. The friction effect for the gradiomanometer is (see Schlumberger, 1973, Chart 6-14): 1096 in the 5-in. liner 1012 in the 7-in. liner nil in the 93-in. casing where pGr= gradiomanometer reading, g/cm3, and p = average fluid density, g/cm3. The corresponding densities are: 0.71 g/cc in the 5-in. liner, 0.69 g/cc in the 7-in. liner, and 0.64 g/cc in the 9i-in. casing. Consequently, the fluid density is around 0.7 g/cc and the decrease in the 9i-in. casing is not due to a lower friction but to the presence of gas. The temperature log indicates a fluid entry at the liner lap. A temperature log was run just after shutting-in the well (see Fig. 11.1-13). It indicates very clearly a cool sink some 550 ft below the liner lap, because the gas leaves the formation at that
410
Fig. 11.1-13. Production continuous tool (PCT) for Example No. 11.1-6-well is shut-in. (Courtesy of Schlumberger.)
depth and enters the well after channeling behind the 741. liner. On the flowing temperature log, there is a very slight change of slope at that depth. It might have been overlooked if a shut-in log was not available.
Production tubing 9 5 0 m
-
D 9 8 5 994 m
Survey tubing 9 7 0 m
7-in,,2 6 - l b coring
C 1000-1009m
B 1045-1052m
perfo
- 4 in SC
A 1063- 1075m
Fig. 11.1-14. Well sketch for Example No. 11.1-7. (Courtesy of Schlumberger.)
411
Example 11.I - 7
The well was rod-pumped and produced oil through four perforated intervals A , B, C and D (see the well sketch in Fig. 11.1-14). The production G/O ratio being very high, it was decided that a dual production head and an additional tubing string would be installed in order to run production logging surveys and determine the origin of the free gas. Production logs A packer flowmeter and PCT (thermometer-gradiomanometer combination) logs were run.
TEMPERATURE 66
(OC
m II
A
B
Fig. 11.1-15. Logs for Example No. 11.1-7. (Courtesy of Schlumberger.)
412
Figure 11.1-15.B is a composite log showing the thermometer and gradiomanometer readings, and the flow profile as determined from the stationary readings of the packer flowmeter. Quick-look interpretation The gradiomanometer shows that: (a) There is a standing column of water up to 1029 m. Oil production from zones A and B is bubbling through this water. (b) Above 1029 m, the continuous phase is oil (no water); above 1010 m, gas bubbles through oil. Zones C and D produce gas. The temperature log indicates that below 1065 m, there is a natural geothermal gradient. The sharp change in the temperature gradient at 1065 m confirms that oil is produced at the top of the perforated interval A . Production from interval B is not apparent on the temperature log. A small cooling effect around zone C confirms the production of gas. Quantitative interpretation There are two types of two-phase flow: (a) From 1065 m to 1029 m: oil and water two-phase flow. (b) Above 1010 m: gas and oil' two-phase flow. The interpretation of the flowmeter and gradiomanometer logs leads to the following results:
Zone A Zone B Zone C Zone D
Total production (m3/D) 25 7
52 12
Oil production (m3/D) 25 7
small small
Gas production (m3/D) small small 52 12
These results are summarized in Fig. 11.1-15.A where the flow profiles of oil and gas are sketched. Conclusion Zones C and D produce most of the gas. They can be squeezed to suppress gas production without reducing the production of oil. The interested reader is referred for further information to excellent publications by Schlumberger. SAMPLE PROBLEMS AND QUESTIONS
(1) Explain the operation of (1) spinner survey and (2) tracer survey in detail. (2) List three uses of running temperature surveys and indicate how temperature gradients would appear on the recording chart. Give the reasons.
41 3
(3) State the applicability of the logging devices described in this chapter for the following cases: (a) Openhole, (b) cased hole, (c) flowing injection well, (d) flowing production well, (e) shut-in well. (4) A gas well is completed with a 9 5/8-in., 47 lb/ft casing. A flowmeter survey run above the perforations gave a fluid velocity of 300 ft/min. Calculate the downhole flow rate in ft3/day. (5) A 9500-ft deep oil well is completed with two zones A and B as follows: top of zone A’= 7800 ft, bottom of zone A = 7950 ft, top of zone B = 9300 ft, and bottom of zone B = 9350 ft. The bubble point pressure for oil in zones A and B is 2200 psia and 2900 psia, respectively. The viscosities are 0.7 CP for zone A, and 0.6 CP for zone B. Static pressure at 8000 ft, measured before the production logging survey, was 3600 psia. The PCT indicated a 3400 psia flowing pressure at 8000 ft, implying that there is no gas in the wellbore. The flowmeter data obtained is as follows: Cable speed,
Spinner frequency (rps)
(ft’min)
At station 1, at 7700 ft
At station 2, at 8500 ft
Going down
60 100
14.0 16.63
8.37 11.oo
Going
50 100 135
6.70 3.35 1.o
1.05 0.88 3.13
UP
(a) What is the fluid velocity at station 2, and the corresponding flow rate? (b) Assuming a zero threshold velocity at station 1, what is the fluid velocity and flow rate at station l? (c) What is the % error in flow rate at station 1 if it is assumed that the threshold velocity at station 1 is the same as that at station 2?
REFERENCES Agnew, B.G., 1966. Evaluation of fracture treatments with temperature surveys. J. Per. Technol., 18(7): 892-898. Allen, T.O. and Roberts, A.P., 1978. Production Operations, Vol. 11. Oil and Gas Consultants International, Tulsa, Okla., pp. 11-52. Bogart, A.J. and Woodruff, W.E., 1969. Technical services-wireline services. In: G.V. Chilingar and C.M. Beeson (Editors), Surface Operations in Petroleum Production. Am. Elsevier, New York, N.Y., pp. 307-338. Champion, C.A., Schaller, H.E. and Jackson, B.R., 1965. Some recent applications of radioactive tracers in determining subsurface flow behauiour. In: 40th Annu. Fall Meet., SOC. Pet. Eng. A.I.M.E., Denver, Colo., Oct. 1965, SPE 1246, 12 pp. Clavier, C., Hoyle, W. and Meunier, D., 1971. Quantitative interpretation of TDT logs. J . Pet. Technol., 23(6): 743-763.
414 Cooke Jr., C.E., 1978. Radial differential temperature (RDT) logging- A new tool for detecting and treating flow behind casing. In: 53rd Annu. Fall Tech. Conf. and Exhibition, Soc. Pet. Eng. A.I.M.E., Houston, Tex,. Oct. 1-3 1978, SPE 7558, 5 pp. Curtis, M.R. and Witterholt, E.J., 1973. Use of temperature log for determiningflow rates in producing wells. In: 48th Annu. Fall Meet., Soc. Pet. Eng. A.I.M.E., Las Vegas, Nev., Sept. 30-Oct. 3, 1973, SPE 4637, 12 PP. Dale, J.R., 1965. Temperature surveys pinpoint steam-injection problems. Pet. Eng., 37(12): 67-70. Doering, M.A. and Smith, D.P., 1974. Locating extraneous water sources with the gamma ray log. In: 49th Annu. Fall Meet., Soc. Pet. Eng. A.I.M.E., Houston, Tex., Oct. 6-9, 1974, SPE 5116, 3 pp. Hammack, G.W., Myers, B.D. and Barcenas, G.H., 1976. Production logging through the annulus of rodpumped wells to obtain flow profiles. In: 51st Annu. Fall Meet., Soc. Pet. Eng. A.I.M.E., New Orleans, La., Oct. 3-6, 1976, SPE 6042, 4 pp. Johnson, W. and Morris, B.P., 1964. Review of radioactive tracer surveys. Prod. Mon., 28(12): 20-24; 29(1): 13-17. Loeb, J. and Poupon, A., 1965. Temperature logs in production and injection wells. In: 25th Meet., Eur. Assoc. Explor. Geophys., Madrid, May 5, 7, 9, 1965. Meunier, D., Tiuier, M.P. and Bonnet, J.L., 1971. The production combination tool-a new system for production monitoring. J. Pel. Technol., 23(5): 603-613. Peacock, D.R., 1965. What you can learn from temperature logs. Pet. Eng., 37(10): 96-111. Schlumberger, 1973. Production Log Interpretation, pp. 21-23. Schlumberger, 1980. Cased Hole Services Seminar, Production Logging, 77 pp. Smith, R.C. and Steffensen, R.J., 1970. Computer study of factors affecting temperature profiles in water injection wells. J. Pet. Technol., 22(11): 1447-1458. Smith, R.C. and Steffensen, R.J., 1975. Interpretation of temperature profiles in water injection wells. J. Pet. Technol., 27(6): 777-784. Wade, R.T., Cantrell, R.C., Poupon, A. and Moulin, J., 1965. Production logging-the key to optimum well performance. J . Pet. Technol., 17(2): 137-144. Witterholt, E.J. and Tixier, M.P., 1972. Temperature logging in injection wells. In: 47th Annu. Fall Meet., Soc. Pet. Eng. A.I.M.E., San Antonio, Tex., Oct. 8-11, 1972, SPE 4022, 11 pp.
415
Chapter 12
GAS LIFT JOHN 0. ROBERTSON Jr., GEORGE V. CHILINGARIAN, WILLIAM G. CARTER and SANJAY KUMAR
INTRODUCTION
The first recorded use of an air lift system was in 1782 to remove water from a flooded mine shaft in Hungary. Subsequently, the air lift technique was further developed for similar use in the mining industry to lift large volumes of fluids from flooded mine shafts. In early 1864, air was used in the oil industry to lift water along with some oil from shallow wells located in Venango County, Pennsylvania, U.S.A. The air lift technique was used in 1899 for lifting oil in the Baku oil fields, Azerbaijan S.S.R., U.S.S.R. In California, U.S.A., air lift was introduced around 1909 to lift fluids out of stripper wells in the Kern River Field, Kern County, California, U.S.A. When applied to lifting hydrocarbons, three major problems were encountered by an operator using air lift: (1) oxygen in the air was corrosive and attacked all downhole iron; (2) a mixture of air and natural gas (hydrocarbons) was explosive and represented a fire hazard; and (3) when air was mixed with the produced gas, the heating value (Btu content) of the gas dropped and the gas was often unsaleable. Natural gas lift was introduced later because of these problems with the air lift. Gas lift gained widespread popularity in the U.S.A. in the early 1900s due to the successful application of the method in lifting fluid from wells in the Gulf Coast area, U S A . The percentage of wells put on gas lift has increased since the end of world war 11. This popularity is due both to improvements in equipment and a better understanding of the process. A considerable amount of research work was done during the 1952-1965 period on determining the pressure losses occurring in two-phase vertical flow. Research was also done for many types of flow occurring in vertical tubing strings. This has changed the application of gas lift from an empirical approach to a blend of practical experience and science. The efficiency of this system was improved with the increasing value of gas. At the present time, the engineer cannot afford to simply cycle the gas; instead economics dictate that one should sell the produced gas which is not required for gas lift or injection. Economics also force operators to carefully design all gas lift operations to insure that gas is not tied up in a lift system when it could be sold to help the cash flow of the project.
416
Brown (1973, p. 182) has summarized the following periods of development for gas lift: (1) Prior to 1864: Laboratory experiments performed with possibility of one or two practical applications. (2) 1864-1900: Lifting of fluids with compressed air, which was injected through the annulus or tubing. (3) 1900-1920: Gulf Coast “air for hire” boom. For example, the Spindletop Field in U.S.A. was produced by air lift. (4) 1920-1929: Straight gas lift utilizing natural gas, e.g., the Seminole Field, Oklahoma, U.S.A. (5) 1929-1945: Development of a multitude of flow valves. More efficient rates of production and proration acted as a stimulus for the development of the flow valves. (6) 1946-1967: Development of pressure-operated valves which resulted in practical replacement of all other types of gas lift valves. The concentric gas lift valve, developed since 1953, was popularized by slim-hole and dual completions. REVIEW OF GAS LIFT FUNDAMENTALS
Pressure gradients
Inasmuch as knowledge of liquid pressure gradients is required in gas lift studies, specific weights of various liquids are compared to that of fresh water in Table 12-1. For example, if a brine is 1.04 times heavier than fresh water, its specific gravity is 1.04 and the pressure gradient is equal to 0.45 psi/ft (= 1.04 X 0.434). The relationship between the pressure gradient and salt content in water is presented in Fig. 12-1. If the API gravity of an oil is given, its specific gravity at 60°F can be determined as follows: sp. gr. at 60°F = 141.5/(131.5
+ “API)
(12-1)
For example, the specific gravity of a 37”API oil is equal to: sp. gr. = 141.5/ (131.5 37) = 0.84.
+
Derivation of pressure at bottom of gas column
The pressure at point 2, p 2 , in psi in a gas column is equal to: P2
=P 1 +
(Y x AL)/144
(12-2)
where point 2 lies distance AL below point 1; p 1 is the pressure at point 1 in psi; and y is the specific weight of gas in lb/cu ft. On using the equation of state and considering 1 lb of gas: pv
= ZNRT
(12-3)
417
TABLE 12-1 Fluid weight conversion table (modified after Zaba and Doherty, 1956; courtesy of Gulf Publishing Co.) Gravity (OAPI)
60.0 59.0 58.0 57.0 56.0 55.0 54.0 53.0 52.0 51.0 50.0 49.0 48.0 47.0 46.0 45.0 44.0 43.0 42.0 41.O 40.0 39.0 38.0 37.0 36.0 35.0 34.0 33.0 32.0 31.0 30.0 29.0 28.0 27.0 26.0 25.0 24.0 23.0 22.0 21.0 20.0 19.0 18.0 17.0 16.0
Specific
Specific weight
gravity (SG)
(Ib/gal)
0.739 0.743 0.747 0.751 0.755 0.759 0.763 0.767 0.771 0.775 0.780 0.784 0.788 0.793 0.797 0.802 0.806 0.811 0.816 0.820 0.825 0.830 0.835 0.840 0.845 0.850 0.855 0.860 0.865 0.871 0.876 0.882 0.887 0.893 0.898 0.904 0.910 0.916 0.922 0.928 0.934 0.940 0.946 0.953 0.959
6.16 6.20 6.23 6.26 6.30 6.33 6.36 6.40 6.43 6.46 6.51 6.54 6.57 6.61 6.65 6.69 6.72 6.76 6.81 6.84 6.88 6.92 6.96 7.01 7.05 7.09 7.13 7.17 7.21 7.26 7.31 7.36 7.40 7.45 7.49 7.54 7.59 7.64 7.69 7.74 7.79 7.84 7.89 7.95 8.00
(Ib/cu ft)
46.1 46.4 46.6 46.8 47.1 47.4 47.6 47.9 48.1 48.3 48.7 48.9 49.2 49.5 49.8 50.0 50.3 50.6 50.9 51.2 51.5 51.8 52.1 52.4 52.7 53.0 53.3 53.6 53.9 54.3 54.7 55.1 55.4 55.7 56.0 56.4 56.8 57.2 57.5 57.9 58.3 58.7 59.0 59.5 59.8
Fluid head (Ib/bbl)
259.0 260.0 262.0 263.0 265.0 266.0 267.0 269.0 270.0 271.0 273.0 275.0 276.0 278.0 279.0 281.0 282.0 284.0 286.0 287.0 289.0 291.0 292.0 294.0 296.0 298.0 299.0 301.O 303.0 305.0 307.0 309.0 311.0 313.0 315.0 317.0 319.0 321.0 323.0 325.0 327.0 329.0 331.0 334.0 336.0
Height
Pressure (Ib/sq in/ft)
(ft/lb)
0.320 0.322 0.324 0.325 0.327 0.329 0.330 0.332 0.334 0.336 0.338 0.340 0.341 0.343 0.345 0.348 0.349 0.351 0.354 0.355 0.357 0.359 0.362 0.364 0.366 0.368 0.370 0.372 0.375 0.377 0.380 0.382 0.384 0.387 0.389 0.392 0.394 0.397 0.399 0.402 0.405 0.407 0.410 0.413 0.416
3.13 3.11 3.09 3.08 3.06 3.04 3.03 3.01 2.99 2.98 2.96 2.94 2.93 2.92 2.90 2.87 2.87 2.85 2.82 2.82 2.80 2.79 2.76 2.75 2.73 2.72 2.70 2.69 2.67 2.65 2.63 2.62 2.60 2.58 2.57 2.55 2.54 2.52 2.51 2.49 2.47 2.46 2.44 2.42 2.40
Buoyancy factor immersed) 0.906 0.905 0.905 0.904 0.904 0.903 0.903 0.902 0.902 0.901 0.901 0.900 0.900 0.899 0.898 0.898 0.897 0.897 0.896 0.896 0.895 0.894 0.894 0.893 0.892 0.892 0.891 0.891 0.890 0.889 0.889 0.887 0.887 0.886 0.886 0.885 0.884 0.883 0.883 0.882 0.881 0.880 0.879 0.879 0.878
418 TABLE 12-1(continued) Gravity
Specific
Specific weight
(OAPI)
gravity (SG)
(lb/gal)
15.0 14.0 13.0 12.0 11.0 10' A.P.I. or} Pure Water
0.966 0.973 0.979 0.986 0.993
8.06 8.11 8.16 8.22 8.28
60.3 60.7 61.0 61.5 61.9
l.OO 1.01 1.03 1.06 1.08 1.10 1.13 1.15
8.34 8.4 8.6 8.8 9.0 9.2 9.4 9.6
Salt Water}
1.154
9.625
1.18 1.20 1.22 1.25 1.27 1.29 1.32 1.34 1.37 1.39 1.41 1.44 1.46 1.49 1.51 1.53 1.56 1.58 1.61 1.63 1.65 1.68 1.70 1.73 1.75 1.77 1.80 1.82 1.85 1.87 1.89
9.8 10.0 10.2 10.4 10.6 10.8 11.0 11.2 11.4 11.6 11.8 12.0 12.2 12.4 12.6 12.8 13.0 13.2 13.4 13.6 13.8 14.0 14.2 14.4 14.6 14.8 15.0 15.2 15.4 15.6 15.8
Pressure (lb/sq W f t )
Height
Buoyancy factor
(ft/lb)
~
339.0 341.O 343.0 345.0 348.0
0.419 0.421 0.424 0.427 0.430
2.39 2.38 2.36 2.34 2.33
0.877 0.876 0.875 0.874 0.874
62.4 62.8 64.3 65.8 67.3 68.8 70.3 71.8
350.0 353.0 361.0 370.0 378.0 386.0 395.0 403.0
0.433 0.436 0.447 0.457 0.468 0.478 0.488 0.499
2.31 2.29 2.24 2.19 2.14 2.09 2.05 2.00
0.873 0.872 0.869 0.866 0.862 0.860 0.856 0.853
72.0
404.0
0.500
2.00
01853
73.3 74.8 76.3 77.8 79.3 80.8 82.3 83.8 85.3 86.8 88.3 89.8 91.3 92.8 94.3 95.8 97.3 98.7 100.0 102.0 103.0 105.0 106.0 108.0 109.0 111.0 112.0 114.0 115.0 117.0 118.0
412.0 420.0 428.0 437.0 445.0 454.0 462.0 470.0 479.0 487.0 496.0 504.0 512.0 521.0 529.0 538.0 546.0 554.0 563.0 571.0 580.0 588.0 596.0 605.0 613.0 622.0 630.0 638.0 647.0 655.0 664.0
0.509 0.519 0.530 0.540 0.551 0.561 0.571 0.582 0.592 0.603 0.613 0.623 0.634 0.644 0.655 0.665 0.675 0.686 0.696 0.706 0.717 0.727 0.738 0.748 0.758 0.769 0.779 0.790 0.800 0.810 0.821
1.96 1.93 1.89 1.85 1.81 1.78 1.75 1.72 1.69 1.66 1.63 1.61 1.58 1.55 1.53 1.50 1.48 1.46 1.44 1.42 1.39 1.38 1.36 1.34 1.32 1.30 1.28 1.27 1.25 1.23 1.22
0.850 0.847 0.844 0.841 0.838 0.835 0.832 0.829 0.826 0.823 0.820 0.817 0.814 0.810 0.808 0.804 0.801 0.798 0.795 0.792 0.789 0.786 0.783 0.780 0.777 0.774 0.771 0.768 0.765 0.762 0.759
(lb/cu ft)
Fluid head (Ib/bbl)
~
~
~
~
e
419 0.36
-...
0.38
g
0.40
F 0.42
E
5
0.44
W
3 In
0.46
W In
g
0.48
0'500
10
20
30
40
50
60
70
80
90
100
PERCENT SALT WATER
Fig. 12-1. Relationship between pressure gradient and percent of salt water in water-oil mixtures. (After Thrash and Brown, 1965, p. 8, chart 1.) If sp.gr. = 1.07, then the pressure gradient = 1.07 (0.434) = 0.465. Pressure gradient of oil salt water mixture = [(% salt water/lOO)X(water gradient)]+[(% oil/lOO)x(oil gradient)]. For example, in the case of 50-50 oil-water mixture, if oil is 42' API (0.354 psi/ft gradient) and sp. gr. of water is 1.07 (0.465 psi/ft pressure gradient), then the pressure gradient of a mixture = (0.50)(0.354)+ (0.50)(0.465) = 0.41 psi/ft.
+
where p = absolute pressure in lb/sq ft, u = specific volume in cu ft/lb, Z = compressibility factor, N = number of moles of gas, R = universal gas constant, and T = absolute temperature in OR. Thus: u = ZNRT/p
(12-4)
Inasmuch as: u=l / y
(12-5)
y =p / Z N R T
(12-6)
On substituting eq. 12-6 in eq. 12-2: p2= p l +(p/l44ZNRT)AL
(12-7)
Rearranging: (p2- P I )
=(p/l44ZNRT)AL
(12-8)
and
( d p / p )= (dL/144ZNRT)
(12-9)
In integral form, eq. 12-9 becomes: j P b c d p / p= j 0L ( 1 / 1 4 4 Z N R T ) d L PS
(12-10)
420
Thus: (12-11) and
( P bc/Ps)
= exp( L/144zav
NRTav
(12-12)
)
(12-12.a)
or P bc 'Ps exp( L/144zav
NRT,v
where Z,,= average compressibility factor computed at T,, and pa,; N = number of moles of gas; R = universal gas constant, which is equal to 10.7 for one mole of gas; ps = pressure at surface, psia; pbc = pressure at bottom of gas column, psia; L = depth of gas column, ft; and T,, = average temperature in column, OR.On simplifying the exponent and considering one lb of gas: (12-13) where Gg= gravity of gas as compared to that of air (sp. gr. air = 1). The average temperature in the gas column can be computed from the following equation :
T,, = T, + [( D/2)(dT/dL)]
+ 460"
(12-14)
where T, = temperature at the surface, O F ; D = vertical depth of well, ft; and dT/d L = geothermal gradient, which is around 2OF/ft. Geothermal gradient, however, varies with locality. Energy utilized in lifting fluids
The following energies are utilized in lifting fluids from an oil well: (a) reservoir energy, which is equivalent to the working submergence; (b) energy of the gas in the well fluids; and (c) supplementary energy contained in the compressed gas introduced at the surface. Types of gas expansion
The types of gas expansion and the energy derived therefrom can generally be classified as: (1) isothermal, (2) adiabatic, and (3) polytropic. Isothermal expansion Isothermal expansion ( pv = constant) is applicable to cases where gas/oil ratios
421
are low and the heat necessary to maintain approximate isothermal conditions is supplied by the oil. If the initial pressure (bottom of tubing) is p 1 and the discharge pressure (top of tubing) is p 2 (both in psia), then the work, W , in ft-lb done by 1 cu ft of gas (measured at pressure p 2 ) on expanding from p 1 to p2 is equal to:
Adiabatic expansion Adiabatic expansion ( p v k = constant) is applicable to open flows where the gas velocity is so great that little or no heat is transferred to the gas from the oil and, consequently, there is a drop in gas temperature due to expansion. The work, W , in ft-lb/cu ft gas is equal to:
(12-16) where k = ratio of the specific heat at constant pressure to the specific heat at constant volume. The value of k for natural gas with a gravity of 0.7, compared to air, is about 1.26. Polytropic expansion Polytropic expansion ( p u n = constant) is the most applicable to practical gas lift situation, because there is usually isothermal expansion at the lower end of the tubing and practically adiabatic expansion at the upper end. Thus:
(12-17) where n is around 1.20 in most cases. Volume of gas necessaty for gas lift
Assuming 100%efficiency, the volume of gas, 6 ,generally corrected to standard conditions in cu ft, necessary to lift one barrel of oil to the surface at a discharge pressure, p2,can be calculated from the following equation:
6 = 349G0L,/
W
(12-18)
where Go = specific gravity of oil; L, = working lift in ft; W = work of expansion of 1 cu ft of gas, calculated from eqs. 12-15, 12-16, or 12-17; and the constant 349 lb/bbl (= 42 gal/bbl x 8.33 lb/gal) is for water. Gas lift efficiencies
The gas lift efficiencies [(theoretical G/O)/(actual G/O)] are low and range from 2 to 30% (average is about 12%). Energy losses occur in the following ways:
422
(a) Leakage Leakage can occur in either the casing or tubing; however, it can usually be prevented or remedied. (b) Entrance and discharge losses These losses, which are usually small, occur when fluids enter pipes of different diameters. (c) Slippage losses Slippage losses occur in the lower portion of the eductor tube, caused by the fluids dropping back through the ascending gas. The slippage losses can be reduced by the use of foot pieces, which can be attached to the lower end of the eductor tube. The use of foot pieces results in the attainment of .a more thorough or intimate mixture of the circulated gas with the liquid to be lifted. This method, however, is not very common now. Another method of controlling slippage losses is the maintenance of high velocities in the lower end of the tubing (2 to 40 ft/sec).
(d) Friction losses The greater portion of the friction’losses occur in the upper part of the eductor tube. These losses ( hf ) can be controlled by changing the velocity of the fluid and gas. The friction factor, f, is a function of the Reynolds number ( V d p / p ) , where p = mass per unit volume, i.e., y / g , and p = dynamic viscosity of the fluid. Velocity in the upper end of the tube is about ten times greater than that at the lower end. The head loss due to friction, h,,, is equal to f ( l / d ) ( V 2 / 2 g ) ,where 1 = length of flowline in ft, d = diameter of flowline in ft, V = velocity of fluid in ft/sec, and g = gravitational acceleration (32.2 ft/sec-sec). As greater velocities give lower slippage but greater friction, slippage and friction losses should be balanced in order to give a minimum total loss. (e) Back pressure at discharge Losses can occur when there is excessive high back pressure in the flowlines and surface facilities. Kick-off pressure (without valves)
Two important pressures in a gas-lift operation are: (1)kick-off pressure, and (2) working pressure. Kick-off pressure is the pressure (measured at the input gas line at the surface) necessary to start the flow in a gas-lift well. It is the pressure required to “kick-off‘’ the well and begin movement through the flowline, when gas pressure alone is used (no rocking). The kick-off pressure is a function of the method used to start flow. A schematic of the pressure-time relationship for a typical gas lift well is presented in Fig. 12-2.
423
5
Fig. 12-2. Pressure-time relationship for kick-off through tubing (without rocking). (1) Gas is first introduced into annulus (any pressure). (2) Gas enters bottom of tubing. It begins to escape up the tubing and consequently the pressure does not rise as fast from (2) to (3) as it does from (1) to (2). (3) Liquid reaches the surface (kick-off pressure). (4) Oil standing in the well has been discharged and sufficient oil has not yet entered the wellbore from the formation to attain working fluid level. ( 5 ) Steady state-constant working pressure is achieved.
The kick-off pressure required for gas lift through tubing can be derived as follows: Assuming that both tubing and casing are of constant cross-section: S, x ~ / 4 (d: - d:
+ d 2 )= h ( 7774)d 2
(12-19)
and thus: (12-20) where S, = starting submergence in ft; d = i.d. of tubing in inches; d , = 0.d. of tubing in inches; d , = i.d. of casing in inches; and h = height of the fluid without gas in tubing in ft. The pressures at the bottom of the casing, Pbc, at the top of the casing, p t c , and at the top of the tubing, p t t , are related as follows: Pbc
+
= p t c A p , = p t t + 0.433Gh + A p t + A p ,
(12-21)
Assuming that the pressure of the column of gas in the casing, Ap,, is offset by the sum of pressures due to the column of gas introduced into the tubing, A.pt, and the pressure loss caused by the friction in the tubing, A p , , then the kick-off pressure, P k o , is equal to:
Pko = P t c ' p a
+ 0.434Gh
(12-22)
or p k o = [0.434S,G(d:-d,Z+d2)]/d2 where G = specific gravity of well fluid, and p a = atmospheric pressure.
(12-19)
424
The kick-off pressure for gas lift through the casing, on the other hand, can be calculated from the following formula: p k o =[ 0 . 4 3 4 S s G ( d : - d : + d 2 ) ] / ( d : - d : )
(12-23)
The following methods were used to reduce the kick-off pressure: (1) Lowering the head of the liquid in the tubing with a swab. (2) “Rocking” the well by applying pressure alternately to the tubing and casing. As a result, the inertia of moving liquid may allow gas to enter the tubing. (3) Utilizing a “stage kick-off” whereby the gas is introduced into the eductor tube at progressively lower levels by surface activated valves. (4) Kicking off through the casing. The common range of kick-off pressures is around 300-1000 psig.
Gas volumes necessary for gas lift The gas volumes necessary in gas lift operations are commonly around 1000-2000 cu ft/bbl, and range up to 12,000 cu ft/bbl. The volume input necessary to maintain flow is a function of: (1) amount of formation gas produced with oil, (2) amount of fluid to be lifted, (3) physical properties of fluids, (4) desired level of back pressure on the formation, and (5) length and diameter of the eductor tube. Figure 12-3 shows the relationship between the pressure at the bottom of tubing and the gas volume. If the volume of the introduced gas is too small, then the greater density of the oil in the eductor tube and the slippage losses cause heading, resulting in higher average operating pressures and lower efficiencies. On the other hand, if the volume of gas introduced is too large, the friction losses can increase the operating pressure and reduce production due to higher pressure at the rockface. Greater gas volumes, higher operating pressures, and higher kick-off pressures are required when the tubing is lowered and the submergence is increased. As the diameter of the tubing is decreased, the friction loss and the operating pressure increase, whereas the slippage and the required gas volume decrease.
-Effective R a n g e-I
GAS
,
‘-Increase
VOLUME
to
in Pressure d u e Friction
__t
Fig. 12-3. Relationship between pressure at bottom of tubing and gas volumes required to lift fluids
425
.
I-
LL
3 0
Gas O u t p u t for MaximumProduction
3 n
\,Gas Output for M a x . Liftin Efficienc
PRODUCTION, B B L 1 DAY
Fig. 12-4. Relationship between gas output and production.
Figure 12-4 illustrates the' relation between the output gas and oil lifted in gas-lift operation of a well. Such a graph permits the choice of gas output for either the maximum production or maximum lifting efficiency. Production would go to zero if the bottom tubing pressure were equal to the formation pressure, and the density of fluid were the same in the wellbore as it is in the formation. The tangent through the origin would give the minimum slope in [(cu ft/min)/(bbl/day)]. Competition or present value considerations may require maximum production rates, whereas a lack of competition or proration may offer a choice between the maximum lifting efficiency and maximum oil recovery. The maximum oil recovery is probably obtained with a minimum gas/oil ratio, which is controlled by regulating the back pressure on the formation (Fig. 12-4). Fluid velocity in eductor tube
The fluid velocity in the eductor tube can be calculated from the following formula:
where Vx= velocity in eductor tube at point x , ft/sec; qo = volume of oil produced, cu ft/day; q,=volume of produced gas (measured at pressure p 2 ) , cu ft/day; p 2 = discharge pressure, psia; p x = pressure at point x where velocity is measured, psia; and A = area of flow tube, ft2.
426
Average density offluid in eductor tube The average specific weight of fluid in the eductor tube, ya (lb/cu ft), can be computed from the following formula: ya =
[(349G0) +(0.0029R’Gg(p, +p2))]/(5.61
+ R’)
(12-25)
where Go = specific gravity of oil at average well temperature; R’ = cu ft of gas produced per bbl of oil at average pressure pa [ = 1/2( p1 + p z ) ] ; p1 = pressure at bottom of tubing, psia; p z = discharge pressure, psia; and G,=gravity of gas (referred to air = 1.0) at the discharge temperature and pressure. Paraffin accumulation in wells operated by gas lift Certain crude oils leave waxy deposits (paraffins) in the tubing or casing owing to the gas expansion resulting in cooling. Inasmuch as most of the gas expansion occurs in the upper portion of the well, most of the paraffin deposition occurs in this area. More paraffin is deposited in the gas lift wells than in the flowing wells because of the introduction of cold gas, as well. as extra gas expansion. Commonly, suspended particles of inorganic silt serve as the nuclei around which the oil and water emulsify to form a waxy deposit. The methods of removing paraffins include the following: (a) intrpduction of gas solvents, such as benzol, gasoline, and heavier distillates (frequently the solvent is preheated in order to increase the solubility of the wax); (b) direct application of heat in order to melt the wax; (c) preheating of the input gas; (d), use of scrapers to mechanically remove the wax; and (e) use of explosives. PRINCIPLES AND METHODS OF GAS LIFT
The mechanism used in the earliest air lift application (1864 in Pennsylvania) was that of a two-pipe system, in which air was injected down one pipe and returned through the second pipe at a shallower depth. The air returned to the surface pushing some fluid ahead of it in the second pipe. The first United States Patent for an “oil ejector” was issued in 1864 (Fig. 12-5). The use of gas lift in the U.S.A. gained popularity in the Gulf Coast area, where it was necessary to lift large volumes of fluid from shallow wells. Later, it was used in Louisiana and east Texas for shallow wells, where large volumes of gas were available. Successful development of many fields has been attributed to gas lift, such as the Evangeline Field in Louisiana and the Smackover and Spindletop fields in east Texas. Initially, gas lift was used as an intermediate production lift system after termination of the natural flow period of the well and up to the time of sucker-rod pump installation. The early method of gas lift was often referred to as “U-tube” technique (see Fig. 12-6): a short string of tubing was inserted into the well and
427 FLOW LINE
1
TUBING CASING
Fig. 12-5. First United States patent (No. 47,793) for an “oil ejector” issued to A. Brear, May 23, 1865. (After API, Vocational Training, 1965, p. 2, fig. 1-1; in: Brown, 1973, p. 182, fig. 8.1.)
higher-pressure gas was introduced either down the casing or down .the tubing. The fluid level in the casing-tubing annulus was displaced to the bottom of the tubing, when the gas was injected through the casing, and the gas escaping up the tubing lifted fluid both by carrying it in the gas stream and as a result of “lightening” of the fluid column with dispersed gas above the bottom of the tubing. If gas was injected through the tubing, the fluid level would also be at the bottom of the tubing and the
FORMATION OIL. WATER, AND GAS
Fig. 12-6. U-tube type of gas lift. (a) Open-end tubing with gas being pumped down tubing and fluid being produced through the casing. (b) Open-end tubing with gas being pumped down the casing annulus and fluid being produced through the tubing.
428
moving fluid column in the annular space between the casing and tubing would be lightened. Although simple, this system was very inefficient and initially required a high pressure to initiate the flow (" kick-off" pressure). The system pressure was lowered after the beginning of the flow and lightening of the fluid column above the bottom of the tubing. The problem of high "kick-off" pressure was solved by the development of gas lift valves, which were placed in the tubing string at various levels (successively lower injection points). They were successively uncovered by the lowering of the fluid level in the annulus. The differential pressure between the tubing and the annulus (through the valve) or the flow of gas through the valve (depending upon the type of valve) causes each valve to close as flow begins through the valve directly beneath it (Fig. 12-7). In the middle thirties, the intermittent gas lift method was introduced as a production method from a reservoir having relatively lower pressures. Initially, the intermittent-type valve was operated by using a lever on the surface connected to a wire extending to the valve on the tubing string. A timing device at the surface activated a gas-driven piston, which pulled the lever arm and the wireline, opening the particular valve. Usually, the valve in the tubing string was designed with a rod that could pass in and out of the valve to control the open and closed positions. The valve in the tubing consisted of one or two balls and seats, with the ball protruding into the tubing as shown in Fig. 12-8. On passing into the valve, the rod forced the ball off its seat, thus opening the valve. Disadvantages of this type of valves were: (1) wireline could wear a hole in the tubing, (2) wireline was subject to corrosion and frequently broke, and (3) the problems of unloading the fluid in the tubing. The wireline-operated valve system became less popular with the advent of pressure-operated valves, which are easy to operate.
-c
CASING TUBING
4
FLOW LINE
f
GASIN
DISK VELO CLOS VA LV
OIL L
E
V
Y
.
STAN DIN G VALVE
Fig. 12-7. One of the first types of kick-off valves developed in the 1930s. This type of valve is referred to as " velocity-controlled". As the gas and liquid stream velocity increases, the valve is closed. (After Brown, 1973, p. 184, fig. 8.10.)
429
SECTION A - A
Fig, 12-8. Mechanically controlled valve. This valve was opened and closed by use of the wireline that extends to the surface. (After Brown, 1973, p. 186, fig. 8.18.)
Gas lift terminology
The following terms are encountered in gas lift operations (refer to Fig. 12-9): Gas lift: A method of lifting fluids which utilizes energy contained in compressed gas to lift well fluids through an eductor pipe from a lower to a higher level. The lifting of fluid using gas is achieved by one or a combination of the three following processes: (1)work of expansion of the compressed gas, (2) aeration (lightening) of the fluid column, and (3) displacement of the oil by compressed gas. Lift, L: Vertical distance between the fluid level and level of discharge at the surface. Static fluid level: The level to which the fluid will rise in a wellbore under conditions of pressure equilibrium. Vertical static head, H,: Vertical fluid column distance between the static fluid level and midpoint of perforations. It can be expressed in pressure units, e.g., psi. Static submergence, S,: Vertical distance between the static fluid level and the bottom of the tubing. Working fluid level: Theoretical fluid level which is used in computing working submergence and working lift, defined as the level to which the oil column would rise behind the production string owing to the working pressure at the bottom of tubing. It is assumed that the annulus behind the production string is connected by a gas column to the surface discharge (Chilingar and Beeson, 1969).
430 (FLUID DISCHARGE
fi
,WATER
^..
UIL
STRING
^S _I K_I.N. .C_,
ST AT IC F L U I D L E V E L
W
E
-
Fig. 12-9. Schematic diagram illustrating gas-lift terms (theoretical).
Working head, H,: Distance between the working fluid level and the midpoint of perforations in the bottom producing interval of the well. It can be expressed as pressure in psi, corresponding to the height of fluid column H,. Working lift, L,: Distance between the working fluid level and the fluid discharge point. Working submergence, S,: Vertical distance between the working fluid level and the bottom of the tubing ( L ,- L,). It is equivalent to the height of a column of well fluid which would exert a pressure equivalent to the working bottom-tubing pressure less the surface-discharge pressure and weight of the connecting gas column.
43 1
Total lift, L,: Sum of the working submergence ( S , ) and the working lift ( L,): L, = ( L , S,). Percentage working submergence: Working submergence ( S , ) divided by the total lift ( L , )expressed as a percentage: [ ( S , / L , ) X 1001.
+
TYPES OF GAS LIFT
Generally, the subsurface gas-lift installations can be divided into three types: (1) fluid discharge occurs up the tubing with gas being injected into the annular space; (2) fluid discharge occurs up the annular space with gas being injected down the tubing; and (3) there are two connected tubings, with gas being injected down one and fluids being lifted in the other. Fluid discharge up the tubing, as shown in Fig. 12-10.c, has the advantage of higher lift efficiencies, and uses lower volumes of gas per barrel of fluid lifted. This system requires a higher surface gas injection pressure than classes 2 or 3. Thus, there is a higher back pressure on the producing formation with resultant reduction in fluid flow from the formation. This system should not be utilized where the gas injected down the tubing can be lost to low-pressure thief zones in the formation. If the gas is injected down the tubing and fluids are lifted up the annulus (Fig. 12-10.d), larger volumes of 'injected gas are required compared to method 1, but the surface injection pressures are lower. This method is discouraged in the presence of H,S or air, because corrosion can occur in the casing. Corrosion problems are more serious in this case, because it is much more difficult to repair casing than it is to repair or replace tubing.
Fig. 12-10.Types of gas lift.
432
The method of utilizing two pipes, as shown in Fig. 12-10.e, is advantageous in cases (1) where the downhole formation pressure is near the downhole injection pressure, and (2) where holes in the casing or “thief zones” exist (above the bottom of the tubing) that would take a large percentage of the injected gas. The gas, which is injected down one tubing and lifted up the other, does not come in contact with the formation. Straight gas lift
Brown (1973) pointed out that the early methods of gas lift, which were commonly referred to as “U-tubing”, involved insertion of a short string of tubing into the wellbore and injecting gas at a pressure exceeding that at the bottom of the tubing. As the column of the fluid above the bottom of the tubing became lighter, the required gas injection pressure became lower. Inasmuch as high pressures were initially required to “kick-off‘’ the well, an operator had to install larger and more costly compressors than required by later methods. After development of gas lift valves (inserted in the tubing), the high “kick-off” pressures were no longer required. A smaller compressor was required during production as the fluid was aerated in the tubing in stages rather than in ope step. Due to its inefficiency, the “straight gas lift” or “U-tube” system required large volumes of gas for lifting fluids. The advantages of the U-tube system included: (1) few moving parts in the well; (2) the compressors and other machinery were located on the surface so that all repairs to the well equipment could be made on the surface; (3) handling of large volumes of produced fluid; (4) the system was readily adaptable to automation; (5) well cleanout work was seldom necessary because produced solids were carried to the surface; ( 6 ) the system was adaptable to directionally-drilled (“crooked”) wells; and (7) means of controlling back pressure under which the well is produced. Disadvantages of gas lift include: (1)a large capital outlay for initial expenditures for compressors and other surface equipment (however, salvage values also tend to be high); (2) the creation of oil-gas-water emulsions which required more effort to break; (3) aggravation of corrosion tendencies when H,S, water, or oxygen are present; and (4) deposition of paraffins and/or asphalt in flowlines, due to cooling effect of expanding gas. As a result of gas injection in a continuous flow gas lift system, the density of the fluid from the point of injection to the surface is reduced due to aeration. There is a lightening of the flowing pressure traverse or gradient as shown in Fig. 12-11. The pressure gradient is dependent upon the gas/liquid ratio ( G L R ) . With increasing GLR, the pressure gradient for a given fluid becomes lighter. Upon reaching a certain limit, however, any additional increase in gas injection results in a heavier gradient because of head loss to friction caused by higher velocities. This limiting gradient is called the “minimum gradient”. Shaw (1949) studied the relationship between the gas injection rate and fluid production rate (Fig. 12-12; see also Fig. 12-4). He noted that there was little or no flow until the pressure of gas injected was sufficient to lift the wellbore fluids. The
Static Fluid L e v e l , S.F.L. Flowing Wellhead Tubing P r e s s u r e PRESSURE
433
-
C- Operating Casing P r e s s u r e . P,
-
\\
Flowing Gradient T r a v e r s e Above The Paint of Injection
\ Point of Injection
A/
Static Gradient, G s
Differential Preasure A c r o s s The
Valve
SBHP Drawdown
Fig. 12-11. Example of static and flowing pressure gradient curves for a continuous flow gas lift. (Courtesy of Macco, 1966a, p. 3, fig. 1.)
6,
I
NO FLOW
s’c----,
I
1
---.‘-.
1
I
-
FLOW RATE, b b l / d a y
Fig. 12-12. Relationship between the gas output and production for a continuous-flow gas lift system. (After Shaw, 1949; courtesy of Texas Eng. Exp. Star.)
434
maximum capacity represents the maximum amount of fluid that can be produced per volume of gas injected. Beyond this point the efficiency lost due to increasing friction is greater than that gained by a further reduction in slippage. After reaching maximum capacity, further increase in gas injection reduces the fluid flow rate until the point of no flow as shown in Fig. 12-12. This occurs when the flowing gas pressure at the bottom of the tubing becomes equal to the reservoir pressure (corrected for the static head differential between the bottom of the tubing and the perforations). Design of straight gas lift system
The initial steps in the design of continuous flow gas lift systems include determination of (1) the point of gas injection, (2) gas volume required, and (3) the injection pressure. The gradient curve is then prepared using the following data: (1) desired fluid producing rate, BFPD (bbl/day); (2) size of tubing; (3) water/oil ratio, W O R ; (4) gas/liquid ratio, GLR (cu ft/bbl); ( 5 ) flowing tubing pressure at the surface, pwh(psig); (6) static bottomhole pressure, SBHP (psig); and (7) the well productivity index, J (bbl/day/psig of drawdown; the latter is equal to the static pressure minus the flowing pressure). In preparing the gradient curve, the flowing gradient traverse is plotted below the point of injection using the flowing bottomhole pressure, FBHP; production rate, BFPD; gas/liquid ratio, GLR; and water/oil ratio, W O R (Fig. 12-13). The flowing gradient traverse above the point of injection is plotted next starting with the flowing wellhead tubing pressure, using a reasonable GLR. The point of injection is where this traverse intersects the traverse below PRESSURE
b
1 Flawing Gradient Traverse Below The Point o i l n p c t i a n ( B a s e d o nthe Formation
0 FBHP
SBHP
Fig. 12-13. Pressure gradient curve. This gradient is dependent upon the gas/liquid ratio of the well. (After Macco, 1966a, p. 4, fig 3.) SBHP = static bottomhole pressure.
435 Flowing Tubing Wellhead Pressure, P,h
r 1-
PRESSURE R e q u i r e d Injection G a s P r e s s u r e
T r a v e r s e Above The Point of Injection ( T o t a l G L R = F o r r n a t i o n t InjeCtiOnGaS)
Point a i Injection
T r a v e r s e B e l o w The P o i n t o f l n j e c t i o n ( F o r m a t i o n Gas Only)
Diiie rential Press"*= A c r o s s The
0
I
FBHP
SBHP
Fig. 12-14. Plot of injection gas pressure gradient. (Courtesy of Macco, 1966a, p. 5, fig. 4.)
(Macco, 1966a). In Fig. 12-14, the injection gas pressure gradient is plotted, using a sufficient pressure differential from casing to tubing to provide the injection gas required. Normally, this differential varies from 40 to 60 psig, depending upon the type of valve used. The following information can be obtained from Fig. 12-14: (1) injection gas pressure at the surface, (2) point of gas injection, and (3) injection
GLR. Valve spacing is determined next. It is necessary to keep in mind that productivity index, J , data are not always reliable and, thus, the point of injection is better determined from previous field experience in many cases. The second method assumes a percentage drawdown (e.g., 50%), taking into consideration the bottomhole pressure decline, increase in water cut, and fluid load. The maximum point of injection and/or deepest point of lift is influenced by these three variables.
Spacing between gas lift valves The spacing between the gas lift valves is determined by a pressure balance between casing and tubing: DBV = Ap/Gr,, where DBV
= distance
(12-26) between valves, ft; A p = differential pressure between casing
P OI W
w
PRESSURE ____)
Producti Casing Pressure
Fluid Being Transferred Into Tubing Through Second Valve
Second V a l v e
Fourth Valve
-
PRESSURE
Productio
-*.
Casing Pressure
Gradient Shortly After Gas Begins t o Enter Tubing at the Second Valve ( ~ r r o w h d i c a t s sDirection Curve
will shirt) Second Valve Third Valve Fourth Valve Fluid Being Transferred Through Third
\
Fig. 12-15. Unloading operations. (a) Above the top valve. the fluid in the tubing is being aerated to the surface by injection gas a s the fluid in the annulus continues to be transferred into the tubing through lower valves. (b) Injection gas is entering the tubing through the top and second valve. The gradient becomes heavier (dotted line) immediately after the second valve is uncovered. The gradient lightens and shifts to the left (dark line), as more gas enters the tubing. (Courtesy of Macco, 1966a. p. 7.)
P W 4
438
and tubing at the valve, psig; and Gr,, = static fluid gradient, psi/ft. The schematic diagram illustrating the uncovering of the second valve is presented in Fig. 12-15. In order to derive the valve spacing equation (fluid-operated valves), it is assumed that the injection gas pressure is equal to the tubing pressure at the moment of
Fig. 12-16. Illustration for spacing equations for tubing pressure operated valves. (Courtesy of Macco, 1966b, p. 21, fig. 7.)
439
uncovering of the valve. The general equation for determining the location of the gas lift valve is as follows (see Fig. 12-16):
where DBV = distance between two valves, ft; pc at La = injection gas pressure at the valve above, psig; P w h = wellhead tubing back pressure, psig; DVA = depth of valve above, ft; SF = spacing factor, psi/ft; and Gr,, = fluid static gradient, psi/ft. Generally, with increasing opening pressure of the valve for any given injection gas pressure, the valves must be placed closer. Although with increasing wellhead back pressure a smaller liquid column is required to operate the valve, this pressure does not affect the valve spacing calculations. Spacing equations are derived on assuming that the tubing pressure at the valve above will not decrease below the opening pressure (OP)of the valve. If a fluid is unloaded with no inflow of fluids from the formation, the following method for determining spacing was proposed by Thrash and Brown (1965) to obtain a wider spacing of valves: Distance between valves is equal to 1 ft per increment of 1 psi of injection pressure. For example, if 600 psi injection pressure is available, valves are spaced 600 ft apart. The test block opening pressures are increased by an amount equal to the increase in casing pressure due to the weight of the gas column, in order to keep a constant differential pressure across each successively deeper valve. This method does not take into consideration the inflow from the formation and/or liquid fallback. These can supply the added pressure necessary to trip the top valve. The fallback, which is approximately 10% at the top valve, increases with depth. Inasmuch as this is usually enough to open the valve above, the system continues to unload. When the bottomhole pressure can deliver a fluid head into the tubing sufficient to open the fluid valve at its proposed depth, all remaining valves may be spaced by using the 1 ft/psi rule (Thrash and Brown, 1965). To compensate for the fallback and feed-in from the formation, it is also possible to increase the tubing pressure to open the upper valve and unload the well. Continuous flow gas lift (unbalanced valves) graphical design After establishing the feasibility of gas lift, to design a continuous flow installation, the following information is required: (1) well depth; (2) tubing and casing size; (3) required wellhead pressure (determined by the surface flowline size and length-the produced fluids should be able to flow to the separators); (4) desired producing rate and percent water cut; (5) injection gas gravity; (6) injection gas pressure (operating gas pressure) and/or maximum volume of gas available; (7) well inflow performance relationship and static bottomhole pressure, BHP;(8) bottomhole temperature, BHT,and geothermal gradient (or surface flowing temperature); (9) oil, water, and solution gas gravities; (10) amount of solution gas; (11) formation gas/liquid ratio, G/L,and formation fluid gradient; (12) kick-off pressure, p k o ;and
440 PRESSURE. D s i a
TEMPERATURE,
'F
Fig. 12-17. An example of graphical design for continuous flow gas lift (unbalanced valves). (See Example 12-1.)
(13) static kill-fluid gradient (minimum fluid gradient for the reservoir fluid plus gas to stop flow). The flowing bottomhole pressure is determined for the desired flow rate using the inflow performance relationship. The formation fluid pressure gradient is plotted using the flowing bottomhole pressure. The maximum operating gas pressure at the surface, p,,, is then plotted and, using the gas gravity, the flowing injection pressure in the casing is drawn (Fig. 12-17). The point of gas injection may then be determined as the depth where a difference of about 100 psi exists between the formation-pressure gradient line and the casing flowing injection-pressure gradient line. The actual well flowing-pressure gradient line can be drawn by connecting the wellhead pressure to the point of gas injection on the formation-pressure gradient line. Knowledge of the surface wellhead pressure, tubing size, well flow rate, depth to the point of gas injection, and water cut enables the computation of the required gas/liquid ratio, G / L , using the applicable vertical flow correlation (as described in Chapter 9). The injection gas rate, qig, required for the system is then computed by: qig=
[(G/L)vfc-(G/L)flql
(12-28)
where ( G / L )vfc = G / L determined from the vertical flow correlation, scf/bbl;
(G/L),= formation G / L , scf/bbl; and q, = well liquid (oil plus water) flow rate, bbl/day. The casing static-pressure gradient line is drawn using the gas-pressure gradient and the surface kick-off pressure. The design tubing-pressure gradient line represents
441
a safety factor on the actual flowing-pressure gradient line. It is drawn from the gas injection point on the formation-pressure gradient line to a single surface pressure of ( P w h + 200) or ( Pwh 0.2pc0), whichever is greater. The valve spacing determinations are then made by plotting the static kill-fluidpressure gradient lines on the graph beginning at the pwhand extending down the hole (Fig. 12-17). The first line is drawn from Pwh to the casing pressure gradient line, whereas all the successive lines are drawn from the tubing design pressure gradient line to the flowing injection pressure gradient in the casing. The valve depths are determined as the points of intersection with the flowing (or static, in the case of the first valve) injection-pressure gradient in the casing. At the valve depth, the pressure read on the tubing design-pressure gradient line is the valve opening pressure in the tubing, pvt0,whereas the pressure on the casing flowing injection pressure gradient line is the valve opening pressure, p v 0 , in the casing. The valve can be selected for the desired spread or valve sizes can be
+
DEPTH, f t
OPENING PRESSURE OPPOSITE VALVE
Fig. 12-18. Dome pressure determination. (Modified after Thrash and Brown, 1965, p. 8, chart No. 2; courtesy of Otis Engineering Corporation, Dallas, Tex.) Determination of operating pressures (average) of Otis balanced gas-lift valves. Assumptions: (1) natural gas charge = 0.60; (2) deviation is considered; (3) sp. gr. of lift gas (SG air = 1) = 0.65; (4) wellhead temperature = 100OF; (5) effective temperature = [70°+(1.6 xdepth/100)]; (6) average temperature in the casing = (T, + T v ) / 2 = (100+ T.,)/2. Sample calculation: Given dome setting at 80”F= 700 psig and well temperature at the valve = 200”F, determine charge in the dome at 200OF. (1) Select temperature on the right-hand vertical side (ZOOOF), (2) select dome setting at 80°F on the diagonal lines (700 psig), (3) move horizontally to the left from the temperature (200°F) until intersecting charged dome diagonal line (700 psig), (4) move vertically downward from this point to the bottom of the chart, and (5) read the opening pressure opposite the valve - 903 psig. (Note: this chart should not be used for a valve having a “spread”.) (Alsosee Kirkpatrick, 1955a.)
442
specified in some instances due to non-availability of other sizes or because of other limitations such as casing-tubing clearance. A low spread is generally desirable. The dome (or bellows) charge pressure at the valve depth, P d , , can be calculated using the following relationship: (12-29) where R ( = A J A , ) , can be determined from Table 12-11 for different valve and port sizes. The dome pressures at 60°F rack conditions can be determined using Fig. 12-18. The examples below serve to illustrate this procedure. The extra pressure exerted by a gas column in a well can be determined from Fig. 12-19.
PRESSURE, psi/lOOO f t
Fig. 12-19. Estimation of the weight of gas column. (Modified after Thrash and Brown, 1965, p. 13, chart No. 3; courtesy of Otis Engineering Corporation, Dallas, Tex.) Based on gas column weight/1000 ft depth; wellhead temperature = 100'; geothermal gradient = 1.6°F/100 ft; surface temperature = 70'F; effective temperature = 70°F+depthx 1.6°F/100; average temperature = [100+(70+(1.6 xdepth/100))] : 2; A p (corrected) = A p (chart) [(chart avg. temp., 'R)/(actual avg. temp., OR)]. Pressure at the valve = surface operating pressure+ gas column weight. Procedure: (1) obtain SG of gas column; (2) mark surface pressure on the ordinate and proceed horizontally to the right to the proper SG line; (3) proceed vertically downward from this intersection to obtain the gas column weight in psi/lOOO ft; and (4) multiply the value obtained in (3) by the valve depth in thousands of ft. (Note: also see Brown and Lee, 1968.)
Example 12-1 Given: Static BHP
= 2000
psi; oil production desired = 400 bbl/day with a 50% water
TABLE 12-11 Macco gas-lift valve specifications: a = fluid operated, b = casing pressure operated (courtesy of Macco Oil Tool Lks i g na t r o n I
onvenlmnal
Rerrievabls
I
Pvrt Sizes
Bellows Area
sq. In.
N
~
Port Slze(ln.)
Fl ui d Ope r a t e d
1-0.. ~nc.)
port ~
Area 2)
(1"
~
~ CaSl"8
Efiecr Factor(%)
~
I
(l -A p / A b )
Nominal Snee (i",)
Port
Casing
Area (in.2)
Effecl laclor(%)
(I-Ap/~b!
Nominal SIZC
Port
Area
Effect
(&".I
(1n.Z)
Fa ct o r( % )
Casing ( l -A p / A b l
CMO-FS
CMOFS-AK CMOFS- BK
0.1246
118
0.0123
10. 8
0. 9013
118
0.0123
LO. 8
0.9013
'-10164
0.020
0.8393
CM1-FS
CMLFS-CK
0. 3189
3/16
0. 0302
10. 5
0 9053
118
0.0123
4.0
0.9614
. 114 5/32
0. 0554 0.020
0.8263 0.9373
CM2-FS
CMZFS-RC
0.7096
114
0 0554
8.5
0. 9217
118
0.0123
1. 7
0.9826
-
(l-*,/nb)
0.0947
0.9774
0.0226
0. 1123
0.9899
0.0101
1
0.0797
5/32 7/32
0.0394
Ap/Ab
12.7
0.8877 0.9718 0 9443
0.020
I
Mi"irn"rn
Apl Ab
. 5/16
Other I:Maxirnurnl Nominal Port Size(&".)
> 9/32 118 10164 I3lbl 11/64 114 318 7/61 8/64 9/64 10164 11/64 12/64 13/64 14/64 114
Port
Area (m.2) 0.0652 0.0123 0.020 0.0340 0.0394 0.0554 0.1142 0.0098 0.0123 0,0161 0.020 0.0243 0.0302 0.0340 0.0394 0.0554
Tubing illcct Factor(%)
(I-Ap/Ab)
25.7 4.0 6.7 11.9 14. I 21.0 19.2
0.7957 0.9614 0.9373 0.8913 0,8765 08262
1.4
0.9862 0.982b 0.9770 0.9718 0.9658 0.9571
1.77 2. 36 2.9 3.55 4.45 5.05 5.9 8.48
0 8191
0.9520
0.9444 0.9218
0.2043 0 . 0386 0.0627 0 . Lob7 0. 1235 0.1718 0 1609 0.0138
0.0174 0.0210 0. 0282 0.0142 0.0426 0.0480 0.055b 0.0782
-
444
cut; depth of perforations = 6500 ft; static kill-fluid-pressure gradient = 0.5 psi/ft; J = 5.33 (based on total production); B H T = 160'F; flowing temperature at the surface = 110'F; kick-off pressure = 1500 psi; maximum operating gas pressure = 1350 psi; gas gravity = 0.65; flowing wellhead pressure = 200 psi; formation G / L = 25 scf/ bbl; pressure gradient of reservoir fluid at G / L ratio of 25 (= 0.44 psi/ft); tubing size = 2 in.; casing size = 4.5 in.; specific gravity of water = 1.04; and oil gravity = 35'API. Assume 1-in. O.D. valves with 3/16-in. ports. For a continuous flow gas lift system design, determine the following for this well: (a) Valve depths assuming 1-in. valves. (b) Opening pressure in the casing for the valves. (c) Opening pressure in the tubing for the valves. (d) Dome pressures at the valve depths. (e) Dome pressures at 60°F rack conditions. (f) Gas injection rate required for the gas lift system. Solution:
Thus, pwf= p R- q , / J = 2000-800/5.33 = 1850 psi. Using eq. 12-13, for p k o = 1500 psi at the surface, the kick-off pressure at a depth of 5500 ft (any depth less than well depth can be taken) = 1680 psi. For pco = 1350 psi at the surface, the operating gas pressure at a depth of 5500 f t is equal to 1510 psi. The solution steps are as follows (see Fig. 12-17): (1) Draw the formation fluid-pressure gradient of 0.44 psi/ft parsing through the flowing BHP of 1850 psi. (2) Draw the casing static column-pressure gradient and the flowing injectionpressure gradient in the casing. (3) Place the gas injection point at the depth where a difference of 100 psi exists between the formation-pressure gradient line and the flowing injection-pressure gradient in casing line. This occurs at a depth of 5500 ft. The tubing pressure at this point is 1410 psi (as read on the formation-pressure gradient line). (4) Use Fig. 12-20 to get a match for G / L ratio in the section of the tubing above the point of gas injection. Iterate for G / L to satisfy the condition of pwh= 200 psi, and pwf= 1410 psi at a depth of 5500 ft. On doing this, the best fit is obtained for a G / L = 400 scf/bbl. (5) pwh 200 = 400 psi and p w h + 0.2pco= 200 + (0.2)(1350) = 470 psi. Use the greater value, i.e., p w h = 470 psi. Draw the tubing design-gradient line from p = 470 psi at the surface (zero depth) to p = 1410 psi at a depth of 5500 ft. (6) Draw top valve line from p w h = 200 psi with a gradient of 0.5 psi/ft. (7) Draw further lines having a pressure gradient of 0.5 psi/ft between the design
+
445 4
8
12
16
20
24
28
VERTICAL F L O W I N G PRESSURE G R A D I E N T S
-
(50% OIL 50% WATER) Tubing Size 2 in. I.D. Producing Rate 400 BblsJDay Oil API Gravity 35" API Water Specific Gravib 1.074 Gas Specific Gravity 0.65 Average Flowing Temp. 120°F
Fig. 12-20. Vertical flowing pressure gradient for Example 12-1. (After Brown, 1973, fig. C-178; courtesy of Petroleum Publishing Company.)
446 TABLE 12-111
Dome valve pressure data for Example 12-1
Temp. (OF)
131 140 145 148 151 152
(C)
(4
( ft)
(b) Open. press. in csg (Psi)
Open. press. in tbg (Psi)
Dome press. at BHT (Psi)
(el Dome press. at 60°F (Psi)
2800 3800 4500 5000 5400 5500
1430 1460 1480 1490 1500 1510
950 1120 1250 1320 1390 1400
1385 1428 1458 1474 1490 1500
1202 1218 1232 1240 1246 1253
(a) Valve depth
tubing-pressure gradient line and the casing injection operating-pressure gradient line. (8) Stop when next valve is below the injection point. (9) Determine the valve depth, pvto, pvco, and temperature for each valve from Fig. 12-17. For example, for the first valve, the depth to the valve= 2800 ft, p,,(, = 950 psi, pvco= 1430 psi, and temperature from the temperature gradient line = 131°F. (10) From Table 12-11, for a l-in. O.D. valve with a port size of 3/16-in., R = A , / A , = 0.0947 and (1 - R ) = 0.9053. For valve No. 1, using eq. 12-29, the dome pressure at the valve depth is equal to: pd,= 1430 (0.9053)
+ 950 (0.0947) = 1385 psi
Similar calculations are performed for the other valves (No. 2-6). (11) Use Fig. 12-17 to determine the dome pressures at 60°F rack conditions. The answers (a) through (e) are listed in Table 12-111. (12) Total liquid production = 400 bbl/day oil + 400 bbl/day water = 800 bbl/ day. G / L = 400 scf/bbl was obtained for the actual tubing flowing-pressure gradient line and the formation G / L = 25 scf/bbl. Therefore, the gas injection rate required is 300 Mscf/day [ = (400 - 25) X 8001. Example 12-2 Gioen: Gas lift is proposed in a 7300-ft deep well with a formation G / L = 50 scf/bbl. A wellhead pressure of 160 psig is to be maintained for allowing flow through surface lines. The maximum amount of gas available is only 750 Mscf/day. Because of existence of sufficient compressing capabilities, gas operating pressure is not a limiting factor. The productivity index, J , of this well is 2.5 (based upon total liquid; assume linear relationship) and the static bottomhole pressure = 2200 psig. Additional data available is as follows:
441 PRESSURE, psig
000
400 I
l
12QO l
pco&
-TI
1600
'
- 2000 -
c
I3000-
! i
w 0 4000-
5000t 6000
8000'
100 120 140 TEMPERATURE, O F
Fig. 12-21. Graphical solution for Example 12-2.
Kick-off pressure at surface = 1400 psig; oil production required = 500 bbl/day, 50% water cut; gas gravity (air = 1) = 0.65; surface temperature = 100°F; bottomhole temperature, BHT, at a depth of 7300 ft = 160°F; reservoir fluid-pressure gradient (with G / L = 50) = 0.40 psi/ft; static kill-fluid-pressure gradient = 0.5 psi/ft; tubing size = 2 in.; casing size = 4.5 in.; valves are 1-in. O.D. with 3/16-in. port size; specific gravity of water = 1.05; oil gravity = 35"API. For an optimum continuous flow gas lift design for this well, determine: (a) The valve depths. (b) Opening pressure in the casing for the valves. (c) Opening pressure in the tubing for the valves. (d) Dome pressures at the valve depths. Solution (refer to Fig. 12-21):
pWr= jjR - q / J
= 2200 - (1000/2.5) = 1800 psi
(1) Draw a line for the formation fluid pressure gradient of 0.4 psi/ft, passing through pwr= 1800 psi at a well depth of 7300 f t ( G / L = 50 scf/bbl). (2) Use vertical flow correlation for 2-in. tubing and 1000 bbl/day liquid production rate. Maximum gas available = 750 Mscf/day, or 750 scf/bbl. Plan to use all this gas to maximize economics, i.e., to use the least number of valves. Inasmuch as formation G / L = 50 scf/bbl, the flowing tubing G / L = 750 + 50 = 800 scf/bbl. Thus, G / L = 800 must be used on vertical flow correlation. pwh= 160 psig. At any depth, e.g., 6500 ft, pressure = 1400 psi from correlation. Using this, draw the actual flowing gradient line. The formation-pressure gradient and the actual flowing-pressure gradient lines intersect at a depth of 6100 f t (at 1320 psi).
448
TABLE 12-IV Example 12-2 data on valve opening pressure in the casing and tubing Valve No. 1
2
3
4
5
6
7
A. Valve depth (ft)
2700
3750
4550
5150
5600
5975
6100
B. Opening pressure in casing (psig)
1310
1340
1370
1390
1405
1415
1420
C. Opening pressure in tubing (psig)
805
965
1085
1175
1250
1300
1320
(3) At 6100 ft, flowing injection gas pressure = 1320 + 100 = 1420 psig or 1435 psia. This pressure, p,,, is equal to: p v = p s exp[(0.01875GgD)/(2T)] (see eq. 12-14), where p\ = pressure at the surface. Use Gg = 0.65, pc = 668 psia, and = 375"R. Assuming pa\ = (1435 + 1315)/2 = 1375 psia and T = (100 + 150)/2 = 125°F or 585"R, then pr = 1375/668 = 2.06; Tr= 585/375 = 1.56; and 2 = 0.84. Thus p s = 1435 exp[( - 0.01875 X 0.65 x 6100)/(0.84 X 585)] = 1233.5 psia or 1219 psig ( = pco). (4) pko= 1400 psig. Assume p = (1415 + 1535)/2 = 1475 psia, pr = 1475/668 = 2.21; and 2 = 0.83. pdeplh = 1415 exp[(0.01875 X 0.65 X 6100)/(0.83 X 585)] = 1649 psia or 1634 psig. The casing static column-pressure gradient line and flowing injection in casing pressure gradient line can now be drawn (Fig. 12-21). (5) P u h + 200 = 160 + 200 = 360 p i g . puh + 0.2pc0 = [160 + (0.2)(1219)] = 404 psig. Inasmuch as the larger value is to be selected, the design pressure would be 404 psig. Then the design pressure gradient can be drawn (Fig. 21-21) and the system design completed (see Table 12-IV).
TABLE 12-V Example 12-2 data on dome valve pressure
B Opening pressure in casing (psig)
C Opening pressure in tubing (PSM
D Dome pressure at valve temperature (PW
2700 3750 4550
1310 1340 1370
805 965 1085
1262 1304 1343
5150 5600 5975
1390 1405 1415
1175 1250 1300
1370 1390 1404
6100
1420
1320
1411
A Valve depth (ft)
449
For a 1-in. O.D. valve and 3/16-in. nominal port size, R = 0.0947; (1 - R ) = 0.9053; pdt= p,,,(l - R ) +pvtoR= 0.9053p,,, + 0.0947 pvto.Table 12-V can then be prepared. Intermittent gas lift The intermittent gas lift system involves expansion of a high-pressure gas bubble or slug ascending to a low-pressure outlet (5ee Fig. 12-22). Complete pressure and volume expansion control of gas entering the tubing is handled by a valve with a large port. It either regulates the lift of the accumulated fluid head above the valve with a maximum velocity to minimize slippage, or controls liquid fall back by fully ejecting it to the tank with minimum amount of gas. Intermittent gas lift is often used in conjunction with a surface time controller (intermitter). In general, the intermittent gas lift method is used on wells producing low volume of fluid and having high productivity index ( J ) and low bottomhole pressure ( B H P ) , or low productivity index and high bottomhole pressure. As pointed out by Neely et al. (1981), intermittent gas lift is an excellent choice when an adequate, good-quality, low-cost gas supply is available and it is intended to lift a relatively shallow, high GOR, low J , or low BHP well with a bad dog-leg and, possibly, producing some sand. The efficiency of the intermittent gas lift system depends primarily on the degree of injection gas breakthrough and liquid fallback, which should be minimal. A greater drawdown pressure can be achieved using an intermittent gas lift system. Production rates below 250 bbl/day and J values of 0.5 bbl/day-psi are considered to be in the range of intermittent gas lift (Thrash and Brown, 1965). Design of intermittent gas lift system Maximum production rate by intermittent gas lift is affected by the following factors: (1) injection pressure; (2) depth of lift; (3) tubing capacity; (4) injection gas volume; (5) injection gas breakthrough and liquid fallback; (6) wellhead tubing back pressure; (7) productivity index; (8) bottomhole pressure buildup characteristics; (9) well conditions, such as presence of emulsions and paraffins; and (10) gravity of crude oil. Productivity index, bottomhole pressure buildup, the characteristics of intermittent lift system (cycles/day, and volume of fluid lifted per cycle) determine the maximum producing rate for the intermittent well. Injection cycle time can be calculated as follows: (1) size load of gross liquid, L, in bbl/cycle; (2) cycles/day = (bbl/day)/(bbl/cycle) = cycle frequency; (3) 1440 min/day: cycle frequency = operating cycle, min/cycle = 1.5 minutes (L)/lOOO ft, where L = depth of lift, ft. For determining the maximum volume of fluid produced per cycle, Macco (1966b) proposed the following empirical method: (1) The starting slug volume (in bbl) at the top valve, V,, is equal to:
V, = [
(12-30)
e
0
(A)
Fluid F r o m Formation Has Built Up Above The Bottom Fluid Operated Valve.
(B)
P r e s s u r e Due To Fluid Head Opens The Bottom Valve And G a s Is Injected Under The Slug.
(C)
Liquid Slug Being Lifted By Injection G a s Entering The Tubing Through T h r e e Valves Below The Slug.
(D)
Liquid Reaches The Surface A s Injection Gas E n t e r s The Tubing Through Open Valves.
Fig. 12-22. Intermittent gas lift cycle of operation for tubing-operated valves. (Courtesy of Macco. 1966b. fig. 1 , p. 4.)
451
where I/, = starting slug volume, bbl; Q,, = tubing capacity, bbl/ft; Gr,, = static gradient of well fluid, psi/ft; and pol = opening pressure of the first valve, psi. (2) The produced slug volume, V, (in bbl/cycle) may be estimated from the following relationship: Vp = 0.90V,
(12-31)
(3) The estimated maximum production rate, Q,, in bbl/min is equal to:
Q,
=
VpFc,
c,
(12-32)
= cycle frequency, min/cycle. where The produced slug volume, V,, calculated using eq. 12-31, is used as the producing slug volume attainable at each succeeding depth. The basic assumption is that all formation gas is produced between cycles and the gradient is essentially that of the fluid alone. The starting slug volume, V,, increases with depth owing to an increase in valve operating pressures, p,,. The difference between the starting and produced slug volumes is assumed to be the volume of “fallback” (Macco, 1966b). Gas is injected at regular intervals by the intermitter in the case of intermittent gas lift. The intermitter is a motor valve which operates by a connecting timer device that permits selective cycling with controlled gas injection into the casing annulus. The cycle time is adjusted in accordance with the fluid fill-in rate from the producing formation into the wellbore. When the surface intermitter is not used, a valve is required that is somewhat more fluid sensitive than the surface valve and requires a built-in “spread”. Sensitivity of the valve can become a disadvantage in some cases. For example, problems arise when the wells must be lifted into restrictive flowlines or against surface chokes. In some other cases, however, this valve has definite advantages, e.g., in some rotative compressor systems. The main difference between the intermittent lift and continuous flow system is that in the former a liquid slug must build up before gas is injected underneath the slug. The slug is thus propelled to the surface. There is a great variation in the flowing bottomhole pressure during the period between the injection of gas underneath the slug and delivery of the slug at the surface. The use of one large-ported valve to accomplish the lift cycle may result in minimum amount of required gas and maximum fluid recovery. This operating valve is usually the deepest valve exposed to gas in the casing annulus. The valves above, whch are referred to as the unloading valves, enable unloading of well fluids using the existing gas pressure. In the case of intermittent gas lift in water-drive reservoirs, the lowest valve is not always the operating valve. Other valves are also installed for future use, because the point of operation moves down as the well pressure declines. The increase in water cut may also necessitate working further down the well (Macco, 1966b).
452
P I o TURN N TUBING TO CLOSE
e
43-43 -TUBING
Fig. 12-23. Jet collar. (After Brown, 1973, fig. 8.8, p. 184; courtesy of the Petroleum Publishing Company.) GAS LIFT EQUIPMENT
Jet collars and orifice inserts Two types of artificial chokes were developed in an effort to reduce the high initial “ kick-off” pressures, required in “straight gas lift” systems: (1) orifice chokes, whch could be placed in tubing; and (2) jet collars, which could be opened and closed by turning the tubing from the surface. The jet collar (Fig. 12-23) consists of two tight-fitting concentric tubes, with a hole drilled through each one of them. On “lining-up” the holes, the system is opened and fluids (oil, water, and gas) flow from inside the tubing to the outside. Upon closing, the holes are not “lined-up’’ and gas cannot communicate or pass through the openings to the tubing-casing annulus. Jet collars, however, often stick and cannot be completely closed. When this occurs the tubing has to be pulled and the valve repaired. In order to reduce the kick-off pressures, holes were placed in the tubing string at selected intervals. This method was workable; however, the holes in the tubing could not be closed or plugged later and, thus, there was a loss of injected gas at each opening. Using a tool which was run on wireline down the tubing string, the holes were shot (or punched) at the desired depths. Inasmuch as after the fluid level was lowered below a certain point the orifice would only circulate the gas without helping to lift the fluids, it was important to keep the holes small. As a result, large volumes of gas were required to lift the fluid and the gas usage was inefficient. The backflow and turbulence created at each orifice aggravated erosion and corrosion, which enlarged the orifice further. This resulted in the additional reduction of the efficiency.
453 ---c
FLOW LINE
t
GAS I N
-TUBING-
FLAPPER TYPE SPRING
\
Fig. 12-24. Flapper type spring valve. (After Brown, 1973, fig. 8-11, p. 185; courtesy of the Petroleum Publishing Company.)
Kick-off values With the development of kick-off valves, the upper holes in the tubing string could be closed and the volume of required cycled gas was reduced. The upper holes could also be reopened in the case where it was necessary to lift the fluid again at that level or kick-off the fluid again if the well was shut-down. After an initiation of flow, the valve was designed to close. The principle of operation of velocity-controlled valves, which were popular in the 1930s, is that the increased flow velocity of gas and liquids would close the valve (Fig. 12-7). As the upper valve was closed, the lower valve exposed to low fluid velocities remained open until reduction in velocity. The valve was not retrievable by wireline and was generally set in the middle of the tubing string. A second type valve developed during the same period was the flapper type (spring) valve (Fig. 12-24). This valve incorporated a flapper type spring that operated under a differential pressure of 15-20 psi. Although, in general, these valves operated well in the case of continuous flow, they often experienced mechanical problems that resulted in their replacement by the flow type valves. Flow type values The flow type valve (Fig. 12-25) is similar to the orifice valve, but the valve closes as it is uncovered by fluids and remains closed until fluids again cover it. This type of valve reduces the volume of circulating gas that does not contribute to the lifting of fluids. The flow valves gave a greater control over “kick-off‘’ pressures than the previously-designed valves. Flow valves include (1) spring-loaded differential valves (Fig. 12-25), (2) mechanically-controlled valves, and (3) pressure-operated valves.
454
FLOW LINE
Fig. 12-25. Spring-loaded differential valve. (After Brown, 1973, fig. 8-15, p. 186; courtesy of the Petroleum Publishing Company.)
Wireline retrievable valves have been continuously improved since the 1930s. Initially, the valves were set in center of the tubing and replacement of the valve necessitated a costly tubing pulling. Tubing with offset mandrel pockets were developed later to hold these valves. This allowed replacement of the valves in the tubing string by a much less costly wireline pulling job. Differential valves According to Brown (1973) spring-loaded differential-pressure valves (Figs. 12-25 and 12-26) were introduced in 1934. This type of valve incorporates a differential-pressure spring to hold the mechanism in an open position (Fig. 12-26). The differential-pressure for each valve can be set individually and ranges from 100 to 150 psig. When the pressure falls below the set value, the valve closes. Brown (1973) pointed out that due to mechanical problems of sticking, etc., this type of valve requires a great deal of attention. The initial operation of the lift system can be erratic until the proper valve settings are determined from experience. This type of valve is best suited for continuous flow systems and is not applicable to intermittent lift. Mechanically-controlled valves A mechanically controlled valve was developed in the mid-1930s (see Fig. 12-8) for wells that did not need continuous help in lifting fluids, but rather only periodic assistance. A timing device activates the mechanism which pulls the wireline operated from the surface and, thus, opens the valve down in the tubing. Often, a gas-driven piston at the surface is connected to the wireline. A rod is positioned at the bottom of the wireline so as to pass in and out of the valve. On entering the valve, this rod pushes against the seated ball, and opens the valve to fluid flow. As
455
Fig. 12-26. Bryan differential valve. (After Brown, 1973, fig. 8.16, p. 186; courtesy of the Petroleum Publishing Company.)
the rod is pulled back, the ball is reseated and the valve closes. Brown (1973) pointed out that although this type of valve was efficient, the wireline would often break due to corrosion or metal fatigue. The wireline also could cause other problems such as wearing a hole in the tubing. Development of pressure-type valves eliminated much of the need for surface-operated mechanically-controlled valves. Pressure valves
The pressure-differential (or specific gravity) valve was introduced in 1940. A flexible diaphragm in this type of valve is open to the tubing-casing annulus pressure on one side and tubing pressure on the other (Fig. 12-27). Figure 12-27.A shows the valve in an open position whereas Fig. 12-27.B shows it in a closed position. As the fluid column height above the valve exceeds 10-16 ft, sufficient pressure is exerted to hold the valve open. Thus, gas injected down the tubing will circulate out of the orifice and aerate (lighten) the fluid column above the valve. The valve closes when the fluid column height drops below 10-16 ft. As a result, gas will no longer circulate through that orifice and is forced down to the next lower valve. The valve can open again at a later time if the fluid level rises and the valve is again covered by fluid. Brown (1973) pointed out that the valve has proven to be an
456
PROWCTION
(A) Valve closed
(B) Valve open
Fig. 12-27. Specific gravity differential valve. (After Brown, 1973, fig. 8-19, p. 186; courtesy of the Petroleum Publishing Company.)
excellent one; however, it tends to be bulky and is applicable primarily to the continuous and not intermittent lift systems. Pressure-charged valves
There are two widely used types of pressure-charged valves, the rubber element and bellows types. In principle these valves operate similarly. The opening and closing pressures of bellows-type valve are different, w h c h is referred to as “spread”. Inasmuch as different manufacturers have included various modifications and improvements in the valve design, the design of these valves varies widely. As pointed out by Thrash and Brown (1965), the force holding the valve open (Fig. 12-28) is equal to: (12-33)
f c = pd
whereas the force acting to open the valve is equal to:
F,
= Pc( A ,
-A,)
+ P,A,
(12-34)
457
Fig. 12-28. Bellows valve schematic. (After Thrash and Brown, 1965, fig. 5 , p. 21; courtesy of Otis Engineering Corporation.)
Inasmuch as in a balanced condition the forces are equal to each other:
where Pd = pressure in the dome, psia; A , = area of piston or bellows, in2; P, = pressure necessary to open the valve, psia; A , = area of valve seat, in2; and P, = pressure in tubing, psia. Solving for P,: (12-36) and substituting R for A . / A , : (12-37) where R = ratio of the area of the seat of the valve to that of the bellows. In the case of intermittent gas type valve, it opens widely when the operating pressure is reached and gas is injected underneath the liquid column (above the valve) lifting the fluid column in the tubing-casing annulus. Consequently. as pointed out by Thrash and Brown (1965), it is important to have a large port size in the gas valve in order to allow a large volume of gas to pass through the orifice as quickly as possible. As a general rule, the minimum opening for such a valve is 1/2 in. The exact size of the valve seat can be selected to yield the desired pressure spread.
458
Estimation of horsepower required to compress gas
In order to compare gas-lift with other methods of lift, as discussed in other chapters, the engineer needs to determine the approximate horsepower requirements in order to estimate the gas compression costs. The following is a simplified approach developed by Gilbert (1954). As he pointed out, it should be used as an approximation only. The temperature of the gas injected into the well (power gas), due to the low specific heat of the gas, is controlled by well temperatures at points of application. The gas-rate requirements are usually estimated at standard conditions. The gas horsepower required for the gas-lift is:
H , = p , Q , ( k / ( k - l))[(p2/pl)(k-1)'k - 11 :550
(12-38)
where p , = discharge pressure, psfa; p 2 = bottom of tubing pressure (initial), psfa; Q, = volumetric rate of flow of gas, cu ft/sec, measured at p , . For a wet gas with a k = 1.25, this formula may be stated as:
[
H p = 0.223M ( p 2 / p 1 ) o ' 2- I]
(12-39)
where M = Mcf/day at 14.7 psi. According to Gilbert (1954), this is the horsepower upon whch quotations should be based with the given (1) pressure ratio, (2) input gas temperature, and (3) input pressure. It may be from 20 to 40% lower than the manufacturers's brake-horsepower ratings depending upon the deviation from Boyle's law, the auxiliaries used, and the overall plant efficiency attained. Figure 12-29 shows the relationships discussed.
0.25-
0.20
-
0.15-
0.00'
I
1.5 2
I
I
4
I
' I "
6 810
I
I
20
'
I 30
'50
P2/Pl RATIO
Fig. 12-29. Estimated gas-horsepower requirements for gas lifting. (After Gilbert, 1954, fig. 22, p. 141.) k -1.25; p z = compressor outlet pressure, psia; p 1 = compressor input pressure, psia.
459 CONCLUSIONS
Gas lift, where applicable, offers many advantages over other lift systems. Several of these advantages were listed by Thrash and Brown (1965): (1) relatively low installation and maintenance costs where high-pressure gas is readily available, (2) low lifting costs, (3) ability to handle a wide range of production rates without great changes in production equipment, (4) simplicity of operations, (5) ability to handle small volumes of sand and other erosive materials, (6) easily adapted to central operation and automated controls, (7) adaptable to deviated holes, (8) cleanout of the wellbore is seldom necessary, and (9) there are few moving parts that can breakdown. The disadvantages of the system become pronounced when there is no adequate volume of high-pressure, low-cost gas - compressors must be installed and maintained. In addition, corrosion is augmented and paraffin and/or asphalt deposition in flowlines is intensified for a gas-lift system as compared to other lift systems. Thrash and Brown (1965) offered the following thirteen rules of thumb for gas-lift installations: (1) Gas volumes required: (a) continuous flow - 150 to 250 scf/bb1/1000 f t of lift; (b) intermittent flow - 250 to 400 scf/bb1/1000 f t of lift. (2) Pressure required: 100 psi per 1000 ft of depth; up to 800 psi with a minimum pressure of 300 psi. (3) Allowable continuous flow depth (ft) = [( pc - p,)/0.15], where pc = casing pressure, psia, and pt = tubing pressure, psia. (4) Slug velocity for intermittent lift = 1000 ft/min. (5) Minimum time (min) required for intermittent cycles = 1.5 X depth. (6) Gas volume used per cycle, Q,/cy, on intermittent lift = (drop in pressure on casing, psia) (storage volume in casing, cu ft) : 15. (7) Approximate maximum rates possible from tubing sizes under normal lift conditions (continuous lift) are as follows: Tubing size (in.)
Maximum volume (bbl/day)
3; 3
2
5000 4000 3000 2500
1;
1000
1; 1
600 350
2;
4 8 ) Productivity index, J, which is equal to:
J
= kh/p.,
B,
where k = formation permeability, darcies; h = formation thickness, ft; p o = oil viscosity, cP; B, = oil formation volume factor.
460
(9) Approximate production rate to change from intermittent flow to continuous flow under normal conditions for straight tubing: Tubing size (in)
Production rate (bbl/day)
3 2; 2 1; 1: 1
300 250 200 75-125 50-75 25-50
(10) Weight of the gas column, Wg= (2.5 X l o p 5 ) ( p , D ) where p , = casing pressure, psia, and D = vertical depth, ft. (11) Dome pressure charged unbalanced gas lift valve - valve set at 60°F will have a surface operating pressure at 60°F. The effects of gas column weight and tubing effect are essentially offset by the temperature effect. (12) Normal pressure in psi expected in a new well is equal to (0.465 X D ) , where D = vertical depth in ft. (13) Normally, temperature expected at a certain depth in a new well along Texas and Louisiana coasts (U.S.A.) is equal to [74"F + (1.6"F/100 ft) X D ft]. Gas lift is a proven method of artificial lift which requires a relatively high-pressure gas (minimum pressure of 250 psi). It is very flexible; for example, if designed properly, it may produce 60 bbl/day as well as 1100 bbl/day. The maximum efficiency is acheved when field operating personnel are properly trained to recognize (and report) the " trouble signs". Continuously checking the behavior of casing and tubing pressure is of utmost importance. The writers strongly recommend that interested readers consult the classical work of Mach and Brown (1980) for further details.
SOLUTION O F SOME FUNDAMENTAL PROBLEMS
Problem 1
The following average data are given for a well which is produced by continuous gas lift through the tubing: (a) bottom of tubing, 6120 ft, (b) pressure at bottom of tubing during gas lift, 825 psig, (c) pressure at top of tubing during gas lift, 105 psig, and (d) specific gravity of well fluid, 0.875. Compute the work of expansion and the number of cubic feet of gas (measured at 105 psig) required to lift each barrel of well fluid for (1)isothermal expansion with 100% efficiency; (2) adiabatic expansion with 100% efficiency; and (3) isothermal expansion with percentage efficiency assumed to be equal to the percentage working submergence.
461
Assumption-The increase in absolute pressure of gas in casing is 2.9% (not compounded) per 1000 f t of depth. Solution L,
=L -
p2 i- 15 105
s,
L - s, + ( pz + 15) [ 1 000
+ (105 + 15)
6120 - S,
1
x (0.029)
j(0.029)
+ 0.434 GS, = p, + 15
+ 0.434 x 0.875 S, = 825 psig
S , = 1858 f t
L,
= 6120 -
1858 = 4262 ft
Isothermal work of expansion: W = 332 pz log,,
= 3 3 2 X 120log,,
( 120 840)
= 33,700
ft-lb/cu ft
Adiabatic work of expansion: W=
Q=
144 x 1.27 x 120 [(7)!A?? 1.27 - 1 - 41,700 ft-lb/cu f t 0.27
]
349 G L, 349 X 0.875 X 4262 33,700
38.7 cu ft/bbl
(1)
Qi=
(3)
% efficiency = % S, = 1858 X 100 = 30.3% 6120
=
38.7 0.303
Q = -= 127.5 cu ft/bbl (with gas measured at 105 psig)
Problem 2
The following average data are given for a well being produced by continuous gas lift through the tubing: (a) bottom of tubing, 6570 ft, (b) static pressure at the bottom of tubing, 1715 psig, (c) gross productivity index, 2.75 (bbl/day)/psi, (d) pressure at top of casing during gas lift, 975 psig, and (e) pressure at top of tubing during gas lift, 65 psig.
462
Find: (a) pressure at the bottom of tubing during the gas lift, (b) gross liquid production rate, and (c) average tubing pressure gradient during the gas lift. Assumption-The increase in absolute pressure of the gas in the casing is 2.5% (not compounded) per 1000 f t of depth. Solution (a)
p 1 = 990 + 0.025 X 6570 X 990 = 1152 psia or 1137 psig 1000
(b)
q = 2.75(1715 - 1137) = 1590 bbl/day
(c) Average tubing pressure gradient
=
p1 - p 2 ~
L
-
1137 - 65 6570
= 0.1632
psi/ft.
Problem 3
This problem concerns 11 wells, all with a tubing depth of 5800 ft and J The following information is given for the month of June: Well no. 1 2
3 4 5 6 7 8 9 10 11
Oil (bbl/mo)
Gas (Mcf/mo)
Casing pressure (Psi&
Tubing pressure (PW
11,800 13,700 11,900 13,500 11,800 12,500 13,900 12,900 12,800 12,900 13,800
15,000 7400 25,000 8000 15,200 7000 6500 9500 7500 8900 8400
1200 1860 1080 1580 1200 1520 1770 1630 1180 1610
550 460 650 520 550 290 5 80 460 730 440 5 80
1560
= 1.0.
Solve for (1) flowing pressure at the bottom of the tubing, (2) static pressure, and (3) gradient in the tubing (av) in psi/lOOO ft. Plot the average gradient versus gas/oil ratio (G/O, R, or GOR). Assuming that static pressure is declining at 20% per year and that there is no change in the gas/oil ratio, when will each well cease to flow? (The limiting factor is not less than 300 Mcf per day. Use the compound interest concept.) Geothermal gradient = 2 O F/ 100 ft; G / Z = 0.84.
463
Sample solution (Well No. 10)
Pbc
Pbc =
p,,
=Ptc
1625
= 430
GL 53.3XZX
T
0.84 - X 5800 53.3X.578
= 1905 psia
or 1890 psig
+ 1905 = 2335 psia
where J = productivity index; pWs= bottomhole pressure, static; pe = external boundary pressure; p b c = pressure at bottom of casing; p , = bottomhole pressure, general; p w f= bottomhole pressure, flowing; q = production rate. (3) Tubing gradient =
(4)
GOR= 8900 ‘Oo0 12,900
(5)
Min gas = 300 Mcf
1890 - 440 = 250 psi/lOOO ft 5.8 = 690
=
ft3/bbl
[ ~ ~ ~ ( 0 -. P8t u)b ”, c o l , ] J
X
GOR
300,000 = [2335(0.8)”- 1450]1 x 690 solving for n : n = 0.93 years
SAMPLE PROBLEMS A N D QUESTIONS
(1) Define “Working Fluid Level”, “Working Submergence”, and “Percentage Working Submergence”. ( 2 ) Explain why more “paraffin” is formed in gas-lift wells than in pumping wells. (3) List the advantages and disadvantages of gas lift. (4) Explain the different types of gas expansion. (5) List the sources of energy loss in gas-lift operations. ( 6 ) List the methods which are used in removing “paraffin”.
464
(7) Given the following average data for a well being produced by continuous gas lift through tubing: (1) bottom of tubing = 8230 ft; (2) pressure at top of casing = 925 psig; (3) pressure at top of tubing = 130 psig; and (4) specific gravity of well fluid = 0.816. Assume: (a) increase in absolute pressure of gas is 2.4% (not compounded) per 1000 ft; (b) work done is 332 p z log,, ( p 1 / p 2 )ft-lb per cu ft of gas measured at p z (psia); and (c) percentage efficiency = 100. Compute the work of expansion and the volume of gas (measured at 130 psig) required to lift each barrel of well fluid. (8) Given the following average data for a well being produced by continuous gas lift through tubing: (1) bottom of tubing = 7460 ft; (2) static pressure at bottom of tubing = 1825 psig; (3) gross productivity index = 2.35 (B/D)/psi; (4) pressure at top of tubing during gas lift = 105 psig; and (5) tubing-pressure gradient during gas lift = 0.145 psi/ft. Assume the increase in the absolute pressure of the gas is 2.7% (not compounded) per 1000 ft. Find: (a) pressure at bottom of tubing during gas lift; (b) gross-liquid production rate; and (c) pressure at top of casing during gas lift. (9) Given: Well No.
Oil
Gas (Mcf/M)
5
11,800
15,200
Casing pressure (PSid 1200
Tubing depth (ft) 5800
J
1.o
Tubing pressure (PSk) 550
Assume that static pressure is declining at 20% per year (no change in G/O ratio), and use compound interest concept. Limiting factor is not less than 300 Mcf per day. At what time will the well cease to flow? T/dL = 2"F/100 ft; G / Z = 0.84; surface temperature = 60°F. (10) Derive kick-off pressure for gas through casing. For additional problem assignments see the classical work of Brown (1973). REFERENCES Brown, K.E., 1973. Gas Lift Theoiy and Practice. The Petrol. Publ. Co., Tulsa, Okla,, 924 pp. Brown, K.E. and Beggs, H.D.. 1977. The Technology of Artificial Lift Methods, Vol. I. PennWell, Tulsa, Okla., 487 pp.
465
Brown, K.E. and Lee, A.L., 1968. Easy-to-use charts simplify intermittent gas lift design. World Oil, (Feb. 1): 44-50. Craft, B.C., Holden, W.R. and Graves, Jr., E.D., 1962. Well Design; Drilling and Production. Prentice-Hall, Englewood Cliffs, N.J., 571 pp. Chilingar, G.V. and Beeson, 1969. Surface Operations in Petroleum Production. Am. Elsevier, New York, N.Y., 397 pp. Gilbert, W.E., 1954. Flowing and gas-lift well performance. API Drill. Prod. Prac., pp. 126-157. Kirkpatrick, C.V., 1955a. The Power of Gas. Camco, Houston, Tex. Kirkpatrick, C.V., 1955b. Spacing pressure-loaded gas-lift valves. Oil Gas J . , 53(40): 110. Kirkpatrick, C.V., 1962. Gas lift. In: T.C. Frick (Editor), Petroleum Production Handbook, Vol. 1. McGraw-Hill, New York, N.Y., pp. (5-1)-(5-53). Macco, 1966a. The Maccomatic Continuow Flow Gas Lift System. Macco Oil Tool, Houston, Tex., 32 pp. Macco, 1966b. The Maceomatie Intermittent Gas Lift System. Macco Oil Tool, Houston, Tex., 72 pp. Mach, J. and Brown, K.E., 1980. Gas lift. In: K.E. Brown, The TechnologV ofArtificia1 Lift Methods, Vol. 2a. PennWell, Tulsa, Okla., pp. 95-444. Neely, B., Gipson, F., Clegg, J., Capps, B. and Wilson, P., 1981. Selection of artificial lift method. SOC.Pet. Eng. of AIME, 56th Annu. Fall Tech. Conf., San Antonio, Tex., Oct. 5-7, 1981, SPE 10337, 6 pp. Shaw, S.F., 1949. Flow characteristics of gas lift in oil production. Tex. Eng. Exp. St., Bull., 113: 1-9, 26, 47. Thrash, P.J. and Brown, K.E., 1965. Field Operation Handbook for Gas Lift. Otis Engineering, Dallas, Tex., 88 pp. Trammel, P. and Praisnar, Jr., A., 1979. Designing gas lift for continuous liquids removal from gas wells. Pet. Eng., (July): 28-34. Teledine Merla, 1980. Gas Lift Values. Teledine Merla, Garland, Tex., 50 pp. Winkler, H.W., 1957. How to design a closed rotative gas lift system. Pet. Eng., 29(5): 35, 36, 38-40, 42-44, 46. Zaba, J. and Doherty, W.T., 1956. Practical Petroleum Engineer’s Handbook. Gulf Publishing, Houston, Tex., 4th ed., pp, 568-629.
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
467
Chapter 13
PLUNGER LIFT
'
CARROL M. BEESON, DONALD G. KNOX, MOAYED AL-BASSAM, AND GEORGE V. CHILINGARIAN INTRODUCTION
The use of plunger lift in oil and gas wells is increasing. Several reasons for this stem from general developments in the industry; that is, wells are being drilled deeper, pressure maintenance is on the increase, and more data are being obtained on bottomhole conditions and productive capacities. The first of these developments makes pumping more difficult and the second increases the amount of gas available for plunger lift. The bottomhole data have helped in deriving methods for accurately predicting plunger lift performance. Supplementing these general developments, plunger lift equipment has been continually improved. In addition, new designs for the equipment have led to the production of various types of plungers for use in a wide variety of oil and gas wells. Beauregard and Ferguson (1981) and Ferguson and Beauregard (1983) discussed applications, advantages, limitations, and economics of plunger lift. Because of the many variables involved, there is no definite answer to the question whether or not eventually a pumping unit will be required on a well utilizing a plunger lift. As pointed out by Ferguson and Beauregard (1983), the mechanical wiping action of the plunger interrupts the formation of paraffins because of their continuous removal before hardening. In addition, because of the faster rate of fluid removal as compared to the normal flow, there is a lower temperature drop and shorter time available for deposition. The present chapter covers the following areas: (1) Beginning with the history of plunger lift and some developments in the equipment, the first section also includes the well data used to put plunger lift on a quantitative basis. (2) The second section describes the determination of equations for gas and pressure from the well data by the method of least squares. (3) The third section explains the methods used in testing the effects of change in tailpipe, water cut, and oil gravity, and the methods used in constructing nomographs to simplify the calculations.
'
Largely based on Chapter 12 in: Suifuce Operutioi7s ii7 Petroleum Production, by George V. Chilingar and Carrol M. Beeson, Am. Elsevier Publ. Co., Inc.. 1969, pp. 238-306. The help extended by the Petroleum Engineer Publ. Co. and Ferguson Beauregard Inc. is indeed greatly appreciated.
468
(4) The next section explains how to use the nomographs to plot operating lines, and contains supplementary figures representing the equations. ( 5 ) This is followed by the nomographs, along with examples illustrating their use; also discussions of the accuracy to be expected from results determined by the equations or the nomographs. (6) Then are described the use of the equations without plotting an operating line, and types of oil wells suitable for plunger lift. In addition, this section discusses the intermittent flowing and plunger lift system, as well as applications of plunger lift in gas wells. (7) The final section covers prediction of plunger lift performance. At the end of the chapter, there is a list of definitions of the symbols used throughout the chapter. HISTORY
Free-cycling plunger
The earliest plungers were made by the Hughes Tool Company. The inventors must have realized that the addition of a piston to separate the lifting gas from the lifted oil would use less gas than flowing oil in a spray. The operation of the piston with small loads of oil would also exert less pressure on the formation than the gas lifting of large slugs of oil without a piston. These early gas-lift pistons were uncontrolled and were merely allowed to shuttle between the surface and the bottom of the tubing. It was found that a regular cycle could be established by adjusting the choke or bean in the continuously open flow line. The choke or bean setting depended on the gas/liquid ratio, production rate, and depth. The plunger was made to drop and return by using a valve in the body of the device. Upon arrival of the plunger at the surface, the valve opened and this permitted the rising gas and liquid to pass through the hollow body of the piston as it fell. The valve was mechanically closed when the device struck the footpiece spring near the bottom of the tubing. Then the plunger became a solid piston capable of lifting the accumulation of oil that had bubbled up into the tubing since the previous cycle. This type of operation worked wonderfully in some wells and not at all in others. Its use, therefore, was limited and somewhat unpredictable. Special tubing was required, and the footpiece could be removed only by pulling the tubing. In addition, the plunger was subject to short stroking, that is, it sometimes failed to reach the bottom or the top. This made restarting the well a frequent necessity. Cycle-controlled expanding plunger
The National Supply Company took over the plunger in 1944 and, with their experience in the field of gas lift, they began to improve the equipment. Valve locks
469
were added to the body of the device, whch prevented short stroking. A cycle controller operated by pressure was used to open and close, in regular cycles, a motor valve in the flowline. Use of the controller added to the efficiency of the gas-lift piston all the benefits of intermittent flow, including substantial savings in gas. The device also permitted varying the size of the load by allowing the plunger to be held at the bottom for practically any desired length of time. The earliest cycle controllers depended on well pressure alone for both opening and closing the flowline. Then an important contribution was made by development of a trigger and a vent valve for closing the motor valve upon arrival of the plunger at the surface. This resulted in additional savings of gas. The trigger development also permitted the closing function of the cycle controller to be used as a safety shutoff, in the event that the plunger failed to strike the trigger. The expanding plunger was designed to eliminate the need for special tubing. A retrievable footpiece and broaches were designed for use with a wireline, making it possible to install the equipment without pulling the tubing or killing the well. Lea (1981) presented a description of a dynamic model of plunger lift operations. which includes calculation of the plunger velocity as the plunger and liquid slug travel up the tubing. He also presented an analysis of plunger cycles in high gas/oil ratio wells, to indicate the maximum (1) casing pressure necessary to lift the plunger and accumulated liquids, and (2) rate of slug buildup.
EQUIPMENT DEVELOPMENTS
Plunger equipment for gas wells, gas-lift wells, and high gas/oil ratio wells is illustrated in Fig. 13-l.a, b, c. Improvements in the equipment, as well as new applications for variations in the device, have been made continually. Some developments include Type M Christmas tree, turbulent-seal plungers, tandem plunger, improved removable footpiece, Taylor Type K cycle controller, and time cycle controller with attachments. In intermittent gas-lift wells, there is a dramatic increase in efficiency upon placement of plunger above the operating valve between the lift gas and the liquid to be lifted (Beauregard and Ferguson, 1981). Type Mplunger lift Christmas tree
The main improvement in the plunger lift Christmas tree is the magnetically operated trigger. This trigger always stays in adjustment, owing to the fact that the plunger does not make physical contact with the trigger. It has a foolproof action, because the trigger functions regardless of the position or speed of the plunger on its arrival at the surface. The problem of packing-off between the tubing and atmosphere has been eliminated, because there is no mechanical linkage extending into the interior of the tubing.
470 EQUIPMENT FOR QAS L I F T W E L L S
O LUBRICATOR WITH FLOW C
CATCHER N I P P L E ______
CONTROLLER BALL VALVE ~~
_MOTOR ___
C
VALVE
IVENTIONAL GASLIFT VALVES
DOWNHOLE SHOCK ASSEMBLY WITH C O L L E T T HOLD DOWN
__
S I D E POCKET MANDREL WITH VALVE
~
Fig. 13-1.Schematic diagrams of free piston systems: (a) Plunger equipment for gas-lift wells. (Courtesy of Ferguson Beauregard Inc.) (b) Plunger equipment for gas wells. (c) Plunger equipment for high gas/oil ratio wells. (d) Retractable segmented plunger called Vertipig. (Courtesy of Ferguson Beauregard Inc.) (e) Operation of plunger lift system.
471 E P U I P Y E N T FOR OAS WELLS
LUBRICATOR WITH FLOW COUPLINO
CONTROLLER
CATCHER N I P P L E
I-
MOTOR VALVE
LOWER FLOW
OUTLET
BPLL VALVE
-PLUNGER
/ DOWNHOLE SHOCK ASSEMBLY WITH
STANDING VALVE C A G E
,
STANDARD PUMP SEAT NIPPLE
472 EQUIPMENT FOR HIGH ___
-
:TI -
CONTROLLER
MOTOR
RATIO -- O I L WELLS
-
LUBRICATOR WITH FLOW COUPLING
VALVE
a
*
/ CATCHER N I P P L E
WITH
MAGNETIC SHUT-OFF
n
DOWNHOLE
,
e
G
P R E S S U R E SENSOR
SHOCK
STANDING VALVE CAQE
/
STANDARD PUYP S E A T NIPPLE
I
Fig. 13-1 continued.
413
MULTI-FLEX
VERTI-PIG
H O W THE FREE PISTON SYSTEM WORKS 1. Well closed, plunger falls bygravity. Casing pressure low, and starts increasing.
2. Upon striking footpiece, plunger valve closes. Casing pressure still increasing.
3. Casing pressure peaks, plunger startslifting oil column, tubing pressure drops.
4. Well opened, casing pressure drops, tubing pressure drops, then rises as fluid reaches surface.
5. Wellclosed. plunger strikes bumper, valve opens and it is ready to fall again. Casing pressure is low, completing cycle.
Fig. 13-1 continued.
Turbulent-seal plungers
Turbulent-seal plungers were designed for certain types of oil and gas wells where the expanding plunger is either unnecessary or not suitable. Two plungers of this type have been developed, One features a nylon brush and the other involves the use of teflon. The brush-seal plunger forms an effective seal because the bristles provide a flexible contact with the tubing, and a turbulent seal is created between the spirals of the brush. The plunger is intended primarily for gas wells which produce small volumes of water or condensate, and wells that produce small cuts of sand. It is built, therefore, without a valve but with a standard Otis-type fishing neck. Wear is confined to the replaceable brush, and there is little chance of sticking the plunger. This is true because a sufficient clearance is provided between the tubing and the steel portions of the plunger, and there are no places for sand or other solids to collect in the device. The teflon-seal plunger creates a turbulent seal by means of horizontal grooves in teflon surrounding the narrow center section of the steel body. The grooved teflon is
474
made in cylindrical sections which are easily replaced. This plunger is intended primarily for the rapid cycling of wells, so it is built with a valve. It has proved particularly useful in wells producing fine, abrasive sands which do not affect the valve, but which wear away the expanding metal bars and segments of the expanding plunger. Tandem plunger
The extended or tandem plunger was designed so that gas would lift it through a tubing string containing retrievable valves. The shorter, standard plunger would not rise through the enlargement of such a valve, because of the blow-by of gas. The tandem unit consists of two brush-sealing elements, each of which is screwed and locked onto the end of a pony sucker rod. The over-all length is about 10 ft, and therefore, one of the elements is always making a seal within the regular-sized portion of the tubing string. Because of the length of the tandem unit, a specially designed hinged lubricator is mounted above the master gate for installing and recovering the plunger. Segmented retractable pads A retractable segmented plunger (Fig. 13-1.d) is manufactured by Ferguson Beauregard, Inc., called Vertipig. The stainless steel pads are cast to conform to the “nominal” I.D. of the tubing when expanded on the upstroke of the plunger. A cam mechanism in the tool is activated in the lubricator to retract the seals O.D. by in and create a bypass for the tool on its downward stroke. This type of bypass permits fast fall time and is very effective where fast cycle time is critical. In addition, it is superior for paraffin removal, inasmuch as it falls through the paraffin in the retracted position and only wipes the tubing on the way up. Ferguson Beauregard, Inc., developed an extremely light-weight expanding steel pad plunger that has been highly effective in very low-pressure, low-differential applications.
a
Removable footpiece
One type of footpiece has a built-in pulling rod. The upper end of the rod is an Otis-type fishing neck, and the rod extends through the compression spring into the hold-down adapter. Whatever type of hold-down is used, the footpiece can be easily fished from its locked position with a wireline hoist, because the rod extending into the adapter allows a positive pull to be applied to the footpiece. Taylor Type K cycle controller
This recording cycle controller was designed so that both the opening control point and the safety shutoff point could be set directly on the desired pressures.
47 5
Synchronization of the pen with the opening and closing pointers is easily understood and accomplished. The heart of the instrument is made in one compact, replaceable unit. Time cycle controller with attachments
Attachments are available which may be added to any time cycle controller, so that the motor valve will be closed by the plunger upon arrival at the surface. The arrangement is such that the instrument shutoff is retained as a safety feature which will close the flowline in event the plunger fails to reach the surface. The time intervals for opening the flowline and for the safety shutoff may be varied at will. Time controllers are used where pressure control is not feasible. This is the case in nearly all gas wells where plungers are used to remove accumulations of water or condensate. Time controllers are used also to ensure that each well in a group starts flowing at a different time. This prevents the overloading of compressors and gas-gathering systems. Electronic controller
The electronic controller is basically a time cycle control with digital display of the cycle time. The unit is powered by “D” cell batteries and uses solid state circuits for accuracy and longer battery life. Cycle setting is made by entering the on period (mode) and the off period (mode) into the control. Selection for the length of time of each mode is from 1 min to 99 hrs and 99 min. The control then repeats the setting of each mode to form the cycle time. This is a distinct advantage over mechanical timer, which uses the 24-hr timing wheel where cycle must be divisible into the number of lugs on the wheel. Digitrol I1 by Ferguson Beauregard offers external sensors such as Hi-Lo pressures for casing or tubing, plunger arrival shutoff, high-liquid level shut down, or differential monitors which are connected to the terminal board in the controller panel. These units may be connected in such a way that they will suspend all timing in the “off” mode, or they may just override the time and switch to the next mode. Any number of sensors may be used on the control at one time. Inasmuch as these are electric switch sensors, there are no high-pressure lines connected to the control panel, which is a definite safety feature.
EARLY PREDICTION METHODS
For several years while the early stages in equipment development there was no reliable method of predicting plunger lift performance. some charts which related operating pressure and production rate depths. There was also a means of predicting effectiveness, that is, the expected rate of production to the rate computed for a zero flowing pressure.
took place, There were for various ratio of the bottomhole
416
These methods were based on data obtained with the free cycling Hughes plunger. They soon proved inadequate for the better scaling and controlled plunger. Very little study had been made of gas requirements. The available charts on requirements of gas volume were unreliable, and so rules of thumb often were used in their place. WELL DATA FOR CYCLE-CONTROLLED EXPANDING PLUNGER
By 1954, it had become apparent that the older charts should be replaced by prediction methods based on data obtained with the cycle-controlled, expanding plunger. Accordingly, an intensive effort was made to obtain field data on this type of equipment that would be representative of operations throughout the oil-producing areas of the United States and covering wide ranges of operating conditions. The data have been listed in Tables 13-1 and 13-11. All of the data received from the operators have been included in the tables, except where inconsistencies were apparent. Of the 70 wells (or well conditions) with 2-in. plunger, 43 were from the midcontinent, 14 from California, and 13 from the East. Of the 75 wells (or well conditions) with 24-in. plungers, 61 were from California, 12 from the Rocky Mountains, and 2 from the midcontinent. The distribution of the other variables involved in the well data also were quite suitable, as shown in Table 13-111. LEAST SQUARES EQUATIONS FOR PLUNGER LIFT
Description of plunger lift and need for equations One of the efficient methods of producing oil involves the use of the cycle-controlled, expanding plunger. The expanding segments make continuous contact with the walls of regular API tubing strings, thereby forming an effective seal between the lifting gas and the lifted liquid. On reaching the surface, a valve in the plunger is opened mechanically, permitting the plunger to fall through the gas and oil to the footpiece spring. On reaching the footpiece, the valve is closed mechanically, so that gas under pressure may lift the plunger to the surface with another load of oil. The plunger starts upward when the motor valve in the flow line is opened by the cycle controller, usually after the pressure in the casing has reached some predetermined value. The motor valve in the flowline ordinarily is closed by the arrival of the plunger at the surface. (See Fig. 13-1.d.) In plunger lift operation, the flowline is opened and closed to gain the advantage of controlled intermittent flow. Combined with this advantage is the piston-like efficiency of the expanding plunger. Some wells do not produce sufficient formation gas to operate the plunger, so use is made of circulated gas. In that case, a convenient source of supply is connected to the casing, and gas is injected continuously at a rate sufficient to supplement the produced formation gas.
TABLE 13-1
Well data for 2-in. plungers Well
Period averaged
Gross liquid
Net oil
Circulated gas
Gross gas
Casing pressure (psig)
Tubing pressure (psig)
Trap press.
Cycles Oil per gravity day
Casing gas gravity referred ("API) to air
Choke or bean
Casing size to
Plunger footpiece a
Tubing depth
setting
foot-
(in.)
piece (in., 0.d.)
(ft)
(ft)
5: 7 5:
7673 7002 7855
7720 7017 7870
(days)
(bbl/ day)
(bhl/ day)
(Mcf/ day)
(Mcf/ day)
max
min
max
min
(psig)
1 2 3
-
20
0 0 0
108 50 50
620 380 225
550 290 200
550
-
20 4 10
-
490 20 36
25 20 36
35 10 80
44 42 40
0.72
32/64 48/64 48/64
4
-
-
12
0
50
280
220
-
30
30
9
40
-
32/64
5;
7713
7728
5
-
-
14
0
150
360
310
-
20
20
24
40
~
32/64
5:
7852
7867
6
-
-
15
0
150
370
310
-
20
20
18
40
-
40/64
5:
7830
7860
7
-
13.8
13.8
290
250
240
120
20
28
41
-
24/64
5:
7715
7770
-
12.4
12.4
0 0
138.2
8
144.4
340
300
340
150
20
23
41
-
24/64
5;
7715
7770
9
1
17
9
65
81
470
415
280
160
60
18
50.0
0.775
30/64
5f
8946
9084
10
1
14
4
67
74
440
410
350
150
60
12
50.0
0.775
30/64
5:
8946
9084
I1
1
15
2
77
81
575
520
400
90
60
14
50.0
0.775
30/64
5:
8946
9084
12
1
11
3
61
66
575
520
400
90
60
14
50.0
0.775
30/64
5;
8946
9084
158
163
365
325
340
120
10
16
50.0
0.775
30/64
5:
9114
9157
128
365
325
340
120
65
16
50.0
0.775
30/64
5:
9114
9157
-
-
-
13
1
9
3
14
1
7
2
124
15
1
8
4
84
91
455
410
350
85
70
16
50.0
0.775
30/64
5;
9114
9157
16
1
9
70
79
425
380
300
100
70
16
50.0
0.775
30/64
5;
9114
9157
17
1
6
5 3
67
72
400
350
320
90
70
16
50.0
0.775
30/64
5f
9114
9157
18
1
9
4
67
74
420
380
350
90
70
18
50.0
0.775
30/64
5;
9114
9157
19
30
21
13
40.0
53.4
200
180
140
30
30
40
36.9
0.78
Open
4600
4612
20
31
11
33.0
43.7
180
160
110
30
30
37
36.9
0.78
Open
5: 5;
4621
4625
21
31
18
37.0
48.0
210
180
140
30
28
43
36.9
0.78
Open
5:
3808
22
30
0
33.7
400
360
340
200
30
21
33.3
0.78
3/8
5;
7711
3883 7718
9.5
6.7 11
8.2
4 P 4
TABLE 13-1 (continued) Well
Period averaged
Gross liquid
Net oil
Circulated gas
(bbl/ day)
(bbl/ day)
(Mcf/ day)
Gross gas
Casing pressure (P&)
Tubing pressure (P&)
max
min
max
min
Trap press.
Cycles Per day
Oil gravity
Choke or bean setting
Casing size to footpiece
(OAPI)
Casing gas gravity referred to air
(in.)
(in., 0.d.)
0.78
Plunger footpiece =
Tubing depth
5:
6630
6664
9: 7 7 7 7
6400 11501 11501 11665 6094
6455 11506 11506 11670 6511
23
28
35
35
0
187.4
500
440
440
125
125
35
37.0
24 25 26 27 28
-
6 4 1 61
17.6 63.2 61.1 16.0 27
17.6 51.0 47.3 4.8 26
0 0 0 282.1 88
48.2 500.2 470.7 435.4 276
425 372 390 310 195
410 362 378 280 185
390 352 320 250 145
15 90 100 70 81
10 85 85 85 81
18 67.5 71.5 47 84
40 37.7 37.5 26.5 35.6
0.732 0.728 0.633 0.789
1/2 40/64 19/32 19/32 64/64 32/64
29
30
25
21
0
193
190
180
100
60
60
96
36
0.70
3/8
11:,7
5959
5965
30
30
24
20
0
221
165
160
130
60
60
72
36
0.70
3/8
11$,7
5959
5965
31
31
22
18
0
202
175
170
140
60
60
72
36
0.70
3/8
11:,7
5959
5965
32
30
21
18
0
152
175
170
140
60
60
72
36
0.70
3/8
11:,7
5959
5965
19 6.3 5.1 7.2 16.7 13.8 7.6 5.3 4.8 6.9 6.9 15.8 4.7 5.5 6.2 16.1
16 4.6 1.7 5.9 16.7 11.3 7.1 5.0 4.8 6.5 6.9 15.6 ?.5 5.5 6.2 16.1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
116 90.0 62.6 52.9 110 34.8 91.4 57.5 66.0 52.4 38.8 88.5
175 203 180 180 120 180 250 270 125 265 283 268 115 248 205 510
170 190 168 168 110 168 225 250 115 250 260 258 105 225 175 480
140 170 170 170 100 170 230 250 115 250 260 250 105 230 180 480
60 21 7 18 18 18 23 20 16 20 20 24 16 20 24 24
60
72 45 46 28 86 37 24 26
36 38 37 37 36 37 35 35 36 35 35 36 36 35 35 37
0.70
3/8 1/2 1/2 29/64
11i.7
5959 7243 6972 6946 6984 6947 7102 6850 7092 7101 6969 6951 7087 7213 7242 7065
5965
33 31 1 34 35 1 36 1 37 1 38 1 39 1 40 1 41 1 42 1 43 1 44 1 45 1 4 6 1 1 47 48 1
44.0 52.6 184 70.0
44 30 17 33 35 25 27 12
-
34/64 1/2 -
1 /2
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
6990 -
7026 6957 7132
6966 7102 7229 7254 -
49 50 51
1 1 31
49.8 38.6 -
52
31
-
53
1
14
54
1
10.4
55 56
1 1
40 4.2
49.8 34.8 11.07
0 0 0
13.83
0
14
0
187 82.5 165.2 87.5 109
165
310 320 200
22 0 160
0 30
330 480
310
195
-
-
36 1 /2
0.8
1/2 23/64
5;
7514
7576
5;
4755
4760
30/64
5;
5020
5020
32/64 20/64
5: 5;
4785 4140
4820 4144
18/64
5:
3840
3844
30/64
5:
4150
4154
495
430
400
130
30
10
42
680
630
520
70
20
28
45
~
~
28.4
650
605
540
50
20
9
46
38 4
0 0
42 8
720 300
680 260
650 250
30 200
20 15
14 4
-
41
~
~
-
57
1
11
10.5
0
23
350
310
260
150
20
6
42
1
18
17.5
0
30.1
600
550
500
375
20
7
41
~
5:
4422
4430
4185
4185
30/64
5:
5172
5187
28/64
5;
5180
5210
17/64
5:
5190
5194
8/16
4;
3240
3262
-
8/16
5:
3180
3220
-
1/2 5/16
5:
3100
3150
5;
3000
3050
1/2
5;
3439
3439
1/2
5;
3534
3534
3/43
5:
4140
4140
800
700
580
550
103
44
~
360
340
250
40
29
41
~
14
0
32.3
740
700
520
40 70
40
16
42
~
26.8
26.8
0
72.9
690
630
500
100
50
12
42
~
1
16
14
0
35.2
670
635
500
70
20
7
42
~
64
30
45
45
0
50
180
160
125
50
40
96
46
65
30
9
8
0
14
160
160
115
46
1
15.2
15.2
0
25
195
-
-
12 70
48
66
140 175
67
1
12
12
0
75
85
75
-
25
72
68
61
43
43
0
32
400
350
350
155
75
32
38-40
~
69
31
30
30
0
35
400
370
290
110
75
20
38-40
~
70
31
10
10
0
170
315
250
290
175
40
36
38-40
4
61
1
14
62
1
63
a
42 0
4
0
Tubing size to footpiece is 2; in. 0.d. for all wells.
~
’ Values not available.
42
7531
5:
820
80
42
1
-
32/64
500
1
~
34/64
0
0
59
6897 6995 7563
0.8
58 60
6863
42
-
0
8.4
7 7 5;
56 36 26
~
~
-
TABLE 13-11 P
0 W
Well data for 2;-in. plungers Well
Period Gross Net averliquid oil aged
Circu- Gross lated gas gas
Casing pressure (Psig)
(days)
(bbl/ day)
(bbl/ day)
(Mcf/ day)
max min max min (psig)
1 2
31 9
26 4
25.5 17
142 178
153 178
183 245
168 150 200 150
50 50
35 35
3
11
70
49
150
190
205
175 150
50
4
7
31
28
248
258
185
155 150
5 6 7
20 13 18
88 108 42
74 101 23
125 133 234
140 143 234
190 200 235
166 150 172 150 215 150
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
21 21 21 153 61
61.4 11.4 23.5 110 55
61.0 11.3 23.5 110 55 11 22 20 11.6 18.8 20 17 24 29 35 29 39 32 9
0 0 0 0 0 0 0 0 96.0 96.0 81 75 0 147 0 130 363 398 105
438 642 188 418 154 50 150 150 286.3 262.2 207 322 176 288 456.5 165 484 464 110
119 117 390 320 360 210 450 260 300 260 190 210
110 105 358 300 350 160 400 220 255 230 170 190 120 180 128 700 760 435 410
4
-
-
-
-
-
-
1 1 31 31 31 31 123 31 30 30
13.8 23.5 36 18 25 35 37 29 ' 40 60
30
10
140 200 138 760 840 475 415
Oil Casing gravity gas gravity ("API) referred to air
Choke or bean setting (in.)
104 125
37.5 37.5
0.738 0.738
3/4 3/8-1/2
35
110
37
0.720
3/8-1/2
5: 5;
50
35
120
37.5
0.738
3/8-1/2
50 50 50
35 35 35
115 120 120
31.5 37 36.5
0.720 0.720 0.745
3/8-1/2 3/8-1/2 3/8-1/2
82 70 59 90 110 20 150 50 80 80 82 81 82 85 60 300 550 100 160
68 70 48 60 70 20 10 10 85 85 82 81 82 85 60 200 480 75 75
151 163 21 120 40 30 14 11 22 31 63 60 43 75 72 13 32 45 17
41-42 34-35 40 35 35 40 40 38 33.7 28.7 35.0 35.5 31.5 29.5 30 21 40.0 36.2 34.8
0.690 0.655 0.675 0.8 0.8
5/8 Open 3/4
-
48/64 31/64 1 32/64 32/64 32/34 32/64 32/64 32/64 1/2 1/4 48/64 19/64 15/64
Tubing pressure (Psi@
84 82 345 260 310 -
400 220 175 225 -
105 600 750 400
420
Trap press.
~
-
0.629 0.629 0.789 0.789 0.780 0.789 0.70 0.62 0.685 0.685 0.685
~
Casing Plunger footpiece size to footpiece depth (in., 0.d.) (ft)
a
Tubing depth (ft)
3756 3633
3768 3638
3606
3614
5:
3629
3637
5: 7 5;
3634 3793 3155
3640 3799 3760
6; 7
3395 2252 5053 6325 6250 7801 7335 8560 11385 11385 6519 6629 6583 7605 5988 6974 8800 7989 7457
3407 2308 5101 6355 6280 7811
7
5: 7 7 7 7 7 7 7 7 7 7 7 7 7 7 I 7
-
8600 11390 11390 6547 6659 6615 7635 6008 9647 9658 9246 7457
27 28 29 30 31 32 33 34 35 36 37 38
4 3 3 6 3 1 1 1 1 30 30
58 28 98 21 59.6 36.7 16 13 6 27 22.7 50.5
33 22.2 98 21 59.6 36.7 11 12 5 25 22.3 33.1
207 68.5 425 154 159.5 182 0 0 0 0 0 271
39
31
51.5
40.8
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
30 31 31 31 31 31 31 31 31 31 31 31 2 6 3 3 3 8 4 3 5 6 1
39.8 35.0 33.0 62.0 11.0 11.0 20.0 70.0 90.0 54.5 31.3 41.5 84 84 84 69 62 64 71 69 70 71 51
32.2 32.0 31.0 60.0 6.0 6.0 16.0 24.0 10.0 48.4 35.6 40.5 83 82 83 68 60 62 69.1 68 69 69.7 54
-
153 54 197 200 331
150 339 532 296 370 357 330 240 190 250 100 520
135 299 484 260 334 320 295 205 155 215 85 480
159
271
470
440
139.6 130 270 0 0 0 0 150 0 0 0 0 655 60 0 0 100 93 101 92 105 67 25
207 170 300 220 90 110
475 340 450 650 300 290 300 560
430 300 310 570 210 250 260 500 660 260 623 684 655 680 690 680 605 660 640 680 580 530 670
252 103 650 210 195 248 70
yo 375 460 176.7 55.0
83.8 780 254 184 245 160 204 200 181 235 141 303
740
280 653 124 715 740 750 750 670 720 710 740 640 580 770
100 261 425 265 271 300 220 200 170 220 80
50 22.8 46 25.7 27 31 13 35 11 41 84 26
27 24.8 31.9 27.5 28.3 30.0 34.3 33.1 35.0 33.0 36.5 31.6
50
33
50 70 70 70 70
25 26 26 24 14 17 22 30 31 57 12 27 22 19.4 17 19 17 15 16.4 16 21 17 13
20 63 56.4
400
30 79 100 14 70 88 30 30 30 30 45 85
380
85
400 250 380 550 280 270 270 475 650 223 560 630 645 700 630 640 550 625 620 610 530 440
85 15 100 120 75 15 80 120 300 104 400 350 160 120 140 100 60 90 100 90 90 65 65
-
-
62 54 25 25 25 25 42 50
10 70 70 70 85 85 85 -
-
-
53
0.701 0.706 0.690 0.706 0.71 5 0.700
1 48/64 48/64 40/64 48/64 48/64 40/64 40/64 42/64 40/64 36/64 34/64
32.0
0.695
35.0 24.0 26.0 30.0 33.0 33.0 28.0 28.0 42.0 29 29 30 36.4 36.4 35.6 37.3 34.6 34.4 34.9 34.0 34.0 33.9 39.0
0.695 0.692 0.692 0.681 0.751 0.132 0.724 0.692 0.735 0.71 0.75 0.74
0.670 0.688 0.688 0.688 -
-
8; 7 7 7 7 7 7 7 7 7 1;
4774 6425 9174 9964 9690 9092 7610 8609 8417 8100 5750 10963
4714 6486 9206 10028 9121 9156 7670 8669 8531 8160 5750 11678
16/64
7;
11300
11727
17/64 24/64 24/64 24/64 24/64 24/64 24/64 24/64 24/64 24/64 24/64 12/64
7;
11198 9706 9707 8209 10201 9511 10472 10189 10023 4117 4859 4755 9344 9344 9344 9255 9365 9349 9349 9349 9349 9349 9349
11638 9706 9107 8209 10201 10276 10472 10189 10444 4117 4871 4758 9753 9153 9753 9771 9874 9842 9842 9842 9842 9842 9842
-
-
~
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.678
-
7
1 4: 1 7 7 7 7 7 7 7 7 1 7 7 7 7 7 7 7 7 7 7
m P +
TABLE 13-11 (continued)
P
N W
Well Period Gross Net averliquid oil aged (days) (bbl/ (bbl/
63 64 65 66 67 68 69 70 71 72 73 74 75 a
1 1
8 4
4 4 4 4 5 3 3 10 1
Circulated gas (Mcf/
Gross gas
(Mcf/
day)
day)
day)
day)
47 47 28 76 59 67 64 25 32 39 62 18 12
46 46 27 72 57 62 58.5 21 30 37 60 18 11
29 3 53 53 112 102 84 69 54 0 0 36 43
209 183 102 163 175 179 161 123 112 90 126 92 74
Casing pressure (Psig) max min
800 800 620 750 720 690 650 735 760 750 720 570 680
Tubing size to footpiece is 2: in. 0.d. for all wells.
Tubing pressure (Psi@ max min
720 720 545 550 520 680 640 620 640 625 570 600 535 655 640 660 700 650 660 650 670 490 560 590 ~
~
50 50 70 100 100 90 160 90 110 70 140 200 80
Trap Cycles Oil press. per gravity day (psig) (oAPI)
Casing gas gravity referred to air
50 70
0.678 0.678
~
~
~
~
-
61
10 8.7 8 4 15 16 19 6 10 10 13 8 4
36.4 36.4 34.5 34.9 33.0 34.4 30.1 32.4 34.4 34.8 36.5 33.4 34.1
-
0.678
Choke Casing Plunger Tubing or bean size to footpiece a depth setting footpiece depth (in.)
(in., 0.d.)
(ft)
(ft)
7 I 7 7 7 7 7 7 7 7 7 7 7
9836 9836 9063 9782 9927 9960 9960 10876 9855 9855 9900 10016 10016
9867 9867 9063 9804 9938 9987 9987 10906 9855 9855 9921 10048 10048
483 TABLE 13-111 Ranges
a
and averages of well data 2-in. Plunger
Tubing depth, ft Tailpipe, ft Production rate, bbl/day Size load, bbl/cy Oil gravity, OAPI Water cut, %
2:-in. Plunger
Low
High
Average
Low
High
Average
3050 0
11,670 138
6677 32
3407 0
11,727 1,257
8313 185
4.2 0.1
35 0
63.2 2.9 50 86.7
18 0.7 38.0 17.0
10 0.2 24 0
110 5.4 46 88.9
45 2.0 33.7 12.6
The variables ranged fairly uniformly between the low and high values listed, except as indicated in footnotes b-g. One well had a tubing depth of 2308 f t . One well had a tailpipe of 417 ft. One well had a tailpipe of 2673 ft. One well had a size load of 0.07 bbl/cy. One well had an oil gravity o f 26.5’API. g One well had an oil gravity of 21OAPI.
a
Ordinarily, casing pressure is used for controlling the cycles and no tubing packer is set. It is this standard type of installation which is the primary subject of the present chapter. Cycles may be controlled, however, by tubing pressure or by time, wherever casing-pressure control is impracticable. Time control is especially appropriate for removing accumulations of water or condensate from gas wells. This method also is used to prevent overloading gas-gathering systems, by ensuring that each well in a group starts flowing at a different time. The early plungers, not being controlled by an instrument, were allowed to “free cycle” while the flowline was open continuously. This method may be employed in some cases, but the equations obtained from the present investigation do not apply to this type of operation. The cycle controlled plunger is so flexible that it may be used under a wide variety of operating conditions in a given well. It was desirable, therefore, to develop a quantitative method of predicting the requirements and performance for various sets of operating conditions. This would make it possible to choose the most suitable conditions for plunger lift in the well. Then the predicted requirements and performance could be compared with those for other methods of production. In order to put plunger lift on such a quantitative basis, the field data tabulated earlier were collected from the producing areas of the United States. The data were then analyzed to determine the variables controlling the gas and pressure requirements. Equations
The main items of interest for the operation of the gas-lift plunger in a given well are the requirements of net operating pressure and gas-liquid ratio. It is important also to be able to estimate the casing pressure buildup and the maximum production rate.
484
A knowledge of the pressure buildup is useful because the net operating pressure depends upon the maximum casing pressure, whereas the production rate depends upon the average casing pressure, and these pressures differ by one-half the pressure buildup. The maximum production rate is needed as one limit to the large number of operating conditions that may be used with plunger lift. It will be shown how the following six basic equations concerning gas and pressure were derived from correlations of field data by the method of least squares. The method of obtaining the two equations for maximum production rate also will be explained. (For definitions of symbols, see Nomenclature at the end of the chapter.)
Net operating pressure:
2"
P,"" - P F = 77.1Lc + 8.29-
D 1000
+
2i"
P,"" - Ptmin = 3.44LC 5.78-
+ 18.O8Lc-1000 + 67.4
D 1000
+ l l . l l L c -1000 +91.4
(13-1) (13-2)
Gas-liquid ratio gradient:
2"
~
,iff
G/L - 9.90( D/lOOO) + 0.774PP" D/lOOO LC
+ 279 + 201
+
~
47.8( D/lOOO) 0.738P,m'" + 2.54 G/L D/1 000 LC
(13-3)
+ 122
(13-4)
+ 0 . 0 2 o p p + 9.9
(13-5)
Pressure buildup:
2" in
Pcm"- PCm'" = 8.16Lc + 3.43-
5t"
Pcm" - Pcm'"= 3.19Lc
24" in 7"
1000
+ 9.31- 1000 + 0.051P,m'" - 43.1
(13-6)
Maximum production rate:
2" 2t"
q""
=
1440L, 1.5( 0/1000)
1440L, 1.5( 0/1000) + 6Lc
'"" =
+ 8L,
(13-7)
(13-8)
485 METHOD OF OBTAINING EQUATIONS FOR GAS A N D PRESSURE
For each size plunger, the three basic equations were obtained from correlations of field data by the method of least squares. Theoretical calculations and plots were made only to ascertain which variables were indicated as influencing each factor pertinent to the operation of plunger lift. For net operating pressure, the indicated variables were size load, depth, and their product L, x D.For the factors, cycle gas gradient and pressure buildup, the indicated variables were size load, depth, and minimum tubing pressure. A regression equation was formulated with a factor expressed in terms of the indicated variables. The method of least squares (Ezekial, 1941) was applied to determine the constant in the equation and the coefficient of each independent variable. These coefficients were then used (Ezekial, 1941) to compute the coefficient or index of multiple correlation. Throughout the calculations, the depths were rounded off so that a table of squares up to 2000 could be used without interpolation. Values were computed for the differences between the actual data and the amounts computed from the equations. In the case of net operating pressure for the 2-in. plunger and the cycle gas gradient for the 2+-in. plunger, the difference for one well was at least five times the average difference. Accordingly, the data for that well were disregarded (Sherwood and Reed, 1939) and the coefficients were determined again, using the data for all the wells except that one. As applied to curve fitting, the principle of least squares states that the most probable empirical equation is the one for which the sum of the squares of the residuals is a minimum. A residual is the difference between the amount required by the equation and the observed value. An equation obtained by the method of least squares represents the most probable values of a dependent variable for different values of the chosen independent variables. For a chosen set of independent variables, the values of the dependent variable need not be exactly equal to the average values. Accordingly, the sum of the positive residuals need not exactly equal the sum of the negative residuals. Nevertheless, a comparison of the plus and minus sums of the residuals may be used to show that an equation substantially represents average values. No figures have been included to show that the equations represent averages of the data, because four dimensions would be required for most cases. There have been listed, however, the plus and minus sums of the residuals. Inasmuch as the sum of the positive residuals and the sum of the negative residuals are nearly equal in every case, it follows that the equations substantially represent averages of the data.
Wells with 2-in. plungers The equation for net operating pressure was determined in the form given in eq. 13-1. Well 53 was disregarded (Sherwood and Reed, 1939) in determining this equation, because the difference between the reported net operating pressure and
486
that required by the equation obtained by using all the wells was five times the average difference. Using the 69 remaining wells, the index of multiple correlation was 0.710 (instead of 0.661 for all 70 wells). The positive and negative sums of the residuals were +2650 and -2648. The equation for gas volume was obtained in terms of the cycle gas gradient, and was expressed as follows: ~Gc
D/lOOO
- 9.90( D/lOOO)
+ O.774Ptmin+ 279 + 201Lc
(13-9)
Using the 70 wells, the coefficient of multiple correlation was 0.520. The plus and minus sums of the residuals were +6887 and -6901. Dividing both sides of eq. 13-9 by L, yields a relation for gas-liquid ratio gradient based on volumes per cycle:
G,/L, - 9.90( D/lOOO) -D/lOOO
+ O.774Pt"'" + 279 + 201
(13-10)
L C
Equation 13-10, expressing GJL, is equivalent to eq. 13-3 given previously in terms of G / L , which expresses the ratio of volumes for any interval of time. The equation for pressure buildup was obtained as given in eq. 13-5. The 41 wells with Sf-in. casing were used, and the coefficient of multiple correlation was 0.576. The positive and negative sums of the residuals were + 194 and - 197. Wells 25-28 and 34-49 had 7-in. casing instead of the 5 i - h casing. Accordingly, these 20 wells offered an opportunity to test the belief that pressure buildup is inversely proportional to the annular areas. The ratio of the actual pressure buildup to that required by eq. 13-5 averaged 0.424. The inverse ratio of the annular areas involved is 0.515. Consequently, these results support the belief that pressure buildup is inversely proportional to the annular areas involved, for a given size tubing. Wells with 24-in. plungers
The equation for net operating pressure was determined in the form given in eq. 13-2. Using the 75 wells, the index of multiple correlation was 0.945. The plus and minus sums of the residuals were 1849 and - 1981. The equation for gas requirement was obtained in terms of the cycle gas gradient, and is expressed as follows:
+
GC D/lOOO
= 47.8-
1000
+ 0.738P,"'" + 254 + 122Lc
(13-11)
Well 52 was disregarded (Sherwood and Reed, 1939) in determining eq. 13-11 because the difference between the actual cycle gas gradient and that required by
487
the equation obtained by using all the wells was almost eight times the average difference. Using the 74 remaining wells, the coefficient of multiple correlation was 0.639 (instead of 0.599 for all 75 wells). The positive and negative sums of the residuals were +9484 and -9698. Dividing both sides of eq. 13-11 by L, yields an equation for gas/liquid ratio gradient based on volumes per cycle. Analogous to the case of the 2-in. plunger, the latter relation for the 24411. plunger is equivalent to eq. 13-4 given previously for gas/liquid ratio gradient based on volumes for any interval of time. The equation for pressure buildup was obtained as given in eq. 13-6. The 61 wells with 7-in. casing were used, and the coefficient of multiple correlation was 0.837. The plus and minus sums of the residuals were +415 and -417.
METHOD OF OBTAINING EQUATIONS FOR MAXIMUM PRODUCTION RATE
The minimum cycle time involves movement of the plunger from the footpiece to the surface, unloading the liquid at the surface, and movement of the plunger from the surface to the footpiece; the last step might be divided into the fall of the plunger through gas to the top of the liquid and the fall of the plunger through the liquid to the footpiece. From this description, it seems reasonable to suppose that the minimum cycle time depends mainly upon the depth of the footpiece and the size load. When the well has a tailpiece, extra time is required for the gas to break through the slug of liquid below the footpiece. This extra time appears to be about equivalent to the time required for the plunger to rise and fall through gas for a distance equal to the length of the tailpipe. For wells with tailpipes, therefore, the depth affecting the cycle time may be taken as the depth to the bottom of the tailpipe. More generally, the depth may be taken as the point where gas enters the tubing. This also is the depth found to be most suitable for computing the gas and pressure requirements. The operating conditions ordinarily are set so that rate of upward movement of the plunger averages about 1000 feet per minute. The downward fall through gas is about twice that speed. Accordingly, the number of minutes required for the plunger to rise to the surface and fall to the top of the liquid may be expressed as one and one-half times the depth in thousands of feet. Measurements and estimates of the time required for the plunger to unload liquid at the surface, and for the plunger to move down through liquid in the tubing, lead to a value in minutes equal to eight times the number of barrels per cycle for the 2-in. plunger and six times the number of barrels per cycle for the 2:-in. plunger. These considerations lead to expressions for the minimum cycle time: 2" 24"
+ 8L, tp = 1.5( 0/1000) + 6L,
tp = 1.5(
0/1000)
( 13-12) (13-13)
488
Dividing the number of minutes per day by minimum cycle time [eqs. 13-12 and 13-13] and multiplying by size load, eqs. 13-7 and 13-8, respectively, may be readily obtained for the maximum production rate expected from plunger lift. Of the data received and listed in Tables 13-1 and 13-11, ten wells correspond to rates of production that exceed 70% of the values computed from eqs. 13-7 and 13-8. For wells 3, 25, 26, 28, 29, and 37 with 2-in. plungers, the respective production rates were 73.7, 115.3, 119.2, 72.3, 73.5, and 73.5% of the values computed from eq. 13-7. For wells 3, 6, 29, and 48 with 2i-in. plungers, the respective production rates were 70.3, 92.5, 85.0, and 71.2% of the values calculated from eq. 13-8. These comparisons indicate that the rates expressed by eqs. 13-7 and 13-8 are not too difficult to attain.
CONSTRUCTING PLUNGER LIFT NOMOGRAPHS
Need for nomographs
After the equations were obtained, as described in the preceding section, the next step in, applying plunger lift was to derive an expression for production rate. This could be accomplished through use of the equations for net operating pressure and pressure buildup, in conjunction with the productivity index equation. Mathematical operations indicated by the resulting expression for production rate could be simplified by construction of a nomograph. This graphical treatment would also permit ready display of the range of production rates below the maximum. Method of constructing nomographs Estimating average pressure at point where gas enters tubing For a cyclical process, such as is involved in plunger lift, the average pressure at the point where gas enters the tubing may be estimated from the casing pressures at the surface, as described in the following paragraphs. The average casing pressure (at the surface) may be taken as the arithmetic average of the maximum and minimum casing pressures, or the maximum casing pressure less one-half the pressure buildup. The average pressure at the bottom of the column of gas may be obtained by adding the weight of the gas to the average casing pressure. This may be done through multiplication of the average presure at the top of the column by a factor appropriate for an average well. The resulting product may be added to the average pressure at the top. The average well was assumed to have a gas gravity of 0.75 (referred to air), a depth of 8000 ft, and an average casing temperature of 160'F. These values led to an average gas factor of 0.027 0/1000, and to the approximate relation: Pwf= [ P,""
-
+ ( P,""
-
PFin)] (1 + 0.0270/1000)
(13-14)
489
Derivation of operating line for 2-in. plunger In obtaining an expression for production rate to construct the operating line, use was made of the productivity index equation: (13-15)
4=J(Pw,-P,,)
Substituting the values of q = JP,, - J [ P,""
-
f ( P,""
F,,
from eq. 13-14 into eq. 13-15:
- P p ) ] (1
+ 0.0270/1000)
(13-16)
Combining the preceding with eqs. 13-1 and 13-5 gives
q = JP,, - J ( P y i n + 77.7L, + 8.29-
+f J
i
8.16L,
D 1000
+ 18.08L,-
D 1000
ii +
+ 67.4
+
1
0.0271000 (13-17)
Equation 13-17 may be rearranged to express q in terms of L, as follows:
i
1+0.027-
q=JP,,-J
i+
-J 1
0.027-
1000
j!0.990Pyi"
+ 6.58-1000 D + 62.41
D 1000
(13-18)
Equation 13-18 represents a straight line, if q is plotted against L, for given values of J , Pws, D, and P?". This is termed the operating line, and it may be determined by the intercepts on the axes, which are as follows: When q = 0,
L,
=
i
P,, - 1 + 0.027-
1000
)[0.990P,""
+ 6.58-1000 + 62.4) (13-19)
+
(I 0.027-
t 73.61
When L,=O, q=JP,,-J
i
1+0.027-
1000
+
] ( 0 . 9 9 0 P ~ " 6.58-1000 +62.4)
(13-20)
490
Derivation of operating line for 2: -in. plunger In steps analogous to those just described, expressions equivalent to eqs. 13-18, 13-19 and 13-20 have been derived for the 2i-in. plunger as follows: = JP,,
i+
-J 1
0.027-
1000
) (0.974PF'" + 1.125 1000 +113.2) -
(13-21)
When q = 0, P, L,
=
-
i+ i+ 1
0.027-
1
1000
0.027-
) (0.974P,m'" + 1.125 1000 + 113.2) -
+ 1.545)
1000
(13-22)
When L,=O,
1
[O.974Pyin
+ 1.125 1000 + 113.2)
(13-23)
Mathematical operations by nomographs The value of (1 + 0.027 D/lOOO) (0.990 Ptmin)may be isolated in eqs. 13-19 and 13-20 and subtracted separately from P,,. Then the intercepts are fixed by the value of D and the value of P, - (1 0.027 0/1000) (0.990 The left-hand chart in the nomograph for the 2-in. plunger (Fig. 13-2) contains curves of the intercept by eq. 13-19 plotted against depth for various values of P, - (1 + 0.027 D/lOOO) (0.990 PFin). The chart in the upper right-hand corner of this nomograph contains essentially straight lines, which subtract the value of (1 0.027 D/lOOO) (6.58 D/lOOO + 62.4) from P,, - (1 + 0.027 D/lOOO) (0.990 P p ) . The radiating lines in the upper center of the nomograph permit multiplication of the above difference by the productivity index, to obtain intercept 2. Inasmuch as the rate of production is limited by the time required for the plunger to complete a cycle, it was necessary to draw maximum production rate curves on the lower center portion of the nomograph. Each of these curves was plotted by substituting a given value of depth and various values of size load in eq. 13-7. The preceding mathematical operations expressed by the nomograph make it possible to locate graphically the two intercepts. The line drawn between the intercepts shows the operating line, which is that part of the line between intercept 1
+
Pp).
+
49 1
Fig. 13-2. (Nomograph 1). Example illustrated by dashed line is for a well with the following known conditions: depth of tubing = 6600 ft: minimum pressure at top of tubing = 60 psi; static hottomhole prcssurc = 800 psi: and productivity index = 0.12. For 2-in. plunger.
492
Fig. 13-3. (Nomograph 2). Example illustrated by dashed line is for a well with the following known conditions: dcpth of tubing = 9200 ft: minimum pressure at top of tubing = 1 0 0 psi; static bottomhole pressure = 2100 psi; and productivity index = 0.08. For 2l-in. plunger.
493
and the point where the line crosses the appropriate maximum production rate curve. Entirely analogous operations are performed by means of the nomograph for the 2 t - h plunger (Fig. 13-3). Derivation of supplementary operating line for 2-in. plunger The operation of plunger lift in a well with an excessive formation gas/liquid ratio may be determined by the ratio rather than by the anticipated minimum tubing pressure. Consequently, an expression was derived for the supplementary operating line, based on gas/ liquid ratio gradient instead of minimum tubing pressure. and This was accomplished by obtaining from eq. 13-3 the value for substituting it in eq. 13-18. Rearranging to express q in terms of L, gives = JP,,
i
- J 1 + 0.027-
i
1000
+ J 1 +0.027- D 1000
'IL
+ 294)
D + 18.08-1000 - 183.5) L , (13-24)
Equation 13-24 represents a straight line, if q is plotted against L, for given values of J , P,,, D , and ( G / L ) ( D / 1 0 0 0 ) . This is termed the supplementary operating line, and it might be determined by the intercepts on the axes. It is more convenient, however, to use the intersection of the two operating lines and the intercept on the production rate axis. The latter is called intercept 2A and is expressed as follows: When L,=O, (13-25)
It was desirable to use the same base (psi) lines on the nomograph for determining both intercepts, 2 and 2A. The chart in the upper right-hand corner of the nomograph contains lines which subtract the value of (1 + 0.027 D/lOOO) (6.58 0/1000 + 62.4) from P,, - (1 + 0.027 D/lOOO) (0.990 P,""n). Equation 13-25 does not contain a term involving P,"'ln, so the minimum tubing pressure is ignored in determining intercept 2A. This equation also shows that (1 + 0.027 D/lOOO) (6.09 D/lOOO) + 294) should be added to P,,,. To use the base lines in the upper right-hand chart as the first step in obtaining intercept 2A, it follows that the sum of the above difference, or (1 + 0.027 D/lOOO) (12.67 D/lOOO + 356.4), must be added to P,,. In this way, P,, will be increased both by the amount the lines subtract and by the amount shown in eq. 13-25. Thus,
494
the measurements for intercept 2A are made from the base lines drawn for intercept 2. The relation for the intersection of operating lines may be obtained by simultaneously solving the equations for the two operating lines. The result is the equation for gas/liquid ratio gradient which was solved simultaneously with the main operating line in deriving the supplementary operating line. In the case of the 2-in. plunger, the expression for the intersection is the same as eq. 13-3. It is explained below why the intersection of operating lines will lie on the dotted extension of the main operating line. There is also included a graphic explanation of the fact that the intersection may be located by means of the appropriate equation for gas/liquid ratio gradient. The only time the supplementary operating line is of interest is when the formation gas/ liquid ratio gradient is greater than that required for the maximum production rate determined by the main operating line. Along this operating line, the minimum tubing pressure is fixed while the gas/liquid ratio gradient continuously increases with decreasing size load. As a consequence, there must be a point on the dotted extension of this operating line where the gas/liquid ratio gradient required for the size load at the point is equivalent to the formation gas/liquid ratio gradient. This point is the intersection of operating lines. Along the supplementary operating line, the formation gas/ liquid ratio gradient remains constant, while the minimum tubing pressure continuously decreases with decreasing size load. Consequently, there must be a point on the supplementary operating line where the minimum tubing pressure required for the size load at the point is equivalent to the fixed value of minimum tubing pressure along the main operating line. This point also is the intersection of operating lines. From the foregoing, it follows that the lines intersect at the point where the size load has the value required by both the minimum tubing pressure of the main operating line and the formation gas/ liquid ratio of the supplementary operating line. This size load may be computed by means of the equation of gas/liquid ratio gradient. The point corresponding to the computed size load lies on both lines, so it is readily located on the one that has already been drawn. Accordingly, the intersection of operating lines is located at the point on the main operating line that corresponds to the size load computed from the equation for gas/liquid ratio gradient. Derivation of supplementary Operating line for 24 -in. plunger In steps analogous to those described above, relations similar to eq. 13-24 and 13-25 have been derived for the 2i-in. plunger: G/L
+ 11.11- D
- 159.51 L,
1000
i
+ J 1 S0.027-
D 1000
(13-26)
495
When L,
= 0,
(13-27) In using the lines in the upper right-hand corner of the nomograph as the first step in determining intercept 2A, the value that should be added to P,, is (1 + 0.027 D/lOOO) (63.1 D/lOOO + 335). To locate the intersection of operating lines for the 2i-in. plunger, eq. 13-4 should be used. Method of testing effects of various well conditions
The effects of change in tailpipe, water cut, and oil gravity were tested by the method of least squares and found to be insignificant. The method was applied to the net operating pressure and to the cycle gas gradient, for the wells with 2i-in plungers, as described below. After eqs. 13-2 and 13-11 were obtained, the method of least squares was applied again, with one of the well conditions as an added independent variable. In this way, new coefficients of the variables and a new coefficient of multiple correlation were determined. In each case, the coefficient or index of multiple correlation was found to increase so slightly, if at all, that the added variable did not significantly increase the correlation. Consequently, it was believed unnecessary to apply the test to the pressure buildup for the 21-in. plungers or to any of the factors of pressure or gas for the 2-in. plungers. Tailpipe A well was considered to have a tailpipe if a pipe extended below the footpiece by a distance equal to 0.3% or more of the distance from the surface to the bottom of the pipe. By this definition, 44 of the wells with 24-in. plungers had tailpipes ranging from 0.3 to 27.756, and averaged 3.25%. As stated earlier, the least-squares equation for net operating pressure, which was obtained from the 75 wells with 2;-in. plungers, had an index of multiple correlation of 0.945. When tailpipe was included as an added first-power variable, the index of multiple correlation was raised to 0.946. This increase in the index is negligible because these values yield a partial coefficient of multiple correlation that is not significant (Davies, 1949). The least-squares equation for cycle gas gradient, which was obtained from the 75 wells with 21-in. plungers, had a coefficient of multiple correlation of 0.639. After the tailpipe was included as an added first-power variable, the coefficient of multiple correlation was found again to be 0.639. The definition of depth which is used throughout this chapter is the distance from the surface to the point at which gas enters the tubing. This would be at the bottom of any tailpipe. The negligible increase, if any, in the coefficients of multiple correlation indicates that no additional notice need be taken of tailpipes similar to those in the wells studied.
496
Water cut All the equations developed in this chapter include the water in the units of barrels per day and barrels per cycle. Nevertheless it was believed advisable to test the effect of water cut on the requirements of pressure and gas. The least-squares equation for net operating pressure, which was obtained from the 75 wells with 2f-in. plungers, had an index of multiple correlation of 0.945. When water cut was included as an added first-power variable, the index of multiple correlation was found again to be 0.945. The least-squares equation for cycle gas gradient, which was obtained from the 75 wells with 2f-in. plungers, had a coefficient of multiple correlation of 0.639. The water cut, as volume per cent of total production, was included as an added first-power variable; the coefficient of multiple correlation was then found to be 0.641. This increase in the coefficient is negligible (Davies, 1949), so water cut may be ignored in estimating plunger lift requirements. Of course, the produced water must be included in the production rate and in the size load. For the 75 wells with 2i-in. plungers, water cuts ranged from 0 to 88.956, with an average of 12.6%. Oil gravity The least-squares equation for net operating pressure, whch was obtained from 75 wells with 2t-in plungers, had an index of multiple correlation of 0.945. After oil gravity was included as an added first-power variable, the index of multiple correlation was found to be 0.947. This increase in the index is negligible (Davies, 1949). The least-squares equation of cycle gas gradient, which was obtained from the 75 wells with 2i-in. plungers, had a coefficient of multiple correlation of 0.639. When oil gravity was included as a first-power variable, the coefficient of multiple correlation was found to be 0.678. This increase in the coefficient is not significant (Davies, 1949) because these values yield a partial coefficient of multiple correlation of 0.029, which is not significant even for a total number of observations as high as 75. For the oils produced from wells with 2i-in. plungers, the gravities ranged mainly from 24 to 46" API (one value was 21"), averaging 33.7" API. Consequently, no allowance need be made for this property in plunger lift calculations, for oil gravities from 24 to 46' API.
HOW TO USE PLUNGER LIFT NOMOGRAPHS
Purpose of nomographs
The purpose of the nomograph is to permit graphical solution of the problem of estimating plunger lift performance for a given well. Construction of the nomo-
491
graphs was based on formulas derived by the method of least squares from actual field data. Supplementary figures in this series provide a means for determining the probable requirements of gas volume and casing pressure, when size load and production rate are read from a nomograph. Solution of the problem is obtained by plotting the operating line for a well. This line relates size load to gross production rate and, therefore, represents a well’s ability to produce by the plunger lift method. For a given depth, there is an upper limit to the production rate, the limit being fixed by the time required for the plunger to complete a cycle. Accordingly, the nomographs include curves for various depths which represent the maximum production rates attainable by plunger lift for various size loads. The intersection of a well’s operating line and the appropriate limiting depth curve, therefore. will indicate the highest rate of production expected for the well that is being plotted. By accepting lower production rates, the plunger may be operated with larger size loads and at higher net operating pressures to obtain lower gas/liquid ratios. Daily volumes of gas and the corresponding cost may readily be compared with rates of production to determine the most economical operation for the well being studied. The effect of a well’s formation gas/liquid ratio upon the production rate may be predicted by plotting a supplementary operating line. Nomograph instructions In using the nomographs to estimate plunger lift performance in a given well, aside from tubing and casing sizes, the following information is needed: (1) depth, (2) static bottomhole pressure, (3) productivity index, and (4) anticipated minimum tubing pressure. The depth is taken as the distance from the surface to the bottom of the tubing (or to the point at which gas enters the tubing from the casing). The static bottomhole pressure should be referred to this depth. The productivity index is considered to be a constant in the nomographs, and so should be an average value, if known to vary with the production rate. The values used for static bot tomhole pressure and productivity index should be as recent as possible and consistent with the known production rate and the computed or measured flowing bottomhole pressure. The minimum tubing pressure should be estimated from the trap or separator pressure, and from expected pressure losses in the flowline, if not known from field observation. Whenever static pressure and productivity index are unknown, it may be possible to estimate their limits and plot a pair of operating lines to obtain the probable range of plunger lift performance. Throughout the nomographs and figures, interpolation is permitted between lines representing depth, pressure, productivity index, and size load.
Step I Multiply the anticipated minimum tubing pressure by 0.990 (1 + 0.027 0/1000) for the 2-in. plunger or by 0.974 (1 0.027 0/1000) for the 2141-1. plunger. Subtract this amount from the static bottomhole pressure to obtain the value for use in steps I1 and 111.
+
498
Step I1 Enter the left-hand chart of the nomograph and proceed up the appropriate depth line to the curve representing the value obtained in step I. At this point, turn right and move horizontally to the barrels-per-cycle (bbl/cy) axis. This point is designated intercept 1.
Step I I I On the upper right-hand chart, proceed down the proper depth line to the value determined in step I. Now turn left and move horizontally to the line representing the productivity index of the well. Then turn and go vertically down to the bbl/day axis. This point is designated intercept 2 .
Step IV Connect intercepts 1 and 2 with a straight line which is solid from intercept 1 to the appropriate depth curve, and is dashed from that point to intercept 2. The solid portion of this line is the operating line. Directly below the intersection of the operating line and the depth curve may be read the maximum production rate expected for the well. The value on the bbl/day axis opposite this intersection is the size load which will be lifted at the maximum production rate. The number of cycles per day is computed by dividing bbl/day by bbl/cy, and the number of minutes per cycle is equal to 1440 divided by the number of cycles per day. Use Fig. 13-4 (for 2-in.) or Fig. 13-6 (for 2i-in.) to find the required gas/liquid ratio gradient (gas/ liquid ratio divided by depth in thousands of feet) from the predicted size load, the depth, and the expected minimum tubing pressure. Multiply this value (for gas/liquid ratio gradient) by the depth in thousands of feet (0/1000) to determine the required gas/liquid ratio. Multiply the latter figure by the predicted production rate to obtain the gas volume. Use Fig. 13-5 or Fig. 13-7 to find the net operating pressure from the predicted size load and known depth. Add this value to the minimum tubing pressure to obtain the probable maximum casing pressure.
Step V-A Compare the required gas/liquid ratio gradient with the expected formation gas/liquid ratio gradient. If the probable gas requirement is greater than that which the well is expected to make, there are two possible situations. If outside gas is available, it may be circulated at a constant rate down the casing of the well in an amount necessary to increase the gas/liquid ratio gradient to the required value. In the second situation where no outside gas is available, use eq. 13-3 or eq. 13-4 to compute the size load which can be lifted with the expected formation gas/liquid ratio gradient, at the known depth and anticipated minimum tubing pressure. Mark a point on the operating line opposite this value of bbl/cy, and read the predicted gross production rate on the bbl/day axis directly below this point. The values of cycles per day, minutes per cycle, and maximum casing pressure may be computed as in step IV.
499
4
5
7 DEPTH, Tn-D
e
10
I1
I2
FEET
Fig 13-4 Gas/liquid ratio gradient versus depth (part A) and versus minimum tubing pressure (part B) for various size loads Add reading of part A to that of part B (2-111 plunger)
500
ZW MINIMUM
300
T U I l N O PRCIIURE,
400
PI1
Fig. 13-4 continued.
Step V-B I f the required gas/liquid ratio gradient is less than that expected from the well, there are two alternative courses. In the first of these, the excess gas is allowed to
ISd '3tlnSS3Ud DNIlVtl3dO 13N k
Y Y
i
P,
L
%
1
3
P
-
.-c
0
.-
L
P
Fig. 13.5. Net operating pressure versus depth for various size loads (2-in. plunger). 501
502
4
5
0
7 8 DEPTH.THOIJSND
0 FEET
10
II
I2
Fig 13-6 Gas/liquid ratio gradient versus depth (part A) and versus minimum tubing pressure (part B) for vanous size loads Add reading of part A to that of part B (2i-in plunger)
flow out of the tubing or the casing, either intermittently or at a constant rate, to obtain the maximum production rate predicted in step IV. The second alternative involves operation of the plunger in the standard manner while accepting the production rate obtainable under these conditions. In t h s second alternative, where no provision is made to remove the excess gas,
503
MINIMUM TUBING CRLSUIII?, P.S.I.
Fig. 13-6 continued
the effect of the excess gas on the production rate may be determined by plotting a supplementary operating line as follows: (1) Refer to eq. 13-3 (for 2-in.) or eq. 13-4 (for 2t-in.) and compute the size load which corresponds to the expected minimum tubing pressure, depth, and formation gas/liquid ratio gradient. Make a mark at this size load upon the dashed extension
504
Fig. 13-7. Net operating pressure versus depth for various size loads ( 2 l - h plunger).
505
of the operating line on the nomograph. Let this point be designated the "intersection of operating lines." (2) This differs from step I11 in that the minimum tubing pressure is not considered. Furthermore, a value is computed which is to be added to the static bottomhole pressure: for the 2-in. plunger, the added value is (1 0.027 0/1000) (12.67 0/1000 356); in the case of the 24-in. plunger, the value is (1+ 0.027 0/1000) (63.1 0/3000 + 335). Re-enter the upper right-hand chart and proceed down the proper depth line to the appropriate value of static bottomhole pressure plus the value just computed. Now turn left and move horizontally to the proper productivity index line. Then turn and go vertically down to the bbl/day axis. This point is designated intercept 2A. (3) Draw a straight line from intercept 2A through the intersection of operating lines, which is solid above and dashed below the point at which the line intersects the appropriate depth curve. Directly below the intersection of the supplementary operating line and the depth curve may be read the production rate predicted for the given conditions. The value on the bbl/cy axis opposite this intersection is the size load which will be lifted at this production rate. The values for cycles per day and minutes per cycle may again be computed as in step IV. The minimum tubing pressure is computed from eq. 13-3 or eq. 13-4 using the size load just determined and the given gas/liquid ratio gradient and depth. Use Fig. 13-5 or Fig. 13-7 to find the net operating pressure from the size load just determined and the known depth. Add this value to the minimum tubing pressure, to obtain the required maximum casing pressure.
+
+
PLUNGER LIFT NOMOGRAPHS AND EXAMPLES
Types of example wells
The nomograph instructions given in the preceding section have been applied to examples illustrating two different types of wells. For the first example well, illustrated on Nomograph 1 (Fig. 13-2) for the 2-in. plunger, the probable gas/liquid ratio required at maximum production rate is considerably higher than the well's formation gas/liquid ratio. For the second example well, on Nomograph 2 (Fig. 13-3) for the 2f-in. plunger, it is considerably lower, so the situation is reversed. Nomograph Examples Data Plunger size, in.
Tubing size, o.d., in. Casing size, o.d., in. Tubing depth, D , f t Anticipated minimum tubing pressure, P,""",psi Productivity index, J , (bbl/day)/psi Static bottomhole pressure, Pw,, psi Gas/liquid ratio, G / L , cu ft/bbl Gas/liquid ratio gradient, ( G / L ) (0/1000), (cu ft/bbl)/(ft/1000)
2 2; 5;
2;
6600 60 0.12 800 3035 460
9200 100 0.08 2100 6900 750
2; 7
506
bv Nomograph Step I : Pw,, psi P,"'"(0.990)(1+ 0.027 0/1000). psi P,"'"(0.974)(1 +0.027 0/1000). psi Pw, - P,"'"(0.990)(1+0.027 0/1000), psi Pw,- P~'"(0.974)(1+0.027 0/1000), psi
Solution
Step 11: Intercept 1, bbl/cy Step I l l : Intercept 2, bbl/day Step I V : Maximum production rate, bbl/day Size load at maximum production rate, bbl/cy Cycle frequency, cycles/day Cycle time, minutes/cycle Required gas/liquid ratio gradient (cu ft/bbl)/(ft/1000) Required gasJliquid ratio, cu ft/bbl Required daily gas volume, Mcf/day Net operating pressure, psi Minimum tubing pressure, psi Maximum casing pressure, psi Step V - A : Required gas/liquid ratio gradient, (cu ft/bbl)(ft/1000) Formation gas/liquid ratio gradient, (cu ft/bbl)/(ft/1000) Difference between required and formation gas/liquid ratio gradient. (cu ft/bbl)/(ft/1000) First alternative: Required volume of outside gas to make up deficiency and attain maximum production rate. Mcf/day Second alternative: Size load which can be produced on formation gas alone. bbl/cq Production rate at above size load, bbl/day Cycle frequency, cycles/day Cycle time, minutes/cycles Net operating pressure, psi Minimum tubing pressure. psi Maximum casing pressure. psi Step V-B: First alternative: Volume of excess gas to be removed, Mcf/day Second alternative: 1. Size load at given value of minimum tubing pressure and formation gas/liquid ratio gradient, for locating intersection of operating lines, bbl/cy 2. Intercept 2A, bbl/day 3. Production rate, bbl/day Size load at above production rate. bbl/cy Cycle frequency, cycles/daq Cycle time, minutes/cycle Minimum tubing pressure. psi Net operating pressure. psi Maximum casing pressure. psi
800 70
2100
-
122
730 -
1978
2.66
72.5
57.0 0.58 98 15 880 5808 331 236 60 296
-
14.08
146
121.5 2.36 51.5 28 454 4177 508 394 100 494
880 460
454 750
420
-296
158
-
1.51 31.7 21 69 420 60 480
-
-
-
331
-
1.22 241 98 1.6 61 24 421 314 135
-
-
507
Accuracy expected from results
As stated earlier, the equations for gas and pressure were derived from correlations of field data by the method of least squares. This method yields expressions that represent actual field experience. Some of the data may have been submitted for plungers that were not being operated under ideal conditions. Nevertheless, the net operating pressure, gas/ liquid ratio, and pressure buildup that were being used undoubtedly were the values required for the chosen set of operating conditions. In operating a plunger, it has always been possible to control the pressure in the casing and the volume of any outside gas that was being circulated. Before the equations and nomographs were available, however, it was not easy to choose the operating conditions most suitable for a well. The above factors are interdependent with size load and minimum tubing pressure. Consequently, the values chosen for maximum casing pressure and gas/liquid ratio will influence size load and minimum tubing pressure. It follows, therefore, that the well data submitted for various sets of operating conditions were suitable for deriving the equations relating these interdependent variables. The method of least squares permitted determination of these relations in terms of probable or average values, thus minimizing the effects of errors and other variations, and in addition, had the usual advantage inherent in a method based on experience. Another major advantage in such correlation equations is the opportunity of choosing safety factors that correspond to practically any degree of certainty desired in results obtained through use of the equations. This is amplified by the discussion of accuracy in the paragraphs that follow. The accuracy to be expected from use of the equations for gas and pressure is illustrated by Figs. 13-8 and 13-9. To obtain the curves for the figures, the data for a well were substituted into the appropriate equation and a factor, such as net operating pressure, was calculated. This was subtracted from the actual reported data and the difference was expressed as per cent of the calculated value. These difference percentages were arranged in order of increasing magnitude. The values at various fractions of the total number of cases were then plotted and a smooth curve was drawn among the points. Any point on such a curve indicates that the fraction of cases directly below the point had difference percentage equal to or less than the value horizontally opposite the point. As an example of use of the curves, it may be noted in Fig. 13-8 that for a 2-in. plunger the actual net operating pressure for 0.5 of the cases fell within +24% and for 0.9 of the cases within +55% of the calculated values. These would be the variations which should include the actual net operating pressure for one out of two new cases, and nine out of ten new cases, for which the appropriate equation is applied. For the 2i-in. plunger, the corresponding values are +12% and *47%. These conclusions stem from the fact that 0.5 and 0.9 of the cases studied fell within these variations and the same should hold for new cases. A percentage variation (or safety factor) which is allowed in predicting net
508
FRACTION
OF CASES
WITH
DIFFERENCES
LESS
THAN SCALE READING
Fig. 13-8. Accuracy expected from least-squares equations (2-in. plunger).
operating pressure usually should lie somewhere between the above pairs of values, depending on the degree of certainty desired. For practical purposes, the same variation may be allowed in predicting production rate. The last statement does not concern any inaccuracies involved in the productivity index equation. This relation, with its advantages and disadvantages, has been the subject of many articles in the past and doubtless will be the subject of many more in the future. The equation has become an important tool in predicting production rates, and proof of its validity is beyond the scope of the present chapter. As stated in the Nomograph Instructions given in the preceding subsection, an average value for the productivity index should be used, if it is known to vary with production rate. It should also be emphasized that unless both the static bottomhole pressure and the productivity index have been determined quite recently, one of
509
FRACTION
OF CASES WITH
DIFFERENCES
LESS
THAN
SCALE
READING
Fig. 13-9. Accuracy expected from least-squares equations ( 2 :-in. plunger).
these should be computed. For example, if the static pressure is the one more recently determined, it should be substituted into the productivity index equation, along with the production rate and the flowing bottomhole pressure which has been measured or calculated from the well’s operating conditions. This permits computation of the productivity index for use with the nomographs. The accuracy to be expected from use of the equations for gas/liquid ratio gradient may also be read from the curves of Figs. 13-8 and 13-9. For the 2-in. plunger, 0.5 and 0.9 of the cases were within 30% and 5-7376, respectively; for the 2i-in. plunger, 0.5 and 0.9 of the cases were within 21% and 57%, respectively. A similar statistical analysis of accuracy cannot be made for the maximum production rate equations. However, this is not a serious disadvantage. These equations are used in the nomographs only for fixing an upper limit to the gross
510
production rate expected from plunger lift. Furthermore, operating conditions may be set so that the maximum production rates given by the equations are actually attained.
PLUNGER LIFT APPLICATIONS
Use of equations and figures without plotting operuting line
When the static bottomhole pressure and the productivity index are known or may be estimated, a well’s operating line may be drawn on a nomograph, as has been described earlier. Even when such bottomhole data are not available, the equations for net operating pressure and gas/liquid ratio gradient may be used to advantage. Any oil or gas well has limitations as to pressure or volume of gas, and one of these factors is more critical than the other. By substituting a value for the more critical factors into the appropriate equation, the size load may be computed. This computed size load may be substituted into the companion equation, and a value may be calculated for the less critical factor. This process may be repeated, starting with various values of the more critical factor. The method is illustrated in Table 13-IV, and it may be applied either before or after installing plunger lift equipment. Before installation, the calculations are used in predicting future performance. After installation, the method is used in adjusting the operating conditions to those most suitable for the well. Types of oil wells suitable for plunger lift General considerations The plunger can be applied to all wells that are within its capacity and where sufficient gas and pressure are available. These include weak-flowing wells, small producers with high gas/oil ratios, and wells on gas lift that are using excessive volumes of gas. Also included are wells where the accumulation of bottomhole water, paraffin, or an emulsion is a production problem. Plunger lift is especially suitable for the deeper well with a medium-to-high static bottomhole pressure and low productivity index. This method may be economically applied to the difficult intermediate step between flowing and pumping, usually when gas/oil ratios are high. The plunger is considered a depletion device for deep, tight formations. The standard plunger lift installation consists of surface controls and a simple, open system without a tubing packer. Experience has proven that flowing bottomhole pressures can be obtained by this system that are generally lower than those by any other gas-lift method. In some wells, it becomes necessary to lift small loads of oil with each cycle. In such cases, a plunger acting as a piston can lift the oil with less pressure, less volume
511 TABLE 13-IV Example wells for use of equations Description
Value
Example well with limited net operating pressure Data: Tubing size, o.d., in. Tubing depth, ft Anticipated minimum tubing pressure, psi Pressure of outside gas available, psi Anticipated production rate, bbl/day
2; 8000 40 500 70
Solution by equations orfigures: From data, net operating pressure, psi From eq. 13-1 or Fig. 13-5, size load, bbl/cy From eq. 13-3 or Fig. 13-4, gas/liquid ratio gradient, (cu ft/bbl)/(ft/1000) Probable gas/liquid ratio, cu ft/bbl Probable daily gas volume, Mcf/day
460 1.47 466 3728 261
Example well with limited gas/liquid ratio Data: Tubing size, o.d., in. Tubing depth, ft Minimum tubing pressure, psi Gas/liquid ratio, cu ft/bbl
2; 8600 60 4000
Solution by equations andfigures: From data, gas/liquid ratio gradient, (cu ft/bbl)(ft/1000) From eq. 13-4, size load, bbl/cy From eq. 13-2 or Fig. 13-7, net operating pressure, psi Probable maximum casing pressure, psi
465 2.07 346 406
Example gas well unloading water to atmosphere Dara: Tubing size, o.d., in. Tubing depth, f t Casing pressure, psi Water production rate, bbl/day
2; 4200 400 4
Solution by equations or figures: From data, net operating pressure, psi From eq. 13-1 or Fig. 13-5, size load, bbl/cy From eq. 13-3 or Fig. 13-1, gas/liquid ratio gradient, (cu ft/bbl)/(ft/1000) Probable gas/liquid ratio, cu ft/bbl Probable daily gas volume used in unloading, Mcf/day
400 1.94 366 1537 6.46
Example gas well producing to sales line Data: Tubing size, o.d., in. ' Tubing depth, ft Sales line pressure, psi Casing pressure, psi Water production, bbl/day
2; 4000 650 860 150
512
TABLE 13-IV (continued) Description
Value
Solution h.v equutions upid figures: From data, net operating pressure, psi From eq. 13-2 or Fig. 13-7, size load, bbl/cy Cycle frequency, cycles/day Cycle time, minutes/cycle From eq. 13-4, gas/liquid ratio gradient, (cu ft/bbl)/(ft/1000) Probable gas/liquid ratio to lift water, cu ft/bbl Probable daily gas volume used in lifting, Mcf/day
210 1.99 75 19 587 2348 3 52
of gas, less slippage, and less chance of forming emulsions than is the case with any other gas-lift method. For best results, oil gravities probably should be above 24”API. Plunger lift will produce wells with water cuts ranging from 0 to loo%, providing the gross production rate is within the “capacities” listed later in this section. The plunger is excellent for preventing the build-up of paraffin on tubing walls and often in flowlines as well. This is due to the combination of the intermittent gas flows, followed by warm slugs of liquid, and the scraping action of the plunger. The plunger can operate in the tubing, no matter how crooked the hole. It has not been determined how great the deflection from the vertical would have to be to cause an adverse effect upon plunger operation. But it is believed that a steep angle would tend to slow the plunger, so that its maximum effectiveness could not be realized. Capacities The maximum (gross) production rates of wells ordinarily considered suitable for plunger lift are listed below: Depth (ft)
Barrels per day Plunger size
4000 6000 8000 10,000 12,000
2-in.
2 5 -in
130 115 105 95 85
190 175 160 150 140
How to determine if a well is a possible plunger lift candidate The suitability of an oil well for the plunger lift system is best determined by plotting the well’s operating line on a nomograph, as described earlier. A helpful
513
NET OPERATING PRESSURE-POUNDS PER SQUARE
INCH
Fig. !3-10.Application curves based on gas and pressure requirements (2-in. plunger).
preliminary appraisal, however, may be made by using Fig. 13-10 or Fig. 13-11 to determine whether a well is a possible candidate for plunger lift. Use of Figs. 13-10 and 13-11 The curves in Figs. 13-10 and 13-11 represent probable requirements of gas and '
514
-1
U
d
c Y
n
NET OPERATING PRESSURE-POUNDS PER SQUARE INCH
Fig. 13-11. Application curves based on gas and pressure requirements (2i-in. plunger).
pressure. The gas/liquid ratio is plotted against the net operating pressure. The latter is the amount of the casing pressure that is over and above the trap or separator pressure. In computing the gas/liquid ratio for Figs. 13-10 and 13-11; the trap or separator pressure was taken equal to zero.
515
Because these curves represent average values, 2000 f t should be subtracted from the well’s tubing depth, in using the curves to determine whether a well’s operating line should be plotted. In using the figures, the procedure to follow depends upon whether the well must rely entirely on formation gas or there is additional gas available from some outside source; that is, whether gas volume or pressure is the more critical factor. Well depends entirely on formation gas Use Fig. 13-10 for a well with 2-in. tubing or Fig. 13-11 for a well with 2+-in. tubing. Starting at the well’s gas/liquid ratio, proceed horizontally to a curve representing the well’s tubing depth less 2000 ft. Directly below this point, read the value of net operating pressure. If this pressure appears feasible, the well’s operating line should be plotted for a complete analysis. Well has gas available from some outside source Select a casing pressure acceptable for producing the well. Subtract from this value the.pressure of the flowline or separator. Enter Fig. 13-10 (for 2-in.) or Fig. 13-11 (for 24-in.) at this net operating pressure, and proceed vertically to a curve representing the well’s tubing depth less 2000 ft. Horizontally opposite this point, read the value of the gas/liquid ratio. If this ratio appears reasonable, the well is a possible candidate and its operating line should be plotted. Intermittent flowing and plunger lift system Before completing a new well, it may be profitable to estimate the likelihood of its developing into one suitable for plunger lift. In this regard, it is helpful to project into the future the probable ranges of the well’s static bottomhole pressure and productivity index. These estimated values may be used to plot a series of operating lines on a nomograph by the method described earlier. These lines will indicate: (1) whether the well will become suitable for plunger lift, (2) when to start the plunger, and (3) whether the well may be economically depleted by plunger lift. The operating lines plotted for the well at various estimated stages of production may indicate that the well will become suitable for plunger lift. In this case, it will pay to consider equipping the well at the start for the intermittent flowing and plunger lift system. This system involves the following producing steps, during any of which outside gas may be added: (1)continuous flow, ( 2 ) intermittent flow by automatic stopcocking, and (3) producing with plunger lift. Equipping wells at start for intermittent flowing and plunger lift (free piston) system Tubing program (1) When run, the tubing should be drifted or sized to 1.915 in. for the 2-in. plunger or 2.355 in. for the 2i-in. plunger.
516
(2) The tubing should be the same size all the way down and should be run as deep as practicable, inasmuch as this will help in the later stages of depletion. (3) An API pump shoe may be run on the bottom of the tubing string, for later setting a footpiece spring. (4) If a packer is desired in the well, consideration should be given to installing a side-door choke at the same time. In wells suitable for such a practice, this will permit the use of outside gas without removing the packer, in case additional gas is finally required for the operation of plunger lift. Christmas tree A plunger lift Christmas tree should be installed, so that it will later accommodate a plunger. The master gate valve should have a full, round opening of the same size as the tubing string. Such a tree is just as useful for all prior operations, and no major change need be made when the well is ready for installation of the rest of the plunger lift equipment. Well starting There should be a means of starting the well after natural flow ceases. Gas from an outside source, such as a compressor, should be available at sufficient pressure for starting the well, or kick-off valves should be installed to permit the use of gas at a lower pressure. Preferably, these should be outside valves which will permit normal passage of the standard plunger through the string of tubing. Cycle controller for intermittent flowing In the producing step involving intermittent flow by automatic stopcocking, a cycle controller would be installed to open and close the flowline. This controller should be the same one needed later with plunger lift. Advantages of intermittent flowing and plunger lift system The system involving intermittent flowing and plunger lift has several advantages over any other system of producing a well. Some of the advantages are the following: (1) The system is so versatile and flexible that it will handle a well throughout wide ranges of bottomhole conditions, production rates, gas/ oil ratios, and water cuts. (2) The formation gas is always used in lifting, and is never a detriment. (3) From completion to the final step of depletion by plunger lift, the tubing need not be pulled to progress through the steps outlined in t h s system. (4) If no tubing packer is used, as is generally recommended, the operator will have the advantage of more easily obtaining bottomhole data. This is especially true, if a bonnet is installed designed for running instruments into the annulus. ( 5 ) All controls are located at the surface. (6) Inasmuch as the well is always in a state similar to natural flowing, there is a beneficial effect on the formation.
517
Gas well applications
The application of plunger lift is being made in gas wells for removing liquid accumulations which hold back gas production. It has been found that gas production is substantially increased by the automatic removal of even very small volumes of water per day. It is necessary to flow some gas wells at very high rates to produce water, or to blow the wells frequently to atmosphere in removing the water. In such cases, it is found advantageous to use a plunger to gain better control in producing the well. Removing the water with a plunger, while producing gas at a lower rate, results in a much gentler action on the formation. In a well which has sufficient pressure, a plunger may operate while the gas is being produced through a separator directly to the sales line. Where this is not practicable, the gas should be produced from the casing to the sales line, and the plunger should be cycled periodically to remove liquid from the tubing to the atmosphere or low-pressure separator. In this latter application, it is necessary to use the trigger to insure a minimum loss of gas during the periodic unloading of the tubing. Ordinarily a timer is used for the automatic control of plunger lift in gas wells. It is possible, however, to operate the plunger manually instead of automatically, especially if no more than two cycles per day are required. The free piston, being a versatile tool, may be operated in a variety of ways to produce gas wells with wide ranges of gas and liquid production rates. The volume of produced gas may be any value above the minimum gas requirement described in the subsection involving Figs. 13-10 and 13-11. The liquid “capacities” listed in the subsection following Table 13-IV apply also to plunger lift operation in gas wells.
PREDICTING PLUNGER LIFT PERFORMANCE
Need to use tubing bottom as reference depth
Static bottomhole pressures and productivity indexes often are referred to the midpoint of the perforations. For use with wells to be produced by free-piston lift, accuracy of predicting well performance is greatly increased by referring these values to the bottom of the tubing. If gas is to enter at some depth other than the bottom of the tubing, the point of reference should be taken at the depth where gas is to enter the tubing, as was done to derive operating lines for plunger lift. In this way, there is a column of gas in the annulus from that depth to the surface while the plunger is rising. Furthermore the height of liquid above that depth ordinarily is negligible during the rest of the free-piston cycle. This permits adding the weight of the gas column to the average casing pressure in order to obtain the average flowing pressure at the bottom of the tubing. That
518
sum and the static bottomhole pressure were introduced into the productivity-index equation, to obtain the operating line. Both the static bottomhole pressure and the productivity index, therefore, should be referred to the depth at which gas enters the tubing, whxh is usually the bottom of the tubing. It is readily apparent that the static bottomhole pressure varies with depth according to the static-pressure gradient. It is also true that the productivity index usually varies with depth because the static-pressure gradient and the flowing-pressure gradient usually are not equal. This becomes evident upon consideration of the productivity-index equation: 4 = JPW, - Pwf)
(13-28)
If Pw, and Fwfare changed by different amounts in referring them to a new depth, the value of Pw, - Fw,is changed. Accordingly, for a given q, the computed value of J will be different. For a thick zone, the difference may be appreciable, as is illustrated by an actual well described in the next subsection. Example of change in productivity index with depth
According to Horton (1959), Well I was perforated from 7350 to 8590 ft, with the midpoint at 7970 f t and the bottom of the tubing at 8490 ft. The measured static pressure was 394 psi at the midpoint and 600 psi at the bottom of the tubing. When the well was producing 38 bbl/day of 33" API oil with a 34% water cut, the measured flowing pressure was 192 psi at the midpoint and 220 psi at the bottom of the tubing. Substituting these values in eq. 13-28, the calculated productivity index is 0.188 (bbl/day)/psi at the midpoint and 0.100 (bbl/day)/psi at the bottom of the tubing. These values differ greatly with depth and the proper one should be used in predicting free-piston performance. With the bottom of the tubing at 5490 f t as at present, an index of 0.100 (bbl/day)/psi and a static pressure of 600 psi should be used. If the tubing bottom were at the midpoint, an index of 0.188 (bbl/day)/psi and a static pressure of 394 psi should be used. If the tubing bottom were at the top of the zone, still a different pair of values should be used. Obtaining static and index consistent with operating conditions
If a well is being produced by gas lift or by free-piston lift and a fairly accurate value is known for either the static bottomhole pressure or the productivity index, a value is readily computed for the other, which is consistent with the production rate and the operating conditions. The method given herein has been found to increase so greatly the accuracy of predicting performance that it should be used in all cases, unless very accurate recent values have been obtained for both the static bottomhole pressure and the productivity index at the depth where gas will enter the tubing.
519 TABLE 13-V Estimated values of static pressure and productivity index for well 1 Case
Static pressure 7970 ft (psi)
Static pressure 8490 ft (psi)
(psi/ft)
1 2 3 4
394 394 394
600 600 550 589 628 578 617 656
0.300 0.375 0.450 0.300 0.375 0.450
5
6 7 8
-
-
a
’
Gradient
’
’ ’
Productivity index 7970 ft [(bbl/day)/psiI
Productivity index 8490 ft [(bbl/day)/psiI
-
0.100 0.100 0.115 0.103 0.093 0.106 0.096 0.087
-
0.188 0.188 0.188
’
’
Values either not available, or not applicable. Values given or assumed. All other values were computed from a production rate of 38 bbl/day and a flowing pressure of 220 psi at 8490 ft.
a
The method involves obtaining, for a given production rate, an average casing pressure to which is added the weight of the column of gas down to the bottom of the tubing. This yields the average flowing pressure at the point where gas enters the tubing. If either the “static” or the “index” is known for that depth, the value may be introduced into the productivity-index equation, along with the average flowing pressure. This permits calculating the one that is unknown, to obtain a pair of values consistent with the operating conditions. (See cases 1 and 2 of Table 13-V.) In the case for which the static bottomhole pressure is known at some other depth, the value of the static at the desired depth may be estimated from the static pressure gradient. If unknown, this may be taken as the pressure gradient of the oil at reservoir temperature, reduced by an amount consistent with the formation gas/liquid ratio. These values of the average flowing pressure and the static at the tubing depth are introduced into the productivity-index equation, which is then solved for the index. The value obtained is consistent with the given production rate and average casing pressure. (See cases 3-5 of Table 13-V.) In some cases, the productivity index may be known more accurately than the static bottomhole pressure, at some depth other than the tubing bottom. In this event, the index and the static may be estimated for the desired depth with sufficient accuracy for most purposes. The given value of the productivity index is introduced into the productivity-index equation, solving for the static. The resulting value approximates the static at the reference depth for the given index. The resulting value of the static is corrected to the tubing depth, applying the static gradient. The new value is then used in the productivity-index equation to estimate the index at the tubing depth. (See cases 6-8 of Table 13-V.) These calculations give a pair of values consistent with the operating conditions and reasonably accurate. This method always gives a pair of values consistent with the operating condi-
520
tions. The values will be exact, provided that the static at the tubing depth is known, or the static at another depth and the static gradient are known. Table 13-V shows that reasonable estimations of the static gradient yield sufficiently accurate values of the index. When the index is known at some depth other than the tubing depth, Table 13-V shows that sufficiently accurate values of the static and the index may be computed by using the known or estimated static gradient. Changes in operating line with changes in static and index computed from operating conditions For a given pair of values for the production rate and the average flowing pressure, all main operating lines plotted on a nomograph will have a common point fixed as follows: the value on the production-rate axis is the given production rate; the corresponding value on the size-load axis is given by the expression for intercept 1 on that axis, with the given average flowing bottom-tubing pressure being substituted for the static bottom-tubing pressure. Derivation of the common point is given in later paragraphs. Changes in the direction of the operating lines passing through that fixed point are apparent from a consideration of eq. 13-22 of intercept 1 for 2i-in. plunger lift: When q = 0,
L, =
P, - (1 + 0.027D/1000)(0.974Ptmin+ 1.1250/1000
(1 + 0.0270/1000)(11.11D/1000
+ 113.2)
+ 1.545)
(13-22)
The values of D and Ptm"'are fixed for given operating conditions, and so L, increases with increasing values of P,,. This causes the operating line to tend toward the vertical. A decrease in P,, causes the operating line to tend toward the horizontal. The equation for the main operating line for the 2i-in. plunger has been derived in the following form (eq. 13-21): )(0.974Pyin+ 1.125-1000 +113.2)
(13-21) This equation is applied to the same production rate for a well, but to different pairs of values taken for the static pressure and the productivity index. Subscript 1 is used to denote one set of values, and subscript 2 to denote a second set of values. The
521
operating line for condition 2 is subtracted from the operating line for condition 1. At the common point of the lines, this yields the following size load: J 2 ( P W S ) 2 -J1(PwJ1
''
=
(1 + 0.027-
-
1000
I(1.545 + 11.11-)(J2 D 1000
0.974PF'" + 1.125-
1000
-J1)
+ 113.2
D
1.545 + 11.111000
(13-29)
The productivity-index equation is then applied to the same production rate and flowing bottom-tubing pressure for a well, but to different pairs of values taken for the static pressure and the productivity index. The fixed value of average flowing bottom-tubing pressure is denoted by p;,. The productivity-index equation for condition 1 is divided by the productivity-index equation for condition 2 to give the following relation: (I 3-30)
Substituting the value of J1from eq. 13-30 into eq. 13-29 gives:
P;, L, =
i+ i+ 1
+ 1.125 + 113.2) 1000 0.027I(1.545 + 11.111000 1000
0.027-
1
1000
](0.974PFin
(13-31)
Equation 13-31 is the value of intercept 1 which would be obtained by using F;, instead of P,,. Of course, the common point of the two lines lies at the production rate given for the well and is fixed by this given value and eq. 13-31. Actual field example of changes in predictions from operating line One operator has supplied both static bottomhole pressures and productivity indexes for nine wells being produced by plunger lift. This provided an excellent opportunity for obtaining actual well data to illustrate the changes that occur in predictions from an operating line, when different values are used for the static pressure and the productivity index. The well data are listed in Table 13-VI. The table includes data of all wells for which the operator supplied both static bottomhole pressures and productivity indexes. In all calculations with the well data, operating conditions were fixed at those
u? N
TABLE 13-VI
Well data for 2;-in. plungers a Well
Gross liquid
-
(bbl/day)
A B C
90 24 28 64 31 40 .53 42 27
Water cut
14 8 2 4 4 4 6 5 8
Gas/ liquid ratio
Casing pressures
Tubing pressures,
Cycles Per
(cu ft/bhl)
max.
min.
max.
min.
9530 7500 21,000 6500 13,500 7500 4500 8500 6500
360 340 380 580 340 310 590
340 325 345 560 300 290 510 360 380
345 330 345 560 300 330 540 365 350
135 200 170 245 110 145 140 115 90
Plunger footpiece depth
-
(ft)
96 29 79 85 30 58 26 30 52
7417 5873 7799 7665 6540 6687 6250 8535 6051
Tubing depth
7447 5873 7822 7695 6570 6717 6280 8565 6081
Producing interval depth from
to
7450 7431 7881 7698 7324 7060 6329 8655 6900
8152 8209 8799 8320 7450 8150 7349 9761 8236
D E F G H 400 I 400 a The following conditions were reported for all the wells: (1) Gas/liquid ratio involved only formation gas because no gas was circulated in any of the wells. (2) Gas gravity was 0.75 referred to air, and oil gravity was 30-31OAPI. (3) Tubing size was 2; in., and casing size was 7 in. ~
TABLE 13-VII Comparison of actual production rate with maximum estimated from given static pressure and given productivity index Well
'Tubing depth
Min. tubing pressure
Given static pressure
(ft)
(PSid
(Psi&
7447 5873 7822 7695 6570 6717 6280 8565 6081
135 200 170 245 110 145 140 115
1000 1100 1160 1500 1100 1055 1590 2070 1040
90
Given prod. index
0.217 0.11 0.05 0.50 0.05 0.10 0.07 0.03 0.15
Intercept 1
Intercept2
(bbl/cy)
(bbl/day)
6.85 9.50 7.60 10.00 9.10 8.30 15.35
150 88 40 525 41 74 92 53 118
15.00
9.90
Intersection of oper. lines (bbl/cy)
(bbl/day)
0.61 0.59 0.30 1.11
393 193 100 1170
~
0.74
Intercept 2A
-
267
Est. max. prod. rate (bbl/day)
70 75 37 113 40 68 86 51 86
a
Initial prod. rate
Initial
(bbl/day)
(%)
90 24 28 64 31
128.7 32.0 75.6 56.6 77.5 58.8 61.6 82.4 31.4 Average 61.2
40 53 42 27
Est. rnax.
'
a With the given static and given index, wells A, B, C, D, and I produced more gas than was required, and excess gas was not removed; therefore, the estimatcd maximum production rate was obtained by plotting a supplementary operating line. With the given static and given index, wells E, F, G, and H produced less gas than was required for maximum production rate; therefore, no supplementary operating line was plotted for any of these wells.
w W N
TABLE 13-VIII Comparison of actual production rate with maximum estimated from given static pressure and computed productivity index Well
A B C D E F G H I
'
Tubing depth
Min. tubing pressure,
Given static pressure
Average flowing pressure
(ft)
(psig)
(PSk)
(PSk)
7447 5873 7821 7695 6570 6717 6280 8565 6081
135 200 170 245 110 145 140 115 90
1000 1100 1160 1500 1050 1055 1590 2070 1040
420 385 437 688 376 353 646 468 455
Calc. prod. index
0.155 0.034 0.039 0.079 0.043 0.057 0.056 0.026 0.046
Intercept 1
Intercept 2
(bbb'cy)
(bbl/day)
7.25 9.60 6.65 10.10 9.50 10.55 15.35 14.80 9.90
107 25 32 84 35 41 70 46 37
Intersection of oper. lines (bbl/cy)
Intercept 2A
(bbl/day)
a
Est. max. prod. rate (bbl/day)
Initial prod. rate
Initial
(bbl/day)
(%)
68 24 30 77 34 40 68 44 35
90 24 28 64 31 40 53 42 27
132.2 100.0 93.3 83.2 91.2 100.0 77.9 95.5 77.1 Average 94.5
Est. max.
With the given static and computed index, well A produced more gas than was required, and the excess gas was not removed; therefore, the estimated maximum production rate was obtained by plotting a supplementary operating line. With the given static and computed index, each of the other wells produced less gas than was required for maximum production rate; therefore, no other supplementary operating lines were plotted. a
TABLE 13-IX Comparison of actual production rate with maximum estimated from given productivity index and computed static pressure Well
A B C D E F G
H I
Tubing depth
7447 5873 7821 7695 6570 6717 6280 8565 6081
Min. tubing pressure
135 200 170 245 110 145 140 115
90
Given prod. index
0.217 0.11 0.05 0.50 0.05
0.10 0.07 0.03 0.15
Average flowing pressure,
420 385 437 688 376 353 646
468 455
Calc. prod. index
834 603 997 816 996 753 1403 1868 635
Intercept 1
Intercept 2
(bbl/cy)
(bbl/day)
5.15 3.10 6.10 3.60 8.20 4.85 13.15 13.30 4.80
126 24 32 87 36 83 76 47 65
Intersection of oper. lines (bbl/cy)
(bbl/day)
0.61
355
0.30 1.11
0.68 -
Intercept 2A
-
92 829 -
149
~
ESt. max. prod. rate (bbl/day)
67 23 31 90 35 70 72 46 58
a
Initial
Initial
prod. rate (bbl/day)
(%)
90
134.3 104.3 90.3 71.1 88.6 57.2 73.6 91.3 46.5 Average 84.1
24 28 64 31 40 53 42 27
a With the given index and computed static, wells A, C, D, and F produced more gas than was required, and excess gas was not removed; therefore, the estimated maximum production rate was obtained by plotting a supplementary operating line. With the given index and computed static, wells B, E, G, H, and I produced less gas than was required for maximum production rate; therefore, no supplementary operating line was plotted for any of these wells.
wl wl N
TABLE 13-X Comparison of actual production rate with maximum estimated from measured pressure run and computed productivity index Well
Tubing depth
Min. tubing pressure
Meas. static pressure
Average flowing pressure
(ft)
(Psig)
(PSk)
(Psig)
7447 5873 7821 7695 6570 6717 6280 8565 6081
135 200 170 245 110 145 140 115 90
899 671 1030 1480 856 1014 1404 1919 605
420 385 437 688 376 353 646 468 455
Calc. prod. index
0.188 0.084 0.047 0.081 0.065 0.061 0.070 0.029 0.180
Intercept 1
Intercept 2
(bbl/cy)
(bbl/day)
5.80 3.95 6.35 9.90 6.75 7.80 13.15 13.65 2.95
112 26 32 84 38 42 76 48 40
Intersection of oper. lines (bbl/cy)
Intercept 2A
(bbl/day)
Est. max. prod. rate (bbl/day)
65 25 30 76 36 41 73 42 37
a
Initial prod. rate
Initial Est. max.
(bbl/day)
(56)
90 24 28 64 31
138.4 96.0 93.3 84.1 86.1 97.6 72.6 100.0 73.0 Average 93.5
40
53 42 27
~~
With the static obtained from a measured pressure run and the computed index, well A produced more gas than was required, and excess gas was not removed; therefore, the estimated maximum production rate was obtained by plotting a supplementary operating line. With the static obtained from a measured pressure run and the computed index, each of the other wells produced less gas than was required for maximum production rate; therefore, no other supplementary operating lines were plotted. a
527
actually measured immediately after plunger lift was installed. These initial conditions were used to compare the initial production rate with the estimated maximum production rate obtained from the operating line. This one point on the operating line is quite suitable as a standard of comparison for wells such as these, with gas/ liquid ratios sufficiently high to operate near maximum production rates. In Table 13-VII, the given static pressure and given productivity index for each well were taken to be the values at the bottom of the tubing and were used in plotting the operating line. In Table 13-VIII, the given static pressure was taken to be the value at the bottom of the tubing to compute the productivity index used in plotting the operating line. In Table 13-IX, the given index was taken to be the value at the bottom of the tubing to compute the static pressure used in plotting the operating line. In Table 13-X, the given static pressure was corrected from the middle of the perforations to the bottom of the tubing by data obtained in a pressure-bomb run. The productivity index was then computed for the bottom of the tubing before plotting the operating line. Comparisons in the tables were based on maximum production rates estimated by plotting operating lines. This method of comparison was used because the producing gas/liquid ratios were high and the operator hoped to produce oil from each well at a rate near the maximum obtainable by plunger lift. However, tubing depths were high (average of 817 ft) above the top of the perforations. Comparisons also might be based on production rates at any given size load. Nevertheless the use of estimated maximum production rate is quite appropriate as a basis for a well with a high gas/liquid ratio or where outside gas is available. It is apparent from the tables that the given static pressure and given productivity index did not permit reasonably accurate predictions of production rate. The actual initial production rate averaged only 67.2% of the estimated maximum production rate. On the other hand, the tables show that computing the productivity index from the given static pressure, or vice versa, yielded reasonably accurate predictions. For the actual case in which the productivity index was based on a static pressure obtained from a pressure-bomb run, the initial production rate averaged 93.5% of the estimated maximum production rate. The average was only 1.0% higher, or 94.5%, when the index was computed directly from the static. The average was only 9.4% lower, or 84.1%,when the static was computed directly from the index. Even more accurate estimations would have been obtained if the static bottomhole pressure had been corrected to the tubing depth. As was stated concerning Table 13-V, the computed index would be exact in case the reference depth is given for the static and the static gradient is known. It was also shown in Table 13-V that quite accurate values of the static and the index may be computed from the index at some reference depth, using the known or estimated gradient. NOMENCLATURE
D
-depth or distance from surface to point at which gas enters tubing, feet.
528
-volume ratio of gross gas to gross liquid, both measured at 14.7 psia and 60" F, standard cubic feet per barrel. G -volume of gross gas per cycle, standard cubic feet per cycle. J -productivity index, barrels per day per pound per square inch. -size load of gross liquid, barrels per cycle. L C PcmaX -maximum casing pressure (at the surface), pounds per square inch gauge. pcmln -minimum casing pressure (at the surface), pounds per square inch gauge. Ptmln -minimum tubing pressure (at the surface), pounds per square inch gauge. -flowing bottomhole pressure (at depth D),pounds per square PWf inch gauge. -average flowing bottomhole pressure (at depth D), pounds per PWf square inch gauge. -static bottomhole pressure (at depth D), pounds per square inch P W S gauge. -gross production rate, barrels of liquid per day. 4 4 Inax -maximum gross production rate, barrels of liquid per day. -minimum time required for plunger to complete a cycle, minutes t y per cycle. ( G / L ) / ( D/lOOO)-gas/liquid ratio gradient, cubic feet per barrel/(feet/1000). Pcmax - P,mn:net operating pressure, pounds per square inch. P,"" - Pcfin:pressure buildup, pounds per square inch. G/L
ACKNOWLEDGEMENTS
The first part of this chapter includes large portions of the material contained in a series of five articles by Beeson et al. (1958). This material has been reproduced by consent of the Petroleum Engineer Publishing Company, Dallas, Texas. The authors sincerely appreciate the fact that the excellent distribution indicated in Tables 13-1, 13-11, and 13-111 is due to the many engineers and operators who so generously collected and submitted the data. The authors sincerely appreciate release of the data for publication of Table 13-VI. The help extended by E. Beauregard and Don Hollis of Ferguson Beauregard Inc. is also greatly appreciated.
SAMPLE PROBLEMS AND QUESTIONS
(1) Show by a sequence of five illustrations the operation of a plunger lift. (2) Describe types of wells suitable for plunger lift.
529
(3) List the main items of interest in the operation of the gas-lift plunger for a given well. (4) Is there any noticeable effect on the items in (3) due to changes in tailpipe, water cut or oil gravity, and how was this determined? (5) In using a nomograph, what is the datum plane that should be used for the static bottomhole pressure and the productivity index? (6) If the static bottomhole pressure and the productivity index are given for the midpoint of the perforations, what information would be needed to compute the values at the desired datum plane? (a) For a well that is being pumped; (b) for a well that is being produced by continuous gas lift. (7) Indicate the computations that would need to be made with the information listed in (6). (8) What is the distinguishing characteristics of a well for which a Supplementary Operating Line would be plotted on a nomograph? (9) For a 2i-in. plunger with maximum production rate of 110 bbl/day determine: (a) cycle frequency, cy/day; (b) cycle time, min/cy; (c) size load at maximum production rate, bbl/cy; and (d) required daily gas volume. Given: depth = 9000 ft; min. tubing pressure = 100 psi. Max. casing pressure is not given. (10) Given: plunger size-2 in.; tubing depth-10,000 ft; intercept 1-3.66 bbl/cy; intercept 2-100 bbl/day. Determine: (a) maximum production rate-bbl/day; (b) size load at maximum production rate-bbl/cy; (c) cycle frequency; and (d) cycle time-min/cy. (11)Construct nomograph for a 2-in. plunger using the data given in this chapter. REFERENCES Anderson, R.L. and Bancroft, T.A., 1952. Statistical Theory in Research. McGraw-Hill, New York, N.Y. Babson, E.C., 1939. Range of application of gas lift methods, API Drilling Prod. Practice, American Petroleum Institute, New York, N.Y., p. 266. Beauregard, E. and Ferguson, P.L., 1981. Introduction to Plunger Lift: Applications, Advantages and Limitations. Presented at the Southwestern Petroleum Short Course, Dept. Pet. Eng., Texas Tech. Univ., Lubbock, Tex., Apr. 23-24. Beeson, C.M., Knox, D.G. and Stoddard, J.H., 1955. Plunger lift correlation equations and nomographs. Paper 501-G presented at AIME Pet. Branch Meet., New Orleans, La., Oct. 2-5, 1955. Beeson, C.M., Knox, D.G. and Stoddard, J.H., 1958. The plunger lift method of oil production. Pet. Eng., 30(6): B96-102; 30(7): B68-71; 30(9): B58-61; 30(10): B76-77; 30(11): B106-108. Davies, O.L., 1949. Statistical Methoak in Research. Oliver and Boyd, London.
530 Ezekial, M., 1941. Methods of Correlution Analysis. Wiley, New York, N.Y. Well? Presented at the Ferguson, P.L. and Beauregard, E.. 1983. Will Plunger Lift Work in Southwestern Petroleum Short Course, Dept. Pet. Eng., Texas Tech. Univ., Lubbock, Tex., Apr. 27-28. Horton, W.D., 1959. Esperience with Vurious Gus Lift lnstullutions. Los Angeles Busin. Rep. Pet. Eng. Dept.. Univ. South. Calif., Los Angeles, Calif. (unpublished). Kempthorne, O., 1952. The Design und Ana/ysis of Experiments. Wiley, New York. N.Y. Lea, J.F., 1981. D,vnumic unu(vsis of plunger lift operutions. 56th Annu. Fall Tech. Conf. and Exhibition, SOC.Pet. Eng. AIME. San Antonio, Tex., Oct. 5-7, 1981, SPE 10253, 11 pp. Lebeaux, J.M. and Sudduth, L.F., 1955. Theoretical and practical aspects of free piston operations. J . Per. Technol., 7(9): 33. Lloyd. F.T., 1959. Artificial lifting of deep wells in California, A P I Drilling Prod. Practice, American Petroleum Institute, New York, N.Y., p. 359. McMurry, E.D., 1953. Use of the automatic free piston in oil well production problems. Truns. A I M E , 198: 165-170. Shenvood, T.K. and Reed, C.E.. 1939. Applied Muthemutics in Chemicul Engineering. McGraw-Hill. New York, N.Y.
531
Chapter 14
SUCKER-ROD PUMPING DAWOOD MOMENI, GEORGE V. CHILINGARIAN, W.B. HATCHER and AXELSON
INTRODUCTION
The oldest and most widely used type of artificial lift of oil wells is the sucker-rod pumping. Most of the stripper wells (i.e., producing less than 10 bbl/day) use sucker-rod pumps. These stripper wells constitute about 74% of all oil wells. About 28% of the remaining 26% are flowing wells, 20% are lifted with sucker-rod pumps, and 52% are lifted by gas lift, electric submersible pumps, and hydraulic pumps. Usually, it is possible to lift 1,000 bbl/day from a depth of about 7000 ft, and 200 bbl/day from a depth of about 14,000 ft. Presence of H,S changes depth values to 4000 f t and 10,000 ft, respectively (Neely et al., 1981). Unfortunately sucker-rod pumping (1) cannot be used in crooked holes, (2) have limited ability to lift sand, and (3) is inefficient in the presence of scale and paraffin accumulations. In addition, the sucker-rod pump operates inefficiently and tends to gas lock if the annulus is not used efficiently or if the gas-liquid separation capacity of the tubing-casing annulus is too low. Inasmuch as various parts of a sucker-rod pumping system (string, pump, and unanchored tubing) are subjected to fatigue, this system (see Fig. 14-l.a,b) must be more effectively protected against corrosion than other lift systems. The possibility of polished rod stuffing box leaks can be minimized using proper design and operating procedures (Neely et al., 1981).
SUCKER-ROD PUMPING UNIT
A schematic diagram of a walking-beam type of sucker-rod pump unit is presented in Fig. 14-1.a. The power of the electric motor ( 1 ) is transferred by v-belts to,a gear reducer ( 2 ) , which reduces the rpm of electric motor to about 3-25 rpm. The number of double strokes per minute (spm) of the sucker rod is equal to the rpm value. If I , = I,, the length of the stroke of the polished rod ( 3 ) is equal to 2r ( r = length of the crank). Crank is connected to the walking beam ( 6 ) by the pitman
‘
Axelson, Inc. gave permission to reproduce several parts of their excellent “Pump and Rod Engineering Manual”. The help extended by s. (Chip) Cipparuolo is indeed greatly appreciated by the authors.
532
Fig. 14-1.a. Sucker-rod pumping system (European approach). (After Szilas, 1975, fig. 4.1-1, p. 259; courtesy of Elsevier Science Publishers.)
(connecting rod), having length 1. Polished.rod (3) is being moved by the walking beam ( 6 ) and the horsehead ( 7 ) . Walking beam ( 6 ) is supported by the Samson post (a trestle) ( 5 ) . Variation of polished-rod load over the pumping cycle is balanced by a crank counterweight (8) and the beam counterweight ( 9 ) . Polished rod ( 3 ) , which is a specially made and machined top unit of the rod string, is hanging from a carrier ( 4 ) . Plunger ( 1 1 ) is moving up and down by the rod string inside a pump barrel ( l o ) ,which is attached to the installed tubing shoe. During the downstroke, the travelling valve ( 1 2 ) is open and the standing valve ( 1 3 ) is closed. Thus, the plunger sinks in the fluid filling the barrel. On the upstroke, the travelling valve ( 1 2 ) is closed and the plunger lifts the fluid filling the annular space between the tubing and the rod. During this time, the standing valve ( 1 3 ) is open to allow the fluid to enter the barrel through the filter ( 1 4 ) .(See also Figs. 14-1.b and 14-2.) General considerations
All beam-type pumping unit geometries fall into two distinct classes: (1) Class I lever system, having rearmounted speed (gear) reducer, with the fulcrum at mid-beam (e.g., conventional unit). (2)Class I11 lever system, having a push-up geometry with front-mounted speed reducer (e.g., air balanced and Lufkin Mark I1 units), in which
533
WNj WJRIffi
WLL Rutr
Fig. 14-1.b. Sucker-.rod pumping system. (After API, 1983, fig. 1.1, p. 1.)
PE s Fig. 14-2. Schematic diagram showing some production and well completion equiprnents and a sucker-rod pumping unit. (Courtesy of Trico Industries, Inc.)
c
L
d
N P
v; m
e
Fig. 14-3. Pumping units nomenclaturc. (After Lufkin, 1983, figs. 33. 34. and 35, p. 27: courtesy of Lufkin Industries. Inc.) 535
Air Balanced Pumping Unit
Fig. 14-3 continued.
Mark
II
Pumping Unit
537
the fulcrum is located at the rear of the beam (Byrd, 1983). Figure 14-3 illustrates schematically these different pumping unit types. In any sucker-rod pumping installation design, the behavior of all elements of the system should be considered very carefully. These elements should not be treated
(2) RWA RSA
(1) RHA: (2) RWA: RSA
(3) R H B (4) RWB: RSB: ( 5 ) RHT: (6) RWT: RST: (7) TH: (8)TP:
(3) RHB
(4)
RWB
RSB
(5) RHT
(6) RWT
RST
(7)
TH
Rod, Stationary Heavy Wall Barrel, Top Anchor Pump Rod, Stationary Thin Wall Barrel, Top Anchor Pump Rod, Stationary Thin Wall Barrel, Top Anchor, Soft-Packed Plunger Pump Rod, Stationary Heavy Wall Barrel, Bottom Anchor Pump Rod, Stationary Thin Wall Barrel, Bottom Anchor P u m Rod,Stationary Thin Wall Barrel, Bottom Anchor, Softsacked Plunger Pump Rod, Traveling Heavy Wall Barrel, Bottom Anchor Pump Rod, Traveling Thin Wall Barrel, Bottom Anchor Pump Rod, Traveling Thin Wall Barrel, Bottom Anchor, Soft-Packed Plunger Pump Tubing, Heavy Wall Barrel Pump Tubing, Heavy Wall Barrel Soft-Packed Plunger Pump
Fig. 14-4. API pump classification. (From API, 1979, R P 11AR: courtesy of American Petroleum Institute.)
538
individually but rather as a series of elements working in harmony with the other components of the system. According to API(1983) RP 11L standards, the minimum amount of information which must be known, or assumed, in order to Letter Designation Metal Plunger Pumps 7-
Type of Pump Rod Pumps Stationary Barrel, Top Anchor Stationary Barrel, Bottom Anchor Traveling Barrel, Bottom Anchor Tubing Pumps
Soft-packed Plunger Pumps 7
-7
Heavy-Wall Barrel
Thin-Wall Barrel
Heavy-Wall Barrel
Thin-Wail Barrel
RHA RHB RHT TH
RWA RWB RWT
....... . ....., , . .. ....,,
RSA RSB RST .. . .. . ..
........
TP
6.7.2 Complete pump designations include: (1) nominal tubing size, ( 2 ) basic bore diameter, (3) type of pump, including type of barrel and location and type of seating assembly, ( 4 ) barrel length, (5) plunger length, and ( 6 ) total length of extensions when used, as follows:
r -1! x x
--Total
length of extensions, whole feet. Nominal plunger length, feet. Barrel length, feet.
Type seating usembly; C -Cup type M - Mechanical type Location of aeating asaembly; A - Top B - Bottom T - Bottom, traveling barrel
F o r metal plunger pumps
S -Thin-wall
-Type
For soft-packed plunger pumps
pump; R-Rod (91.8 mm) (98.1 m m ) (44.5 mm) ( 4 5 2 mm) (50.8 mm) (57.2 mm) (69.5 m m ) (69.9 mm)
-Tubing
-
sire; 16 1.900 in. OD. 20 -%in. OD. 26 2% in. OD. 30 3 % in. OD.
(48.9 m m ) (60.9 mm) (79.0 mm) (88.9 mm)
6.7.3 In addition t o the pump designation described in Par. 6.7.2., i t is necessary for the purchaser to provide the following information: a. Barrel material b. Plunger material C. Plunger clearance (fit) d. Valv;? material ' ' e. Length of each extension NOTE: Metallic Materials for Subsurface Sucker Rod Pumps for Hydrogen Sulfide Environments are listed in NACE Std MR-01-76.
Example: A 1% in. (91.8mm) bore rod type pump with a 10 ft. (9.048 m ) heavy wall barrel and 2 ft. (0.610 m ) of extensions, a 4 it. (1.219 m ) plunger, and a bottom cup type seating assembly for o era tion in 2% in. (60.Smm) tubing, would be &sip: nated a s follows: 20-126 RHBC io-4-2
I
Fig. 14-5. API pump designation. (From API, 1977, RP l l A R , p. 17; courtesy of American Petroleum Institute.)
539
determine even approximate loads and pump displacements must include: (1) fluid level (net lift), ft; (2) pump depth, ft; (3) pumping speed, strokes per minute; (4) length of surface stroke, in.; (5) pump plunger diameter, in.; (6) specific gravity of the fluid; (7) nominal tubing diameter and whether it is anchored or unanchored; (8) sucker-rod size and design; and (9) unit geometry. With the above data, the engineer can calculate: (1) plunger stroke, in.; ( 2 ) pump displacement, bbl/D; (3) peak polished rod load, lb; (4) minimum polished rod load, lb; (5) peak (crank) torque, in.-lb or ft-lb (when the unit’s torque factor schedule is known); (6) polished rod horsepower; and (7) counterweight required, lb. It should be noted that the API RP 11L calculations assume that there is 100% pump fillage. Also, difficulties are encountered when trying to use these calculations on shallow wells. As stated by Brown et al. (1980) three steps are generally required in designing an installation: (1) preliminary selection of components for the installation must be made; ( 2 ) operating characteristics of the preliminary selection are calculated using the basic formulas, tables and figures presented in this chapter; and (3) the calculated pump displacement and loads are compared with the volumes, load ratings, stresses, and other limitation of the preliminary selection. Subsurface pumps Once a decision has been made to artificially produce an oil well with a sucker-rod-pump system, the size of the pump bore is the first element which must be considered. Inasmuch as the primary reason for the installation is to produce fluid, the quantity of fluid desired is the first controlling factor. The next step is the selection of the pump type. It is narrowed down from eight API types to about two by the time one has selected the bore size. In some cases, however, one may have to consider some of the “special” pumps offered by the manufacturers. The American Petroleum Institute (1983) has adopted a classification system for subsurface pumps, as shown in Fig. 14-4. A complete pump designation is given in Fig. 14-5. Complete description of the API pumps is presented in the next section.
EVALUATION AND SELECTION OF PUMPS
In selecting the pump for use in a particular well, it is necessary first to consider the amount of fluid to be produced. This determines the bore size of the pump. The next consideration is the type of pump to be used. SelectiQn of pump bore size With the required quantity of fluid known, selection of a pump bore is the first step in designing a pump. Charts have been prepared where combinations of net plunger ttavel and strokes per minute have been applied to the most popular bore sizes to obtain the daily production. This production figure is based on 100%
540
r
STROKES
PER
MINUTE
Fig. 14-6. Relationship between production (bbl/D) and strokes per minute (SPM) for a 1&-in. bore and various net plunger travels. (After Axelson, 1982. p. 26.) V.E. = volumetric efficiency.
volumetric efficiency. Thus, it is necessary to determine at what efficiency the pump will operate when it is finally installed. The measured production/computed production ratio is fairly constant for many of the wells in a certain area. This ratio, which is termed "field efficiency", can be applied to other similar wells of that area. If efficiency ratio is not available, it is common practice to use a figure of 80%. As an example, required daily production of 120 bbl was selected. To produce this amount at 80% efficiency, the pump must have a capacity of 150 bbl/D at 100% efficiency (i.e., 120/0.8 or 120 X 1.25). The production charts (Figs. 14-6 through 14-13) are then consulted to determine how the different plunger bores must be
541
STROKES
PER MINUTE
Fig. 14-7. Relationslup between production (bbl/D) and strokes per minute (S;PM) for a 1:-in. bore and various net plunger travels. (After Axelson, 1982, p. 27.)
operated to produce the necessary amount of fluid. The 150 bbl/D point is found in the “production in barrels per day” column on the left-hand side of the chart and a horizontal line extended from that point to the right. This line should intersect one or more radiating lines that represent different lengths of net plunger travel. When vertical lines are dropped downward from these points of intersection, they will fall across the scale which represents the number of cycles or strokes per minute. Thus, for any average production requirement, there may be several pump bores capable of delivery. Each one of those bores may be operated with numerous combinations of net plunger travel and strokes per minute.
542
Y
<
8 0
3 ?i v)
fBl 5
1 STROKES PER MINUTE
Fig 14-8 Relationship between production (bbl/D) and strokes per minute (SPM) for a 1:-in bore and various net plunger travels (After Axelson, 1982, p 28 )
Upon scanning the different bore production charts, Figs. 14-8 and 14-9 covering the It-in. and lf-in. bores seem to offer the most likely applications. The l$-in. bore will produce the desired quantity with 50-in. NPT (net plunger travel) at 11: SPM (strokes per minute); 40-in. NPT at 14$ SPM; or 30-in. NPT at 193 SPM. The If-in. bore using the same NPTs will produce the needed amount at the slower speeds of 8$, 104, or 14 SPM, respectively. The problem may be approached in different ways after the production line has been established on a chart. For instance, if a la-in. bore is to be used and a pumping unit and prime mover were already coupled together to run at 12: SPM, approximately 34-111. NPT would be needed.
543
Fig. 14-9. Relationship between production (hbl/D) and strokes per minute (SPM) for a 1 :-in. bore and various net plunger travels. (After Axelson, 1982, p. 29.)
In some instances, where large bores are involved, the required production may not be listed on the chart. In such cases, the required production must be divided by some factor so that the result is a number of barrels per day on the chart. Using the method outlined earlier, locate the NPT and SPM intersection. Then, either the NPT or SPM must be multiplied by the same factor used to reduce the original production figure, which will give the combination needed to produce the original amount. The simple formula used to construct these charts is presented below. The pump bore may be considered as the diameter and the net plunger travel as the length of a
544
Fig. 14-10. Relationship between production (bbl/D) and strokes per minute (SPM) for a 2-in. bore and various net plunger travels. (After Axelson, 1982, p. 30.)
cylinder-shaped bucket. This bucket is filled and emptied a certain number of times per day. This enables determination of total production Q: Q =A,
X
NPT
X
S P M ( 6 0 X 24/231
X
42)
= O.1484AP X
NPT
X
SPM
(14-1)
where Q = total production, bbl/D; A , = area of plunger, in.’; N P T = net plunger travel, in.; and S P M = strokes per minute. (Conversions: 60 min/hr, 24 hr/D, 231 cu in./gal (U.S.), and 42 gal/bbl.) Inasmuch as for each distinct bore the plunger area remains unchanged, the formula can be further simplified by multiplying the different plunger areas by the
545
STROKES
PER
MINUTE
Fig. 14-11. Relationship between production (bbl/D) and strokes per minute (SPM) for a 2:-1n. bore and various net plunger travels. (After Axelson, 1982, p 31.)
time/volume factor of 0.1484, to obtain a list of production constants. This constant is frequently referred to as d (see Table 14-1). Thus: Q = C x N P T x SPM
(14-2)
If a chart is not available but Q and the pump constant are known, a product of the N P T and S P M can be found and a combination of the two determined for application. For example, if a chart for 1:-in. bore was not available and the same Q of 150 bbl/D was required, by dividing Q by the C value for the I f - in . bore (0.2622) the result is 572. This figure is a product of N P T X S P M needed to
546
STROKES
PER M I W E
Fig. 14-12. Relationship between production (bbl/D) and strokes per minute (SPM) for a 21-in. bore and various net plunger travels. (After Axelson, 1982, p. 32.)
produce 150 bbl/D. By assigning a speed of 10 SPM, the N P T must be approximately equal to 57 in. Referring to Fig. 14-8 again, the N P T and SPM are found to agree with the 150 bbl/D horizontal line. It should be noted that the SPM value can be adjusted over an entire range, whereas the stroke length, which is directly related to N P T , can only be set at 2 or 3 specific values, depending upon the pumping unit. After concluding that either the lf-in. or the l+-in. bore (and their operating combinations) are best suited, Table 14-11 may be consulted as a further step in deciding which one of the two will be best for the other conditions anticipated in this well. Ths table shows the minimum size tubing that can be used with certain
541
STROKES
PER MINUTE
Fig. 14-13. Relationship between production (bbl/D) and strokes per minute (SPM) for a 3 l - h bore and various net plunger travels. (After Axelson, 1982, p. 33.)
bores of the different types of pumps of standard design. In some instances, the tubing size is governed by field practice if all of the hoists are equipped with the same size tongs. Letter designations for several kinds and types of pumps are included under the pump type heading. The first letter of this designation is either an “R” or a “T’. The “ R ’ indicates that the complete pump assembly is installed and retrieved by manipulating the sucker-rods to which it is attached. The tubing in which the pump is used is generally equipped with a means of anchoring the stationary part of the pump at the required depth, while the movable or travelling portion is free to
548 TABLE 14-1 Pump production constants ( C = 0.1484 A P G ) . Production bbl/D 1982, p. 34.) Plunger diameter (in.)
Gross plunger area (in.*)
=
NPT
X
0.3067
0.0455
(0.750) (0.875) l(1.000) 1 (1.0625) 1; (1.125)
0.4417 0.6013 0.7854 0.8866
0.0,656 0.0892 0.1166 0.1316
0.9940
0.1475
1 -0.040’’ (1.210)
1.1499 1.2272
0.1706 0.1821
1 (1.500)
1.7671
0.2622
1f (1.625)
2.0739
0.3078
1: -0.040’’ (1.710)
2.2966
0.3409
1: (1.750)
2.4053
0.3569
1s (1.7813) 2 (2.000) 2; (2.125) 2:(2.250) 2: (2.500) 2; (2.750) 3 (3.000) 3: (3.250) 3; (3.500) 3; (3.750) 4: (4.750)
2.4900 3.1416 3.5466
0.3695 0.4662 0.5263
3.9761
0.5901
4.9087
0.7285 0.8814 1.0490 1.2310
i
5.9396 7.0686 8.2958 9.6211 11.045 17.721
X
C. (After Axelson,
Pump constant, C
5 (0.625)
a 1 a (1.250)
SPM
1.4278 1.6390 2.6297
reciprocate with the rod string. This anchoring mechanism, in conjunction with the holddown assembly of the pump, also serves as a fluid seal. Such pump assemblies are known as rod type pumps. The “T” indicates that the pump assembly is a tubing type pump. In the latter case, installing or retrieving the complete assembly requires manipulation of both the sucker rods and tubing. The second letter designation indicates the kind of rod type or tubing type pump, in reference to the barrel or tube part of the assembly. The letter ‘‘€3’’ indicates that the barrel of the pump is a one-piece, thick-walled tube with a precision finish. The letter “L” indicates that the barrel of the pump is an assembly consisting of an outer jacket in which either a one-piece liner or several sectional liners are inserted and held in position in the jacket with clamping collars or bushings. Both the one-piece and sectional liners have precision finishes, with
549
TABLE 14-11 Minimum tubing size requirement for standard types of pumps. (After Axelson, 1980, p. 35.) Pump type
Tubing size (in.) If
1;
2
2;
3
1; 1: BSC
2;
I f BSC If
2
2;
3;
1:
2;
2q
3q
4
Pump bores (in.) RH rod, heavy wall RW rod, thin wall TH tubing, heavy wall
7 -
8
15
that of the sectional type being considerably finer. The “H” and “L” barrel designations are common to both rod and tubing type pumps. Although the letter designations “S” and “ P” are different, these barrel designations for rod and tubing pumps are the same. The barrel tolerances are quite large and are used with fabric cup and/or ring plunger. The letter “‘W’ denotes a thin-wall precision tube. The API has established bore tolerances for these different barrel designations. The “H” and “W” tubes and one-piece liners “L” are allowed: plus two-thousandths of an inch and minus zero thousandths of an inch ( + 0.002 in., - 0.000 in.) over or under the base bore. The “P” and “S” tubes are allowed: plus sixty-two ten-thousandths of an inch and minus twenty-two ten-thousandths of an inch ( + 0.0062 in., - 0.0022 in.) over or under the base bore. The sectional liners “L” are allowed; plus twelve ten-thousandths of an inch and minus two ten-thousandths of an inch ( + 0.0012 in., - 0.0002 in.) over or under the base bore. Besides the choice to be made in the type of pump, “ R ’ or “T”, the barrel selection is made by weightng the advantages and disadvantages of the different kinds. The “H” variety rod pump is heavy and strong but has a smaller bore and costs more than the “ W” variety. It has the same bore, fewer fittings and lower costs than the “L” variety; however, repair costs may be higher and, in comparison to the sectional liner “ L”, cannot furnish the fine bore tolerances and is not available with the cast wearing surfaces. The “H” variety tubing pump comparisons are the same as the rod type, except that there is no contrast in bores. The “L” variety (sectional liner) has greater precision and, frequently, lower repair costs; however, it has more parts, higher initial cost and, in comparison to the “W” variety, a smaller bore. The one-piece liner, with fewer parts, has its greatest advantages in cost of repair. The material selection is smaller and its precision the same as that of the “H” and “W” kinds, whereas its bore is also smaller than the “ W” type.
550
The “S’and “P” varieties are available in the same materials as the “H” and “W” tubes, but because of the bore tolerances are more adapted to mass production. Both of these barrels are of the soft-packed plunger type; the “S” variety being “ thin-walled’’ and the “P” variety being “ heavy-walled”. The chief advantage of this type of barrel is low initial cost. The “W” variety is the most popular of all barrel designs, and is available in all tube materials, with standard bores having precision finish. It is intermediate in initial cost and repair. Being of the “thin-wall” design, it is not adaptable to the loads and standard handling of a tubing (T type) pump. As shown in Table 14-11, the 1;-in. bore is available in the RS and RW type pumps which require a minimum tubing size of 2 in. The 14-in. bore is available in the R H type, but the minimum tubing size required for its use is 24-in.. The If-in. bore is also found in the 2-in. tubing column (Table 14-11), but is available only in the TH or TP types. The l$-in. bore appears in the 24-in. tubing column and is available in the RH type. This is because the 2i-in. tubing has sufficient space to accommodate the barrel wall thickness of those two types. The standard bores, as shown in Table 14-11, are not restricted only to the tubing sizes shown, because it is easy to adapt them to the larger tubing sizes. This has been common practice in the past where large-diameter tubing was installed for flush, flowing production, but relatively small bores were needed to handle the fluid when pumping took place. The effect of proration should be considered at the beginning in determining the daily requirements. In sizing a pump it is also necessary to make allowances for downtime. To oversize the pump after a bore has been selected to provide for downtime increases operating costs. Making provisions for shutdown time can be handled similarly to that of the .pump efficiency. For example, if the allowable well production of 114 bbl/D for each calendar day has been established, the average monthly allowable, using 30: days per month, will be 3480 (3477) bbl/month. An estimate of the number of shut-in days per year and then per month is made. If 18 days per year was an acceptable figure, the monthly average would be li days. Subtracting that amount from the 30: average days per month, would leave 29 days for producing the 3480 bbl. This would require 120 bbl/D which is the quantity used in the example. If downtime does not occur, the extra day and a half can always be used for preventative maintenance. If 2i-in. tubing was used in the example, the selection of the bore would require the use of Figs. 14-8 and 14-9. The choice is dictated by the surface equipment and sucker-rod string needed to obtain the net plunger travel and strokes per minute combination. The 1:-in. and l+-in. bores are available in identical kinds and types for that size piping. Thus, the 2-in. tubing can be used. Then the types of pumps and their specifications must be examined closely, because the decision at this stage will automatically determine the bore size to be used. One should analyze all the basic pump types to evaluate the features of two pumps in the above examples.
551
Selection of pump setting depth
In selecting the size of the pump, the pump setting depth is very critical. The bore size will have to be decreased as the setting depth of pump increases. Figures 14-14 and 14-15 may be consulted for maximum setting depths of various barrel sizes. Recommended safety factors are also shown on the charts. In wells that are known as “sanders”, it might be necessary to set the pump at a level higher than what might otherwise be considered optimum. Selection of pump types
After the bore size of the pump has been established, one can determine the pump type required to lift the fluid to the surface. A few API pump types are described below. 35000
Recommended Safety Fact0 F o r Bottom Hold Down
30000
25000
I :
20000
w
0
15000
10000
5000
30
50 70 90 MAltltRIAL YIELD STRENGTH (Thousmds)
110
130
R W (THIN WALL)
Fig. 14-14. Relationship between depth (ft) and yield strength of material (thousands of psi) for RW (thin-wall) type pump. (After Axelson, 1982, p. 8A.)
552 35000
30000
25000
I
In - 20000 W
0
15000
10000
'
Recommended S a f e t y F a c t o r s : For Bottom Hold Down Pumps Divide Max. S e t . Depth By 2
I
For Top Hold Down Pumps Divide Max. S e t . Depth By 3
so00
EXAMPLE: 2-112'' RW Pump, Bot. Hdn., Allo) S t e ( Max. Set Depth - 20,500' ,Rec. Safety S e t . Deoth - 10.150'
1
30
90 50 MTERIAL Y I E L D STRENGTH (Thousands)
110
I30
Fig. 14-15. Relationship between depth (ft) and yield strength of material (thousands of psi) for RH (heavy-wall) type pump. (After Axelson, 1982, p. 9.)
( I ) Casing pump Casing pumps (Fig. 14-16) include those generally designed for large-production volumes. The well fluid is produced through the casing and no tubing is used because the pump is installed on the sucker-rod string. Most assemblies incorporate pack-off and hold-down mechanisms. When the pump reaches the desired depth in the well, the rods are manipulated in a manner which actuates the anchor pack-off to engage and seal against the casing wall. These anchor pack-off elements can usually be installed at the top or bottom of stationary barrel pumps and can be used also with the travelling barrel pump type. Some of the pack-off assemblies rely on the hydrostatic force of fluid in the casing to compress the pack-off for more effective seal. Unless a fluid release is available in packers of that type, tremendous loads are involved when it is necessary to pull the
553
CASING PUMP
Fig. 14-16. Schematic diagram of casing pump. (Courtesy of Axelson, 1980. p. 9.)
pump. It is essential to know the exact size and weight of the casing when pulling the pump. This information is also needed when specifying a casing pump packer. All standard type pumps can be adapted to casing pump use. Slim-hole tubingless completions are smaller versions of casing pumps. Rod insert pumps are adapted to run with pump anchors. Macaroni tubing instead of rods is used in some installations to overcome some basic disadvantages of casing pumps. Gas can be vented and production is confined in the macaroni tubing. The main advantage of tubingless completions is reduction in tubular investment. The casing pump is not recommended for handling gas, because all fluids move through the pump and gas cannot be vented. Wells suited for the large-volume casing pumps are generally water-drive wells with gas effects being minor. The initial and repair costs of casing pumps are comparatively high. Advantages of casing pumps include: (1) large volumetric capacity, (2) certain cost savings, and (3) use without tubing. Disadvantages of these pumps are: (1) casing is subject to wear by rods, (2) broken rods are harder to fish in casing, (3) hazardous in the presence of sand, scale, or corrosion, (4) expensive reworking, and (5) inefficiency in the presence of gas.
554
*
4
THE Fig. 14-17. Schematic diagram of tubing pump (THE). (Courtesy of Axelson, 1982, p. 10.)
( 2 ) Tubing pump A tubing pump (Fig. 14-17) has a greater capacity than a standard rod pump for the same size tubing. The pump barrel and standing valve seating shoe are installed as part of the tubing string or it may be dropped in the tubing prior to running the rods, which is not recommended. It may be run with the plunger on the rod string if the plunger is equipped with a standing valve puller. In addition, the plunger may be run inside the barrel with the tubing and connected to the rods later by means of an on-or-off attachment. The fluid is produced up the tubing and gas is vented up the casing annulus. Oversize plungers and barrels may be adapted to the bottom of the tubing and connection made to the rods by means of an on-or-off attachment. With the addition of a tubing drain, the rods are disconnected and pulled without the pump, while the pump is retrieved by pulling the dry tubing. Such an installation eliminates many of the disadvantages found in casing pump operations. Due to the limitations imposed by the strength and stretch of the rod string, the tubing pumps are usually used at shallow to medium depth and where relatively
555
RWT
RHT Fig. 14-18. Schematic diagram of a rod pump with travelling barrel (RWT, RHT). (Courtesy of Axelson, 1982, p. 11.)
large volumes of fluid are being produced. Tubing pumps are constructed from various materials. Advantages of these pumps include: (1) greater capacity than standard rod pumps, ( 2 ) simplicity and ruggedness for severe service, (3) protection of casing against wear and corrosion, (4) large fluid flow areas, and (5) adaptability for producing viscous fluids. Disadvantages of tubing pumps are: (1) tubing must be pulled to repair pump barrel, ( 2 ) additional installation cost (tubing), (3) gas compression ratios are lower than in an insert pump, and (4) in unconsolidated sands large production may give rise to large sand volumes. (3) Rod pump with travelling barrel In the case of rod pump with travelling barrel (Fig. 14-18), the barrel tube travels, whereas the plunger remains stationary. The barrel tube is connected to the
556
sucker-rod string through a connector and a large travelling valve. The standing valve connects directly to the top of the stationary plunger. The plunger is supported by a long, hollow pull-tube, which is connected to the bottom holddown. The surging action of the fluid around the bottom of the travelling barrel tube keeps sand from settling and sanding-in the pump. The large travelling valve carries the upstroke load of the fluid and during periods of shutdown acts as a built-in sand check valve. The barrel tube must resist a collapsing load on the upstroke. On the downstroke, pressure is balanced on the inside and outside of the tube, with the exception of any shock loading caused by fluid pound. Inasmuch as the leakage path for the plunger is from the bottom to the top, gravity aids in preventing sand from scouring and wearing the tube and plunger. The standing valve, which must fit inside the barrel tube, is smaller than the standing valve on stationary tube pumps. Because of the construction of the pump with the open-type cage and the connector on the tube, it is not possible to get quite as good compression ratio as compared to a stationary barrel pump. This generally results in lower volumetric efficiency when compared to the stationary barrel pump. To reach the compression chamber, the produced fluids must pass through a long hollow pull-tube. Because of the long fluid passage, the smaller standing valve, and the comparatively smaller compression ratio, the travelling barrel type pump is not recommended in wells having gas problems. In wells where the fluid level is very low, a large closed-type standing valve, sometimes called a foot valve, is installed between the holddown and the pull-tube. Fluid is then trapped in the pull-tube instead of surging back to the lower fluid level line. On the downstroke, the fluid load is taken by the standing valve. The pull-tube, in turn, must support this compressive load. The lower barrel tube plug serves as a guide and gives some measure of support. Because of buckling tendency and wearing action, a long travelling pump is seldom used in deep wells. The use of oversize connectors to prevent the barrel tube from rubbing against the tubing is quite popular. The travelling barrel pump generally has fewer components than a stationary pump and is cheaper. The advantages of rod pump with travelling barrel include: (1) absence of sand settling due to agitation, (2) good plunger leakage path, (3) built-in check valve, (4) both cages are of open type, (5) rugged construction, (6) stronger pull-tube instead of valve rod of stationary barrel tubes, and (7) lower cost (usually) than stationary barrel pumps. Disadvantages of these pumps are as follows: (1) poor performance in wells with gas problems, (2) not recommended for long pumps in deep wells, (3) poor valve placement, (4) poor fluid flow design, (5) tendency of gas break-out, and (6) poor valve sizing. (4) Rod pump with Stationary barrel and bottom hold-down
Pumps with stationary barrel and bottom hold-down utilize a seating mechanism at the bottom of the barrel tube which holds it stationary, while the plunger is free to travel with the motion of the rods (Fig. 14-19). The standing valve is generally
557
RWB RHB Fig. 14-19. Schematic diagram of rod pump with stationary barrel and bottom hold-down (RWB, RHB). (Courtesy of Axelson, 1980, p. 13.)
larger than the travelling valve which is preferred over the travelling-barrel design where the reverse is true. The produced fluid must flow across the smaller opening of the travelling valve. Any resulting gas breakout does not affect the operation of the pump. In this type of pump, the valve placement is better and the flow design causes less gas interference than the travelling type. The barrel tube is surrounded by stagnant well fluid between the barrel and the tubing. As a result, there is a trap for the accumulation of sand or other sediments as well as an undisturbed area for corrosion to take place. This could stick the pump in the tubing and cause a stripping job. A bottom discharge valve is sometimes used to keep the area surrounding the barrel tube free of sand and sediments; however, the pressure drop across this valve can restrict the filling of the pump chamber and induce gas breakout.
558
The weakest part of these assemblies is the valve rod and its small threads. As the size has been increased, the trouble from this source was reduced but not eliminated. Because of the basic design of this type of pump and distribution of the pressures involved, its use is recommended in the deeper wells. Advantages of these pumps include: (1) adaptability to deep wells, ( 2 ) smaller possibility of having pressure ruptured tubes than in the case of other standard pump types, (3) good valve location, (4) preferred valve sizing, ( 5 ) good flow design, (6) varied material selection, and (7) better design where long pumps are necessary. Disadvantages can be summarized as follows: (1) valve rod is a weak link in sucker-rod chain, ( 2 ) barrel tube is subjected to sedimentation and corrosion, (3) part-time pumping may allow sedimentation to take place in working parts, (4) poor plunger leakage path, and (5) requirement of more parts at generally higher cost than in the case of travelling-barrel type pump.
R WA
RHA Fig. 14-20. Schematic diagram of rod pump with stationary barrel and top hold-down (RWA, RHA). (Courtesy of Axelson. 1980, p. 14.)
559
(5) Rod pump with stationary barrel and top hold-down
In rod pump with stationary barrel and top hold-down, the barrel hangs from the hold-down (Fig. 14-20). Fluid is discharged immediately above the hold-down and keeps sand from settling and sanding up the pump. The pump barrel can be used as a gas anchor for better gas separation. On the downstroke, the entire fluid load is supported by the standing valve. The barrel tube must also withstand this tensile load. Consequently, t h s type of pump is not recommended for deep wells. Inasmuch as formation pressure only acts on the outside of the tube, the fluid column pressure on the downstroke on the inside, which is usually greater than the formation pressure, acts to split the barrel. The standing valve is positioned below the hold-down; hence, in low-fluid level wells, it has the best possibility of being submerged when compared to other pump types. Pumps of this type for tubing larger than 2 in. in diameter require oversize seating nipples for cup hold-downs. The advantages of these pumps include: (1) good tolerance in sandy wells, (2) pump barrel can act as gas anchor, (3) excellent adaptability to low-fluid level wells as standing valve can be submerged, (4) excellent fluid flow design, (5) preferred valve sizing, and (6) good design where long pumps are necessary. Disadvantages of these pumps are: (1) valve rod is a weak link in chain of sucker rods, (2) poor performance in deep wells because of bursting and tensile load on barrel, (3) poor plunger leakage path, (4) part-time pumping may allow sedimentation to take place in working parts, (5) requirement of more parts at generally higher cost than travelling-tube type of pump. (6) Rod pump with stationary barrel and top and bottom hold-down
Rod pump with stationary barrel and top and bottom hold-down is a non-standard type of pump, which utilizes two hold-down mechanisms attached to the top and bottom of the barrel (Fig. 14-21). Its use requires the simple construction of a section of tubing, to each end of which is attached a seating shoe or nipple corresponding to the mechanism on the pump, so that the length will position both hold-downs simultaneously or nearly so. The best pump assembly is the top-cup hold-down and bottom mechanical hold-down, with corresponding seating nipple at the top and mechanical hold-down shoe or nipple at the bottom of the tubing shell. As the “no-go” swell of the top-cup hold-down is removed, it is pushed completely through the cup seating nipple. The mechanical hold-down of the pump will pass freely through that nipple, but will stop and seal when it reaches its mechanical hold-down shoe. With the “no-go” removed, the top-cup hold is then adjusted in its seating nipple as to the length of the shell and provides a seal and hold-down. The advantages of these pumps include: (1) excellent stabilization for any and, particularly, for longer pumps; (2) elimination of sedimentation around barrel tube; (3) great reduction of corrosive attack on exterior of tube; and (4) presence of valve placement, sizing, and flow design of stationary-barrel bottom hold-down pump typeThe disadvantages are: (1) loss of the top hold-down advantages of valve submergence and use of pump barrel as gas anchor, (2) retainment of valve rod
560
c
1 RWAB RHAB Fig. 14-21. Schematic diagram of rod pump with stationary barrel and top and bottom hold-down (RWAB and RHAB). (Courtesy of Axelson, 1982. p. 15.)
weakness (but will reduce its flexing due to stability of pump), (3) retainment of plunger leakage path and sedimentation possibility with intermittent pumping, and (4) increased cost due to the construction of shell.
THEORETICAL ANALYSIS IN SUCKER-ROD DESIGN
Although a sucker-rod pumping system appears to be a simple mechanical device, its mathematical analysis is quite difficult. This is due to the fact that the behavior of each component of the system is dependent upon the actions of other components. Additional problems arise due to the elastic nature of the rods, the tubing, and the produced fluid. Forces are applied to each one of them at various stages of the cycle, and each one will elongate or compress differently.
561
The sucker-rod behavior can be treated mathematically with one-dimensional wave equation, which is a partial differential equation with boundary conditions. Difficulties exist, however, in trying to define the boundary conditions which describe the behavior of the downhole pump. These difficulties arise because the pump behavior is controlled by the motion of the sucker-rod string, which must be established by the wave equation. Much work has been done by Dr. S.G. Gibbs on the solution of the wave equation and he has written numerous papers on its applications (Gibbs, 1963, 1977). The most commonly practiced method of determining sucker-rod behavior is presented in the API (1977) RP 11L. This method uses empirical equations which are based on correlations of actual test data. The designer must realize, however, that unusual conditions might exist downhole that will greatly affect the validity of the calculations. Some of these unusual conditions include: (1) slanted or crooked holes, (2) very viscous fluid, (3) excessive sand production, (4) excessive gas production through the pump, and ( 5 ) well flowing-off. Design of the sucker-rod string
Usually a tapered rod string is used, which is a combination of different lengths of rods of different sizes, with the largest-diameter rods at the top. The minimum and maximum loads expected during the pumping cycle must be known in order to choose and design suitable surface equipment to handle these loads. There are two basic methods of designing tapered sucker-rod strings. In the first method, a point in the string should be determined at which the stress in the rod equals the maximum safe working stress and from that point to the top a larger size rod is used. In selecting the lengths of individual sections, the unit stress at the end of each section is made equal to the maximum permissible working stress (see Craft et al., 1962, p. 295). In the more commonly used second method, the lengths are selected so as to make the unit stresses at the tops of the sections equal. The latter method has a greater safety margin as far as corrosion pitting is concerned and its applications are outlined in this section. The percentages of rods having different sizes in a tapered string can be obtained from Table 14-IV. The stress at any point in a rod string is equal to the stress due to the fluid load on the plunger plus the stress caused by the weight of rods below this point. In the case of elastic deformation, the ratio of the stress applied to a body to the resulting strain is constant. This ratio is called the modulus of elasticity, E . Stress is equal to the force per unit area, i.e., stress = F / A ; whereas strain is the fractional change in length due to stress, i.e., strain = e//. Commonly, force, F , is expressed in lb and cross-sectional area, A , of the unit under stress is expressed in in2. Elongation, e, and original length, I , must be given in the same length units. The force caused by the fluid load results from the pressure differential across the plunger (pump is set at a depth, L ) , having a cross-sectional area of A , : F=ApxA,
(14-3)
TABLE 14-111 Approximate coefficient of stretch, E (courtesy of Axelson, 1982, p. 35) Plunger size (in.)
a.
Rod size (in.) 5
7
7
5
R
4
R
X
3 4
7 U
5
1 4
n
7
R
1
~
0.62
0.47
0.38
0.59
0.44
0.35
0.56
0.42
0.32
0.26
0.86
0.65
0.52
0.81
0.60
0.48
0.78
0.57
0.44
0.36
1.24
0.93
0.76
1.17
0.87
0.68
1.11
0.81
0.63
0.51
1.69
1.27
1.02
1.58 2.06 2.63
1.18 1.53 1.95
0.93 1.21 1.55
1.51 1.97 2.51
1.11 1.44 1.83
0.86 1.12 1.43
0.90
3.11
2.26
1.76
1.41
3.73
2.70
2.12
1.70
0.70 1.15
Formulae: Approximate stretch of rods and tubing for a fluid having specific gravity of 1.0. Stretch = E X ( Dp/1000)2, where E = coefficient of stretch E,L,)Dp/lOOO; where L , = length of top section, ft; E , = coefficient of stretch for top section; and Dp = pump depth in ft. Stretch = ( E , L , E,L, L, = length of center section, ft; E, = coefficient of stretch for center section; L, = length of bottom section, ft; and E , = coefficient of stretch for bottom section.
a
+
+
TABLE 14-IV Rod and pump data (from API, 1977, 1RllL. table 4.1, pp. 7-9; courtesy of American Petroleum Institute). 1
2
3
4
5
6
Plunger
Rod
Elastic
Frequency
Rod string (% of each size)
diam., d, (in.)
weight,
factor,
1:
(Ib/ft)
constant, E, (in./lb-ft)
44
All
0.726
1.990 x 10
54 54 54 54 54 54 54
1.06 1.25 1.50 1.75 2.00 2.25 2.50
0.908 0.929 0.957 0.990 1.027 1.067 1.108
1.668 X 10K6 1.633X10-6 1.584~10-~ 1.525 X 1.460X10-6 1.391 x 1.318X10K6
1.138 1.140 1.137 1.122 1.095 1.061 1.023
44.6 49.5 56.4 64.6 13.7 83.4 93.5
55
All
1.135
1.270~
1 .000
100.0
64 64 64 64
1.06 1.25 1S O 1.75
1.164 1.211 1.275 1.341
1.382X10-' 1.319X10-" 1 . 2 3 2 ~lo-' 1.141 X
1.229 1.215 1.184 1.145
33.3 37.2 42.3 47.4
33.1 35.9 40.4 45.2
65 65 65 65 65 65 65 65 65
1.06 1.25 1.50 I .75 2.00 2.25 2.50 2.75 3.25
1.307 1.321 1.343 1.369 1.394 1.426 1.460 1.497 1.574
1.138X 1.127 X 1.110x10-6 1.090X10-6 1.070x 1.045 X 1.018 X 10K6 0.990X10-6 0.930X10K6
1.098 1.104 1.110 1.114 1.114 1.110 1.099 1.082 1.037
34.4 37.3 41.8 46.9 52.0 58.4 65.2 72.5 88.1
65.6 62.7 58.2 53.1 48.0 41.6 34.8 27.5 11.9
66
All
1.634
0.883 x lo-'
1.ooo
100.0
75
1.06
1.566
0.997 X 10-
1.191
Rod
No.
a
w,
.
F, -
7
1
8
9
10
11
1
1
5
H
4
X
I 2
1.Ooo
100.0 55.4 50.5 43.6 35.4 26.3 16.6 6.5
33.5 26.9 17.3 7.4
v,
27.0
27.4
45.6
m
W
TABLE 14-IV (continued) 2
3
4
5
6
Plunger
Rod
Elastic
Frequency
Rod string (% of each size)
diam., d, (in.)
weight,
cons tan t ,
w,
4
factor, Fc
1;
(lb/ft)
(in./lb-ft)
75 75 75 75 75
1.25 1.50 1.75 2.00 2.25
1.604 1.664 1.732 1.803 1.875
0.973 X 10V6 0.935 x 0.892 X 0.847 X 0.801 x
1.193 1.189 1.174 1.151 1.121
29.4 33.3 37.8 42.4 46.9
29.8 33.3 37.0 41.3 45.8
76 76 76 76 76 76 76 76 76 76
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.25 3.75
1.802 1.814 1.833 1.855 1.880 1.908 1.934 1.967 2.039 2.119
0.816 X 0.812 x 0 . 8 0 4 ~10K6 0.795 X 0.785 X 10K6 0.774 X 0.764 x 10K6 0.751 x 0.722 X 10V6 0.690X 10K6
1.072 1.077 1.082 1.088 1.093 1.096 1.097 1.094 1.078 1.047
28.5 30.6 33.8 37.5 41.7 46.5 50.8 56.5 68.7 82.3
71.5 69.4 66.2 62.5 58.3 53.5 49.2 43.5 31.3 17.7
77
All
2.224
0.649 x
1.000
100.0
85 85 85 85
1.06 1.25 1.50 1.75
1.883 1.943 2.039 2.138
0.873 x 0.841 X 0.791 x 0.738 X
1.261 1.253 1.232 1.201
22.2 23.9 26.7 29.6
22.4 24.2 27.4 30.4
22.4 24.3 26.8 29.5
86 86 86 86 86 86 86 86
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75
2.058 2.087 2.133 2.185 2.247 2.315 2.385 2.455
0.742 x 10K6 0.732 X 0.717 x 0.699X10-6 0.679 x 10 0.656 x 0.633 X 0.610 x
1.151 1.156 1.162 1.164 1.161 1.153 1.138 1.119
22.6 24.3 26.8 29.4 32.8 36.9 40.6 44.5
23.0 24.5 27.0 30.0 33.2 36.0 39.7 43.3
54.3 51.2 46.3 40.6 33.9 27.1 19.7 12.2
1
Rod No.
a
-
7
1
8
7
x
9
10
3 4
5 X
11
I
40.8 33.3 25.1 16.3 1.2
33.0 27.6 19.2 10.5
vl
m P
87 87 87 87 87 87 87 87 87 81 87
1.06 1.25 1S O 1.75 2.00 2.25 2.50 2.75 3.25 3.75 4.75
2.390 2.399 2.413 2.430 2.450 2.412 2.496 2.523 2.575 2.641 2.793
0.612X10-6 0.610~ 0.607X10-6 0.603 x 100.598 X lo-" 0.594x10-' 0.588 x 0.582 x lo-" 0.570X10-6 0.556X lo-" 0.522 X lo-"
1.055 1.058 1.062 1.066 1.071 1.075 I .079 1.082 1.084 1.078 1.038
24.3 25.7 21.7 30.3 33.2 36.4 39.9 43.9 51.6 61.2 83.6
88
All
2.904
0.497 X
1.om
100.0
96 96 96 96 96 96
1.06 1.25 1.50 1.75 2.00 2.25
2.382 2.435 2.511 2.607 2.703 2.806
0.670 x 0.655 X 0.633 X 0.606 x 100.578xlO-" 0.549 x
1.222 1.224 1.223 1.213 1.196 1.172
19.1 20.5 22.4 24.8 27.1 29.6
19.2 20.5 22.5 25.1 27.9 30.7
19.5 20.7 22.8 25.1 27.4 29.8
91 97 97 97 97 97 97 97 97
1.06 1.25 1S O 1.75 2.00 2.25 2.50 2.75 3.25
2.645 2.670 2.707 2.751 2.801 2.856 2.921 2.989 3.132
0.568 X 0.563 x lo-" 0.556 X 0.548 x 0.538xlO-" 0.528X10-6 0.515 x 0.503X10-6 0.475X10-6
1.120 1.124 1.131 1.137 1 A41 1.143 1.141 1.135 1.111
19.6 20.8 22.5 24.5 26.8 29.4 32.5 36.1 42.9
20.0 21.2 23.0 25.0 27.4 30.2 33.1 35.3 41.9
60.3 58.0 54.5 50.4 45.7 40.4 34.4 28.6 15.2
98 98 98 98 98 98 98 98 98 98 98
1.06 1.25 1.50 1.75 2.00 2:25 2.50 2.75 3.25 3.75 4.75
3.068 3.076 3.089 3.103 3.118 3.137 3.157 3.180 3.231 3.289 3.412
0.475 X 0.474~ 0.472 x lo-" 0.470~10-~ 0.468 X 10O.465xl0K6 0.463 x 10-" 0.460~ 0.453 x 0.445X10-6 0.428 x
1.043 1.045 1.048 1.051 1.055 1.058 1.062 1.066 1.071 1.074 1.064
21.2 22.2 23.8 25.7 27.7 30.1 32.7 35.6 42.2 49.7 65.7
78.8 77.8 76.2 74.3 72.3 69.9 67.3 64.4 57.8 50.3 34.3
99
All
3.676
0.393X10-6
1 .Ooo
100.0
"
75.7 74.3 72.3 69.7 66.8 63.6 60.1 56.1 48.4 38.8 16.4
42.3 38.3 32.3 25.1 17.6 9.8
TABLE 14-IV (continued) 1
2
3
4
5
6
7
8
Rod *
Plunger
Rod
Elastic
F~~~~~~~~
Rod string (W of each size)
No.
diam.,
weight,
factor,
1:
1;
(in.)
(lb/ft)
constant, E, (in./lb-ft)
107 107 107 107 107 107 107 107
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75
2.977 3.019 3.085 3.158 3.238 3.336 3.435 3.537
0.524 x 10 - 6 o.517x10-6 0.506 x 10V6 o.494xlo-6 0.480X10-6 0.464 X 100.447 x 0.430~
1.184 1.189 1.195 1.197 1.195 1.187 1.174 1.156
16.9 17.9 19.4 21.0 22.7 25.0 26.9 29.1
16.8 17.8 19.2 21.0 22.8 25.0 27.7 30.2
17.1 18.0 19.5 21.2 23.1 25.0 27.1 29.3
108 108 108 108
1.06 1.25 1.50 1.75
3.325 3.345 3.376 3.411
0.447 x 10-6 0.445 X 0.441 x 10-6 0.437 X
1.097 1.101 1.106 1.111
17.3 18.1 19.4 20.9
17.8 18.6 19.9 21.4
64.9 63.2 60.7 57.7
d,
w,
1
F,
9
10
11
1
3
5
x
4
x
49.1 46.3 41.9 36.9 31.4 25.0 18.2 11.3
108 108 108 108 108 108
2.00 2.25 2.50 2.75 3.25 3.75
3.452 3.498 3.548 3.603 3.731 3.873
0.432 X 0.427 x 0.421 x 10K6 0.415 X 0.400X 0.383 X
1.117 1.121 1.124 1.126 1.123 1.108
22.6 24.5 26.5 28.7 34.6 40.6
23.0 25.0 27.2 29.6 33.9 39.5
109 109 109 109 109 109 109 109 109 109 109
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.25 3.75 4.75
3.839 3.845 3.855 3.867 3.880 3.896 3.911 3.930 3.971 4.020 4.120
0.378 X 0.378 x 0.377 X 0.376 X 0.375 x 0.374 X 0.372 X 0.371 x 0.367 X 10K6 0.363 X 0.354 X 10W6
1.035 1.036 1.038 1.040 1.043 1.046 1.048 1.051 1.057 1.063 1.066
18.9 19.6 20.7 22.1 23.1 25.4 27.2 29.4 34.2 39.9 51.5
81.1 80.4 79.3 77.9 76.3 74.6 72.8 70.6 65.8 60.1 48.5
54.3 50.5 46.3 41.6 31.6 19.9
Rod No. shown in first column refers to the largest and smallest rod size in eighths of an inch. For example, Rod No. 76 is a two-way taper of and rods. Rod No. 85 is a four-way taper of and rods. Rod No. 109 is a two-way taper of 1 and 1; rods. Rod No. 77 is a straight string of rods, etc.
a
i,i, I.
a
568
Inasmuch as it is usually assumed that the pump is set at the working fluid level in the well, the pressure differential is actually equal to the pressure at depth L due to the column of fluid having specific gravity G: Ap
= 0.433GL
(14-4)
For a more general case, where working fluid level is situated at depth D, the pressure under the plunger, which is due to a column of fluid having height ( L - D ) in the casing, should be considered. Thus: Ap
= 0.433GL - 0.4336( L -
(14-5)
D ) = 0.433GD
For a tapered rod string with two sections having lengths L, and L , (in ft), cross-sectional areas of A , and A , (in in.2) and weights M I and M 2 (in lb/ft), the stress at the top of the lower section is equal to: 0.433LAp + L,M,
-
0.433LAp + L R , M ,
A,
( 14-6)
A1
The stress at the top of the upper section is equal to: O.433LAp+ L,M,
+ L2M2
-
0.433LAp + LR,M,
A2
+ LR2M2
(14-7)
A2
where R , and R , are the fractions (of the total rod length) of the lower and upper rod strings, respectively. If stresses at the top of two sections are equal, then: 0.433Ap + R I M l A1
-
0.433Ap + R , M ,
+ R2M2
(14-8)
A2
where: R,
+R, = 1
(14-9)
The above equations can be modified for any tapered string with more than two sections. This, however, makes the analysis more complex. (See Craft et al., 1962.) Example 14-1
It is estimated that production of a well will be 400 bbl/D with a pump setting depth of 6050 ft. Assuming a pump with a l:-in. plunger and using a tapered rod string, consisting of $-in., ?-in., and 1-in. rods, determine the length of each section. The rods are available in lengths of 25 ft.
569
Solution: From Table 14-IV (rod no. 86): R , = 0.406, R , = 0.294, and R , = 0.30. Then: L , = L R , = 6050 X 0.406 = 2456 f t of :-in. rods L , = L R , = 6050 X 0.30 = 1815 f t of :-in. rods L , = L R , = 6050 X 0.294 = 1779 f t of 1-in. rods. Using 25-ft increments: L , = 2450 ft, L , = 1800 ft, and L, = 1775 ft. As mentioned before, the maximum estimated stress must be compared with the allowable working stress for the rods being designed. This would result in an adjustment in the above results. Consideration of polished rod loads and, consequently, the maximum stress at any point in the rod string is the most important factor in this adjustment. This will be discussed later. Rod motion analysis
As discussed before, the pumping motion is supplied to the polished rod through a series of rigid members, with no fluid couplings. The power is transmitted through the various components in a sequence beginning with the gear reducer and then through the crank, pitman, walking beam, horsehead, and, finally, hanger bar to the polished rod. The rotating crank is connected to the walking beam by the pitman, thus causing the walking beam to reciprocate with a motion that simulates a simple harmonic one. Thus, theory and load analysis of beam pumping are based upon simple harmonic motion (SHM). (See Day and Byrd, 1980, p. 28.) Unlike hydraulic units, the upstroke and downstroke loads of the beam units are not constant and have certain maximum and minimum loads. This is caused primarily by the harmonic type of pumping motion (SHM), which begins slowly at the reversal, increases to a maximum velocity at midpoint, and then decreases to the next reversal. As a rule of thumb, the maximum velocity is approximately equal to 1.5 times the average velocity. In addition, the harmonic vibration of the rod string, which is related to pumping rate and rod-string length, affects the loads. These loads must be estimated prior to the selection of equipment for a pumping installation and before designing and selecting a suitable sucker-rod string. Rod load is maximum in the top unit of the string, i.e., in the polished rod. It is subject to a considerable variation during the double-stroke pumping cycle. Instantaneous load is a function of a large number of factors. These could be static and dynamic loads. If sucker rods were suspended statistically from a polished rod or if they were rising or falling at a constant velocity, the weight of sucker rods, W,, would be the only force acting on the polished rod. In dynamic case, however, when the sucker rods are accelerating, an additional acceleration load of [ ( W , / g ) a ] is added to the above force. The acceleration factor, a,is expressed as follows: a = a/g
where a is the maximum acceleration achieved by the sucker-rod string.
(14-10)
570
The most commonly used analysis of simple harmonic motion of the rod string without considering fluid acceleration involves the Mills (1943b) acceleration factor: SN2 70,500
a = 2v2SN2/g=
-
where S
of stroke, in., and N
= length
(14-11) = pumping
speed, strokes per minute (spm).
Example 14-2
Calculate the maximum polished rod load caused by 3000 f t of 1-in. sucker rods if the average speed is 20 spm and the polished rod stroke length is 44 in. The weight of sucker rod is 3.12 Ib/ft. Solution: Weight of sucker rods W,= 3.12 lb/ft X 3000 ft = 9360 Ib; a = 44 X 202/70,500 = 0.25; maximum load = acceleration load weight of rods = 9360 X a + 9360 = 11,700 lb.
+
Effective plunger stroke
The relative movement of the plunger stroke with respect to the working barrel is the controlling factor in determining the volume of oil handled during each stroke of the pump plunger. This relative motion is called the net or effective plunger stroke, which may differ from the polished-rod stroke because of many variables. These include (1) rod and tubing stretch, (2) plunger overtravel resulting from dynamic motion and elasticity of the rods, (3) rod vibration, and (4) subsurface friction effects. The basic stretch of the rod string in a given well fluid essentially depends on the length of the rod string. The string is loaded by its own weight alone only during the downstroke, whereas during the upstroke there is an additional weight of the liquid column acting on the plunger. The change in the liquid load entails a change in stretch, which is described, subscribing to Hooke’s law, as follows: e, = A L ,
FL EA,
=-
(14-12)
where F is force in lb, A , is the cross-sectional area of the rod under stress in in.2, and E is the modulus of elasticity, which is a characteristic of the rod material to which stress is applied, L is the length of the member under stress in ft, and e, ( = AL,) is elongation in ft (e.g., E for steel is equal to 30 X lo6 psi). Normally, magnitudes of elongation and length are in inches and feet, respectively, in which case eq. 14-12 becomes: 12F L e , =AL, = -
EA,
(14-13)
571
Substituting eq. 14-5 in eq. 14-13: e, =A L ,
= 12 X
0 . 4 3 3 G D A p L / (E , A , )
(14-14)
= 5 . 2 0 G D A p L / (E , A , )
In the case of a tapered rod string, the above equation must be applied to each section: e,,
5.20GDApL,
= AL,,
=
EAl
; er2= A L,, =
5 .20GDApL , EA2
; etc.
(14-15)
where AL,, is the elongation of length L , (ft) of rods having cross-sectional area A , ; A L,, is the elongation of length L2 (ft) of rods having cross-sectional area A , ; etc. Thus, the total stretch for the rods is equal to: e,, = AL,, = ( 5 . 2 0 G D A p / E ) (L , / A ,
+ L,/A, + ..
+
( 14-16)
)
Equation 14-14 can be used also for the elongation of the tubing as follows: e , = A L,,
(14-17)
= 5.20GDApL/EA,
where A , is the cross-sectional area of the tubing wall. During each pumping cycle, as the travelling and standing valves of the subsurface pump open and close, the fluid load is being transferred alternately to the tubing and to the rod string. During the downstroke, when the standing valve is closed and the travelling valve is open, a certain amount of elongation occurs in the tubing, which is caused by the fluid load. On the other hand, at the beginning of the upstroke the travelling valve is closed and elongation of the rods result. Stretch comes out of the tubing as a result of opening of the standing valve. The working
TABLE 14-V Tubing data (from API, 1977, R P 11L, table 4.2, p.10; courtesy of American Petroleum Institute) 1
2
3
4
5
Tubing size (in.)
Outside diameter (in.)
Inside diameter (in.)
Metal area (sq in.)
Elas tic cons tan t, E, (in./lb-ft)
1.900 2;
1.900 2.375
2; 3f 4 4;
2.875 3.500 4.000 4.500
1.610 1.995 2.441 2.992 3.476 3.958
0.800 1.304 1.812 2.590 3.077 3.601
0.500 X lo-' 0.307 X lo-' 0.221 x 10-6 0.154X10-' 0.130X10-6. 0.111 x 1 0 - 6
572
TABLE 14-VI Sucker-rod data (from API, 1977, RP 11L, table 4.3, p. 10; courtesy of American Petroleum Institute). 1
2
3
4
Rod size (in.)
Metal area (sq in.)
Rod weight in air,
Elastic constant, Er (in./lb-ft)
w, (lb/ft) 1
2 5 8
3 4
7 8
1 If
0.196 0.307
0.72 1.13
1.990X10-6
0.442 0.601 0.785 0.994
1.63 2.22 2.90 3.67
0.883 X 0.649 x 0.497 X 10K6 0.393 X
1.270 X
barrel moves upward as a result of tubing restoration to its original length and the plunger moves downward due to the elongation of rods. Thus, there is a decrease in effective plunger stroke, which is equal to the sum of rod and tubing elongations resulting from fluid load. In addition to the fluid load, the rod load consisting of the dead weight of rods and acceleration load causes additional rod elongation. The weight of the sucker rods suspended below any element of the string changes uniformly from zero at the bottom to the maximum value of Wr at the top of the string. An average weight of Wr/2 centered at L/2 could be considered effective in causing rod elongation. On considering both dynamic and static loads, the elongation of the rods at the end of the downstroke, ed, is equal to: (14-18)
At the end of the upstroke, when the acceleration loads are in opposite directions, the elongation of the rods is equal to: (14-19)
The net elongation resulting from acceleration, which is termed the plunger overtravel, ep, is equal to: ep = ed - e ,
= 12WraL/EA,
(14-20)
573
For the rod string, having specific weight of yr in lb/cu ft, the weight in Ib is equal to: W, = y,LA,/144
(14-21)
Inasmuch as steel density is equal to about 490 Ib/ft3, substituting eq. 14-20 into eq. 14-21 results ip: (14-22) where a is the acceleration factor which, as presented earlier, is equal to: a=- S N 2
70,500
As pointed out by Craft et al. (1962), some investigators prefer to use the following more empirically correct formula: ep = 32.8L2a/E
(14-23)
On substituting 30 X lo6 for E of steel in eqs. 14-22 and 14-23 gives: ep = 1.36 X 10-6L2a or ep =
~
40.8L2a E
(14-24)
The rod and tubing stretch caused by fluid load decreases the effective plunger stroke, whereas the plunger overtravel increases it. Thus, the effective plunger stroke is equal to:
sP= s + ep - ( e , + e , )
(14-25)
where Sp= effective plunger stroke, in.; S = polished rod stroke, in.; ep = plunger overtravel, in.; e , = tubing stretch, in.; and e , = rod stretch, in. Combining eqs. 14-16, 14-17, 14-22 and 14-25 yields: (14-26) The above equation can be simplified as follows in the case of untapered rod: (14-27)
574
Inasmuch as there is no tubing stretch in the case of anchored tubing, the term A , can be neglected (Craft et al., 1962, p. 293). Pump-size determination
The major factor in the selection of a suitable pump size is the volume of fluid displaced by the pump per unit length of each stroke. The diameter of the pump bore determines the displaced volume; therefore, for a given pumping depth and amount of fluid to be produced, there is an optimum size of pump bore which will result in effective pump plunger travel and optimum speed of operation. According to Craft et al. (1962, p. 299), the theoretical pump displacement Vp in bbl/D is equal to: 1440 min/day
PD = V, = (A,, in.' = 0.1484ApS,N
bbl/day
(14-28)
The factor 0.1484AP is termed the pump constant, K , which is independent of surface operating conditions. Thus, eq. 14-28 can be written as follows: (14-29)
Vp = KSpN
TABLE 14-VII
Pump constants (from API, 1977, RP 11L, table 4.4, p. 10; courtesy of American Petroleum Institute) 1
2
3
4
Plunger diame ter,
Plgr. diam. squared
Fluid load factor a (0.340 x d i ) (Ib/ft)
Pump factor (0.1166X d l )
1.1289
0.384
1.5625 2.2500
0.531 0.765
0.132 0.182
3.0625 4.oooo 5.0625 6.2500
1.041 1.360 1.721 2.125
0.357 0.466 0.590 0.728
7.5625 14.0625
2.571 4.781
22.5625
7.671
0.881 1.640 2.630
d, (in.)
1k If 1; 1; 2 2a 2f 2: 3; 4: a
4
(sq in.)
For fluids with specific gravity of 1.00.
0.262
575
Plunger areas and pump constants for all API pump sizes are presented in Table 14-VII. The volumetric efficiency of the pump is defined as the ratio of the volume of the fluid actually handled to the pump displacement: E,
=
Q/v,
(14-30)
where Q is the rate of well production in bbl/D. In the case of good separation of formation gas in the hole and ample pump submergence, the volumetric efficiency commonly ranges from 70 to 80%. Many factors, such as fluid properties, surface operating conditions, the pump type, depth of pump, and gas interference affect the efficiency. In the case of foamy, gaseous production, the efficiency may be as low as 2 5 4 0 % . For wells with no gas interference and high fluid level the volumetric efficiency may be close to 100%. Example 14-3
Production from a pumping well is 145 bbl/D of 30"API (sp. gr. = 0.87) oil. The l 2 - k plunger is set at a depth of 6100 f t in 2i-in. tubing (2.875 in. O.D., 2.441 in. I.D.). The fluid is at a depth of 4800 f t in the casing annulus. The rod string consists of :-in. and i-in. rods and operates at 19.5 spm. Pump efficiency is 80%. Calculate the following: (a) Effective plunger stroke. (b) Tubing stretch. (c) Rod stretch. (d) Polished rod stroke. (e) Plunger overtravel. Solution: (a) The total pump displacement is equal to:
V, = P D = A , (in.2) X S, (in./stroke)
= 0.1484A,SpN
and Q = PDE,
A,
= 0.1484A,S,NEV
2
= ~ / 4 ( l $ ) = 2.405
in.2
X
N (strokes/min)
X
1440 min/day 9702 in.3/bbl
576
Thus:
Q sp= 0.1484Ap NE,
c
145 (0.1484)(2.405)( 19.5)(0.80)
= 26
in.
= ~ / 4 ( 2 . 8 7 5~ 2.4412) = 1.812 in.2
(b) A, et =
-
5.20GDApL - (5.20)(0.87)(4800) (2.405) (6100) EA t (30 X 106)(1.812)
= 5.860
in.
Assuming that L , and L, are equal to 3233 ft and 2867 ft, respectively: er=
i
5.20 x 0.87 x 4800 x 2.405 30 X l o 6
(d) S,
=S
+ ep - e , - er
Therefore: Sp= 26 = S
+ 0.311s - 5.860 - 21
Solving for S :
S
= 40
in.
= 0.311s = 0.311 X
40 = 12.4 in.
Polished rod loads calculation
In the design of surface equipment for a pumping installation, the anticipated maximum or peak polished rod load must be estimated as accurately as possible. Five major factors affect the net polished rod load: (1) the dead weight of sucker rods, (2) fluid load, (3) buoyant force resulting from submergence of sucker rods in the fluid, (4) acceleration load of sucker rods, and (5) frictional forces.
577
As shown before, the weight of a tapered rod string is equal to:
W,= M I L , + M , L,
+ ..
(14-31)
*
where M is the weight per unit length (lb/ft) and L is the length (ft). The maximum and minimum acceleration loads of the rods can be expressed as follows: Minimum acceleration load
= -
Maximum acceleration load
=
( 14-32)
Wra
W,a
(14-33)
The volume of rod string is equal to:
V, = volume = weight/specific weight = W,/y,
(14-34)
Inasmuch as specific weight of steel is usually assumed to be 490 Ib/cu ft, V , in cu ft is equal to:
V, = W,/490
(14-35)
For a fluid having specific gravity G, the buoyant force, B,, on the rods, which is equal to the weight of displaced fluid, can be determined as follows: B, = - ( WI/490)62.4G= - 0.127WrG
(14-36)
The negative sign indicates the upward direction of this force. The fluid load is equal to the weight of the fluid column which is supported by the plunger. The volume of fluid, V, (cu ft), is the difference between (1) the volume of a column having the sucker-rod string length and the plunger base, and (2) the volume occupied by the rods:
V, = (LA,/144)-( W,/490)
(14-37)
where L is the sucker-rod string length (ft) and A , is the plunger area (in2). Thus, the fluid load on the polished rod during the upstroke is equal to:
W,= 62.46[( ~ ~ ~ 1-4(W,/49O)] 4 )
= 0.433G(LA, - 0.294y)
(14-38)
The frictional forces, F,, can be estimated from the dynamometer tests. They are positive during the upstroke ( + F,) and negative ( - F,) on the downstroke, resulting from their opposition to the direction of motion of the body. Thus, the maximum (or peak) polished rod load, W , , , during the upstroke is equal to: W,,
=
W,+ W, + W,a + F,
(14-39)
578
whereas the minimum polished rod load, Wmin,which occurs during the downstroke, is equal to: Wmin= W, - W,(Y- 0.127W,G - Ff
(14-40)
Inasmuch as the frictional forces usually can be neglected, the above equations can be simplified as follows (Mills, 1943a,b):
w, + w, + W,a
(14-41)
Wmin= W, - W,(Y- 0.127W,G
(14-42)
w,
=
and
It should be noted that the vibration loads, which are not considered in the above equations, will be discussed later. Also, the acceleration loads resulting from the fluid being lifted are neglected.
COUNTERBALANCE DESIGN
The first step in the design and selection of surface equipment for any beam pumping system is the proper design of the counterbalance system. During the first half of the crank cycle (upstroke), when the amount of polished rod work to lift the fluid is high, the counterbalance would release the energy stored in the second half of the cycle (downstroke). If the pumping unit were not properly counterbalanced, the amount of work done by the prime mover during the upstroke would be maximum and consequently the fluid will not be produced effectively and efficiently. The counterbalance system uniformly distributes the loads and torques exerted on the prime mover and reduces the size of the prime mover and gear reducer. Theoretically, the “ideal” counterbalance effect, C,, can be estimated by the average load during a full crank cycle, namely, the prime mover carries the same average load on the downstroke and on the upstroke. This could be expressed as follows: (14-43)
On substituting eqs. 14-39 and 14-40 into the above equation, the following relation is obtained: Ci = Wf/2
+ W, - 0.127G W,
( 14-44)
519
Fig. 14-22. Counterbalance effect of the countenveight. (Modified after Craft et al., 1962, p. 301, fig. 5.4: courtesy of Prentice-Hall, Inc., New Jersey.) Point 0 represents the crankshaft, whereas point P represents the saddle bearing. Fp = the force in the pitman: C, = counterbalance effect; W, = Counterweight: r = distance from crankshaft to pitman bearing: d = distance from crankshaft to the center of gravity of the counterweight.
The above equation shows that the ideal counterbalance effect is equal to the sum of half of the fluid load and the weight of rods in the fluid. As shown in Fig. 14-22, the counterbalance effect, C,, caused by the counterweight of W,, depends on the geometry of a beam pumping unit, the stroke length, and the weight and position of the counterweight. In addition to C,, the counterbalance effect, C,, may result from structural unbalance of the surface installation. The total counterbalance effect C, at the polished rod, therefore, is equal to:
c,= c, + c,
(14-45)
where:
d = the distance from crankshaft to the center of gravity of the counterweight; r = the distance from crankshaft to the pitman bearing; I , = the distance from the saddle bearing to the tail bearing; I, = the distance from the saddle bearing to the bridle; and W, = weight of the counterweight used with the unit, lb. (See Day and Byrd, 1980, p. 39.)
5 80
TORQUE CALCULATION
Basically, torque for a lever arm is defined as the product of the arm length and the force acting at the end of it, which tends to produce rotation and work. For a pumping unit, torque is the amount of force (commonly expressed in in.-lb) caused by the pitman pull due to well loads and by the opposing effect from counterbalance moments and by the prime mover (Day and Byrd, 1980, p. 39). The torque is being applied to the crank by the low-speed shaft of the gear reducer. In any pumping unit, therefore, the actual peak torque must not be more than the maximum torque rating (capacity) of the gear box or speed reducer. Figure 14-23 shows the instantaneous torque on the gear box. The crank makes an angle B with the vertical, which is measured clockwise from crank position at the beginning of the upstroke. The net torque at any position of the crank is the difference between well load torque and counterbalance torque. The torque on the gear reducer (net torque about point 0) is equal to: T,
=
Wr sin B - W,d sin B
(14-47)
In the case when the geometry of the surface installation is not considered and C, = 0, eq. 14-45 is reduced to the following form: C, = 2Wcd/S
(14-48)
On substituting eq. 14-48 into eq. 14-47 the following relation (Craft et al., 1962, p. 303) is given: .T, = S / 2 W sin 8 - S/2 C, sin 8
or T, = ( W - C,)( S/2) sin B
(14-49)
As pointed out by Craft et al. (1962, p. 303), this equation is an approximate expression for instantaneous torque on the gear box. A simple relationship for the W
Fig. 14-23. Instantaneous torque on the gear box. (Modified after Craft et al., 1962, p. 303, fig. 5.5: courtesy of Prentice-Hall, Inc.. New Jersey.) W = polished rod load.
581
peak torque can be derived by substitution of the highest possible values for W and sin 8 , which are W,, and sin 90" ( = l), respectively: (14-50) In the design of any pumping unit installation, T, should be calculated for both upstroke and downstroke, because under some conditions either one can be greater than the maximum allowable torque for the unit. Inasmuch as the counterbalance can be between 90-95% of the ideal value, a simple expression for the peak torque prediction is as follows: T,
=
!sl
(W,,, - 0.93Ci) -
(14-51)
PRIME MOVER HORSEPOWER REQUIREMENTS
In producing fluid (from the pump to the surface), two major power loads must be taken into consideration: (1) hydraulic horsepower, H,, or the power required to lift a given volume of fluid vertically, through a given distance in a given period of time, and (2) the frictional energy loss, H,, between the pump and polished rod. The total polished rod horsepower, HI (or P R H P ) , therefore, is equal to: H, =H,
+ H,
(14-52)
The polished rod horsepower is defined as the rate of the work energy transferred to the polished rod by the pumping unit. Using the safety factor of 1.5, proposed by Kelly and Willis (1954), the brake horsepower, H , , is equal to: H,
= 1.5( Hh
+ Hf)
(14-53)
The hydraulic horsepower, H,, can be calculated as follows:
H,
=
Q X L , X Y, x G (33,000 ft-lb/min/HP)(24 hr/D)(60 min/sec)
= 7.36 X
QGL,
(14-54)
where Q = production in bbl/D, G = specific gravity of produced fluid, L , = net lift expressed in feet of produced fluid, and y, = specific weight of water in lb/bbl ( = 350 Ib/bbl). The total pressure differential as the fluid moves from the pump to the surface is called net lift, L,. It is not equal to the pump setting depth because of the effects of casing and tubing pressures.
582
If the pump is set at the working fluid level, however, the contribution of casing pressure in lifting the fluid becomes zero. On the other hand, if the pump is set at a depth L below the working fluid level, which is at depth D in the casing, the effect of casing pressure is equal to ( L - 0). The tubing back pressure, pt (psi), effect, which acts as a force against the lifting of fluids, actually adds to the net fluid lift (ft). Thus, the net lift is equal to: L,
=L
- ( L - D )+ pt/0.433G
=D
+ pJO.4336
(14-55)
It is also necessary to consider the frictional energy loss between the pump and polished rod. A simple empirical equation, developed by Zaba and Doherty (1956), can be used to calculate the frictional energy loss, H , (in HP) as follows: (14-56) or
H , = 6.31 xlO-’W,SN
’
(14-57)
where N = strokes per minute; 33,000 = conversion factor-ft-lb/min/HP; and 12 = conversion factor-in./ft. It is important to note here that the interested reader should consult excellent books by Craft et al. (1962) and Brown et al. (1980). Example 14-4 An analysis of dynamometer card shows a maximum load of 12,000 lb and a minimum polished rod load of 4000 lb, and polished rod horsepower of 12. The well has a 1:-in. plunger set in 2-in. tubing on 4300 f t of :-in. sucker rods. The tubing is not anchored and the working fluid level is low. When the well is pumped at 20 spm with a 64-in. stroke, the production is 300 bbl/D of a fluid having specific gravity of 0.83. Calculate (1) the ideal counterbalance effect, (2) the maximum sucker rod stress, (3) the polished rod horsepower, and (4) the peak torque if the unit is counterbalanced to within 5% of the ideal value. Solution: a s - -S N 2 - 64
70,500
2o
2o
= 0.363
70,500
According to Atlantic Richfield Co. Manual, for air-balanced Mark I1 systems the constant is equal to 6.25 X lo-’.
583
= 64
+ 4300
= 53.74
4300 [40.8 x 0.363 - 5.20 x 0.83 x 2.405(1/1.304 30 X lo6
+ 1/0.442)]
in.
W, = ML = 1.63 X 4300 = 7009 lb W, = 0.433G( LA, - O.294Wr)= 0.433 X 0.83(4300 X 2.405 - 0.294 X 7009) = 2976 lb W, = Wf + W,(1 + a) = 2976 + 7009 X 1.363 = 12530 Ib Ci = 0.5Wf W,(1 - 0.127G) = 0.5 X 2976 + 7009(1 - 0.127 X 0.83) = 7758 lb H , = 7.36 X 1OP6QGLn= 7.36 X x 300 X 0.83 x 4300 = 7.88 HP H , = 6.31 X 10-'W,SN = 6.31 x l o p 7 x 7009 x 64 x 20 = 5.66 HP Ht = 7.88 + 5.66 = 13.54 HP Ci = +(W, + Wmin)= $(12,000 + 4000) = 8000 lb Maximum stress = W,,/A, = 12,000/0.442 = 27,150 psi Tp = (W,, - 0.95C)(S/2) = (12,000 - 0.95 x 8000)(64/2) = 140,800 in.-lb
+
API RECOMMENDED DESIGN PROCEDURE
Summarized in this section are the equations described in detail in API (1977) R P 11L for a conventional pumping system design through trial and error methods. The design procedure is based upon correlations, data tables, and curves which are presented in this section. The following three steps can be followed in designing an installation: (1) A preliminary selection of components for the installation. (2) Use of formulas, tables, and figures to calculate the operating characteristics of the preliminary selection. (3) Comparison of the calculated pump displacement and load with the volumes, load ratings, stresses, and other limitations of the preliminary selection. Listed below are the minimum amount of information which must be either known or assumed: (1) Fluid level, D ,the net lift in ft. (2) Pump depth, L, ft. (3) Pumping speed, N, strokes per minute. (4) Length of surface stroke, S , in. (5) Pump plunger diameter, d,, in. (6) Specific gravity of fluid, G. (7) The nominal tubing diameter and whether it is anchored or free. (8) Sucker rod size and design. Knowledge of these factors makes it possible to determine the following: (1) Plunger stroke, S,, in.
584
P
c
POLISHED ROD POSITION Fig. 14-24. Basic dynagraph card. (Modified after API. 1979, RP 11L, fig. 3.1: courtesy of American Petroleum Institute.) F, = gross plunger load: W,, = weight of rods in the fluid: Fl = peak polished rod I ; F2 = minimum polished rod load (MPRL)-point 6: I = bottom stroke: load (PPRL)-point 3 = polished rod card for pumping speed greater than zero ( N > 0): 4 = polished rod card for pumping speed. N 0 ; 5 = top of stroke: S = polished rod stroke, in.
( 2 ) Pump displacement, PD (or V,), bbl/D. (3) Peak polished rod load, P P R L (or W,,), lb. (4) Minimum polished rod load, M P R L (or W,,,), lb. (5) Peak crank torque, P T (or T,), in.-lb. (6) Polished rod horsepower, P R H P (or H,). (7) Counterweight required, CBE, lb. These variables can be determined from the following equations:
sp=[(s,/s)x s ] - [ F , / k , ) I
(14-58)
where F, is the gross plunger load, and its significance is shown on a dynagraph card (Fig. 14-24). The value of S,/S is determined from Fig. 14-25. Fo = 0.340 X G X dp’ X D
(14-59)
and l/k,
= E, X
L
(14-60)
where E , is the elastic constant of the tubing and can be determined from Table 14-V. In the case of anchored tubing, l / k , is equal to zero. The pump displacement, P D , is equal to: PD
= 0.1166
x Spx N x dp”
(14-61)
Peak polished rod load, P P R L , is equal to: P P R L = W,, + ( F , / S k , ) X S k ,
(14-62)
585 I .7
1.6
1.5
I.4
1.3
1.2
1.1
SP
1.0
S 0.9
0.8
0.7
0.6
0.5
0.4
u.3
U P
“.I
0.I
0
0.2
0.4
0.3
0.5
0.6
0.7
N -
Nb
Fig. 14-25. Plunger stroke factor, S,/S. (After API. 1977, RP I l L , fig. 4.1: courtesy of American Petroleum Institute.)
where S k , = load (in lb) necessary to stretch the total rod string an amount equal to the polished rod stroke, S ; W,, = weight of the rods in the fluid, which can be determined on using the following formula:
W,, = W, X L
X
(1 - 0.128G)
(14-63)
586 12
1.1
I0
09
08
0.7
Fl Skr
06
05
a4
0 3
a2
0.1
I
0. 0
0. I
02
04
03
I
0.5
I
l
l
0.6
N NO
Fig. 14-26. Peak polished rod load, F 1 / S k , . (After API, 1977, R P 11L, fig. 4.2; courtesy of American Petroleum Institute.) N = pumping speed, strokes/min: No = natural frequency of straight rod string, stroke,/min.
The weight of the rods, W, (lb/ft), can be determined from Table 14-IV, whereas the nondimensional parameter (F,/Sk,) (peak polished rod load) is determined from Fig. 14-26. Minimum polished rod load, M P R L , is equal to: M P R L = W,, - [ ( F 2 / S k , ) X Sk,]
(14-64)
where F2 = M P R L factor which can be determined from Fig. 14-24. The parameter ( F2/Sk,) (minimum polished rod load) can be determined from Fig. 14-27. Peak torque, Tp,is equal to:
Tp = (2T/S2k,)
X
Sk, X T, X S/2
(14-65)
587
N NO Fig. 14-27. Minimum polished rod load, F 2 / S k , . (After API. 1977, RP 11L, fig. 4.3; courtesy of American Petroleum Institute.)
where T = crank torque, lb-in.; T, = torque adjustment constant for values of ( W , , / S k , ) other than 0.3. The latter adjustment can be obtained from Fig. 14-28, whereas the value of ( 2 T / S 2 k , )can be obtained from Fig. 14-29.
588
Fig. 14-28. Adjustment for peak torque for values of W , , / S k , other than 0.3. (After API, 1977, RP 11L, fig. 4.6: courtesy of American Petroleum Institute.) In order to use, multiply % indicated on curve by [( Wr,/Sk,)-0.3]/0.1. For example, for W , , / S k , = 0.600, N/N6 = 0.200, F , / S k , = 0.188. Adjustment = 3% for each 0.1 increase in Wrf/Skr above 0.3. Total adjustment = 3 X 3 % = 9%. T, =1.00+0.09 =1.09. If W , , / S k , is less than 0.3, adjustment becomes negative. (Nd = natural frequency of tapered rod string, strokes/min.)
589
2T -
S2k,
1 %I
Fig 14-29 Peak torque, 2 T / S 2 k , , for values of W r f / S k , = 3 (After API, 1977. RP 11L. fig 4 4 courtesy of Amencan Petroleum Institute ) Use torque adjustment for values of W,,/SI\ other than 0 3 ~
Polished rod horsepower, P R H P , is equal to: H , = PRHP
=
(F,/Sk,)
X
Sk, X S
X
N
X
2.53 X
(14-66)
590
Fig. 14-30. Polished rod horsepower, F 3 / S k r . (After API, 1977, R P 11L, fig. 4.5; courtesy of American Petroleum Institute.)
The parameter ( F , / S k , ) (polished rod horsepower) can be obtained from Fig. 14-30. The counterweight required is determined by using the following equation: CBE = 1.06( W,,
+ +Fo)
(14-67)
591
Fig. 14-31. Percentage increase in fundamental frequency for 1;-, 1-, and :-in. three-way tapered string. (After API, 1977, R P 11L, fig. A.l; courtesy o f American Petroleum Institute.)
In Figs. 14-26, 14-27, 14-29 and 14-30, the value of the term (N/N,) is determined using the following equation: N/N,
= (NL)/245,000
(14-68)
where N = pumping speed, strokes/min (spm); No = natural frequency of straight rod string, strokes/min (spm); and L = pump depth, ft. The term (N/NA) is equal to:
592
CHANGE OF FUNDAMENTAL FREQUENCY
FOR I,
8
a
TAPERED R O D STRINQ
i-,
Fig. 14-32. Percentage increase in fundamental frequency for 1-, and :-in. three-way tapered rod string. (After API, 1977, RP 11L,fig. A.2; courtesy of American Petroleum Institute.)
where F, = frequency factor (a constant of proportionality), which depends upon the rod design. The dimensionless pumping speed (N,”;) is a significant index of the behavior of rod string. The frequency factor, F,, can be obtained from Table 14-IV. For rod strings which are not presented in Table 14-IV, Figs. 14-31 through 14-36 can be used to determine the frequency factor from the following relationship:
F, = 1.O + % from Table 14-IV
(14-70)
Sample step-by-step design calculations are shown in Table 14-VIII using the above-shown formulas.
593
i-, :-,
Fig. 14-33. Percentage increase in fundamental frequency for and :-in. tapered rod string. (After API, 1977, R P 11L, fig. A.3; courtesy of American Petroleum Institute.) DYNAMOMETER CARDS (DYNAGRAPHS)
The dynamometer card is a continuous record of polished rod load versus polished rod position. It can be used for the determination of main design factors: (1) polished rod load, (2) peak load, (3) peak torque, and (4)horsepower requirements. It also enables the evaluation of pumping well problems. The well character and the production history should be known before any card interpretation. It is also important to note here that it is virtually impossible to perform a well analysis by studying a single dynamometer card. The basic principles of card analysis are presented here. As pointed out by Day and Byrd (1980), the following information is developed in a simple card analysis:
594
a, 38i
CHANGE OF FUNDAMENTAL FREQUENCY TAPERED ROD STRINO FOR
i-, i-,
Fig. 14-34. Percentage increase in fundamental frequency for and :-in. three-way tapered rod string. (After API, 1977, R P 11L, fig. A.4; courtesy of American Petroleum Institute.)
(1) Torsional load on the speed reducer and prime mover; the unit’s torque factors must be known and properly applied. ( 2 ) Minimum and peak pumping unit structural loads. (3) Proper counterbalance. (4) Work done by the polished rod against the elevation of the fluid and against friction. (5) Minimum and peak rod loads-rod stress, and load range. (6) Number of rod load fluctuations per crank cycle. The dynamometer profile enables the operator to visualize the polished rod load pattern as well as to evaluate the pump performance.
595 40
>
8
30
U
PERCENT OF LARGEST ROD, L t / L
Fig. 14-35. Percentage increase in fundamental frequency for four-way tapered rod string. (After API. 1979, R P 11L, fig. A.5; courtesy of the American Petroleum Institute.)
A unique single force signal is transmitted ( = 15,800 ft/sec) along the sucker-rod string to the surface, where it is recorded by the dynamometer, for each stroke of the downhole pump. Maximum and minimum rod loads, polished rod horsepower, correct counterbalance, and torque can be obtained from these cards. In addition, the following information can be obtained from these cards: (1) magnitude of a gas or fluid pound, (2) frictional loss, (3) gas locking of pump, (4) condition of the travelling and standing valves, (5) overtravel or undertravel of the pump plunger; and (6) degree to which the well is pumped-off.
VISUAL DIAGNOSIS OF OPERATING CONDITIONS
The dynamometer card will be a rectangle (Fig. 14-37) for an idealized pumping system in the presence of the following conditions: (1) no vibrational or frictional forces within the system, (2) no time lag in transmitting motion from surface to the plunger, ( 3 ) no rod elongation due to fluid load transfer, (4) the standing valve opens and the travelling valve closes instantaneously at the beginning of the upstroke, (5) the travelling valve opens and the standing valve closes at the beginning of the downstroke, and (6) no acceleration forces-well is pumped very slowly.
Fig. 14-36. Percentage increase in fundamental frequency for specific rod string combinations. (After API, 1979, RP 11L. fig. A.6; courtesy of the American Petroleum Institute.)
591
In Fig. 14-37, the line A B represents the upstroke, where the load remains constant and the polished rod load is equal to the fluid load plus the weight of rods in the fluid. The line CD represents the downstroke where the load remains
TABLE 1CVIII Example design calculations for conventional sucker-rod pumping system (from API, 1977, RP 11L, p. 6; courtesy of American Petroleum Institute; and Axelson, 1982, p. 62; courtesy of Axelson, Inc.). Object: To w i r e for-Sp.
PD. PPKL. MPRL. PT.PRHP. and CRF:
Known or Assumed Data:
598
TABLE 1CVIII (continued) COMPANY:
A-a-D
PUMP SIZE:
2'/r ( 2 )
1.
.
CO
TYP1::Rod W e i g h t / F t . X Pump Depth F t .
-SIZE _ _ NU\lBEI? _
FEE7'
IVEIGliT P E R FOOT
X /300 X 2200
X
3240
lC8G
GRAND TOTAL X S p . G r . = WON
X-X-. so00 1.0 = S B S O Lbs. V a l u e s on p a g e 65 a r e g i v e n i n F l u i d Load/ C F t . WR + WON = SL S t a t i c Load
/ o s / ~+
4.
s8So
=
( 6 0 ) (14 1 2 =
.I6
Dynamic Rod Load (1 + . I G ) X lOr/ Peak P o l i s h e d Rod Load
T WR = RDY 12,200 Lbs.
( 1 + T)
8
=
WR ]
[ ( l + T)
+
WON
= PL
- -
(1 + * I b 1 x l o t i 8 + 58SO = / 8 , o r 0 Lbs Minimum Rod Load WR (.76 T) = ML I Oh1 8 X (.76 ./b j = 6 3 / 0 Lbs. P L - ML = RPR Load Range / BOSO 63iO = 117CCO Lbs. Peak Torque R p ~ X (LJ4) = P.T. (est.)
-
b0 ) = 176, 100 Lbs. ( 7
f1ly.O 10.
P o l i s h e d Rod H . P .
11.
Production 0.1484 x
X
14
S t r o k e Loss
-
X Lp
4/,CX
12. 13.
= i4ppR
LN R P R 750,000
0 . 1 4 8 4 X Ap(;
3.19
1 2 (WOC)
seS0
Lbs.
16368
Lbs.
16
Lbs.
T
70,500 5.
'ON
16368
L N ~ 70,500
Impulse F a c t o r
60"
7692.
IJ7*
3.
L:
TOTAL ROD WEIGHT
Weight of F l u i d / F t . X L i f t / F t .
F l u i d Load (Net)
14
WR
3.59 2.84 2.16 1.63
yo00
TOTAL FEET 2.
SPM:
Rod Load ( . 4 i r )
1-1/8" 1"
I3tCH
WELL NUMRI
X
/4
(est.)
N
= =
B/O
'%?I ____
*
599 A
0
ACTUAL CARD
0
A 0
-
D
t
Fig. 14-37. Example of an ideal and actual dynamometer card for a beam pumping unit
constant; the rod string is falling freely and the polished rod load is equal to the weight of the rods in the fluid. For an actual pumping system, however, it is impossible to achieve the ideal conditions and the shape of the card deviates greatly from a full rectangle, depending on existing conditions. A typical pumping cycle on a dynamometer card is shown in Fig. 14-38. An illustration of a typical dynamometer card for a crank-counter-balanced unit is presented in Fig. 14-39. The shape of dynamometer cards is influenced by the following: (1) fluid conditions, ( 2 ) pumping depth, (3) pumping speed, (4) abnormal conditions of the pump, (5) pumping unit characteristic and geometry, and (6) frictional forces. Several operating problems can be diagnosed through examination of dynamometer cards: (1) plunger undertravel and overtravel, ( 2 ) fluid leakage from travelling or standing valve, (3) fluid pound, (4) gas lock, (5) sticking plunger, (6) restriction in the well, (7) excessive friction, (8) vibrations, (9) synchronous pumping speeds, and (10) abnormal load conditions. (See Eickmeier, 1967.) Some of the above conditions are illustrated in Fig. 14-40. a-g. As mentioned before, the shape of the dynamometer card is a good indicator of the performance of pumping equipment. Detailed analysis of the dynamometer card
; ,
Maximum Lood
6 TA
V
V
9\ 8
Minimum Lood
Stroke
Fig. 14-38. A typical dynamometer card. (Modified after Russel, 1953; courtesy of World Oil.) 1 = polished rod down, 2 = travelling valve closing, 3 = recoil, 4 = rods and fluid are being lifted, 5 = deceleration of walking beam, 6 = polished rod up, 7 = standing valve is taking over load, 8 = rods and plunger are falling through the fluid, and 9 = deceleration of walking beam. Load scale: 1 in. = 12,500 Ib.
600
TOP DEAD CENTER
COUNTERBALANCE YELPS L I F T ( P ) , LOWERS REDUCER TORQUE
REDUCER TORQUE LIFTS REMAINING COUNTERBALANCE MOMENT POLISHED
ZERO
@
0
POLISHED ROD LOAD
n
DYNAMOMETER CARD
LOADW HELPS LIFT COUNTERWEIGHTS, REDUCER TORQUE LIFTS REMAINING COUNTERBALANCE MOMENT
COUNTERBALANCE HELPS L I F T (3) LOWERS REDUCER TORQUE
BOTTOM DEAD CENTER
Fig. 14-39. Typical dynamometer card for a crank-counterbalanced unit. (Courtesy of Atlantic Richfield Company; fig. 3-9, p. 3-16; from Artificial Lift - Sucker Rod Pumping Manual.)
601
Well History Nax l o a d lriin load
4-22-42
Range a p ee d stroke O i l
Water Lax l o a d Iriin l o a d Range dpeed Stroke PUP O i l
Water
Max l o a d
&in load Range Speed btroke PUP Oil Water
kax l o a d U i n load Range Speed dtroke PUP
Oil Water
12,800 l b 4,000 l b 8,800 lb 1 6 spm 64 i n . 2 1/4-in. p l g r 453 b b l p e r day 0 b b l p e r day 18,600 l b 5,100 l b 13,500 l b 16 s p a 64 i n . 2 1/4-in.plgr 189 b b l p e r day 199. b b l p e r day
18,600 l b 3,000 l b 15,600 l b 20 spm 64 i n . 2 in.-plgr 1 5 2 b b l p e r day 205 b b l p e r day
18,900
lb
3,600 l b 15,300 l b 22 spm 64 in. 2 1/4 i n . p l g r 129 b b l p e r d a y 239 b b l p e r day
The above dynamometer c a r d s i l l u s t r a t e c h a n g i n g w e l l c o n d i t i o n s o v e r a p e r i o d of f o u r y e a r s , F i e l d s which a r e p r o d u c e d a t a r e l a t i v e l y h i g h r a t e u s u a l l y have a r a p i d d e c l i n e i n bottom h o l e pressure. The horsepower used i n l i f t i n g t h e w e l l f l u i d i s a f u n c t i o n of t h e e n c l o s e d area o f e a c h dynamometer c a r d . This g i v e s some i d e a a s to bottom h o l e p r e s s u r e d e c l i n e and t h e added horsepower n e c e s s a r y when s a l t wtlter h a s t o be produced. Fig 14-40 a.b.c,d.e.f,g Examples of dynamometer cards showing some uell ~onditionsand problems (Courtesy of Bethlehem Steel Compank )
Fig. 14-40.a.
602
Overtravel
24000
24000
m-
-&-
16000
/
8000
>
d-
I6000
/
8000
e z -
NO. 4
NO. I
24000
24000
16000
8000
8 000
__c
-
0
r
NO. 2
NO. 5 24000
24000
l
6
0
0
0
B
-5iz
-
/
8 000
8000
-
/
I
NO. 3
NO.6 P o l i s h e d rod dynamometer c a r d s were t a k e n a t t h e same time bottom h o l e dynagraph c a r d s were r e c o r d e d . T h i s shows s u c c e s s i v e s t a g e s i n a w e l l pumping up. I t i s s i g n i f i c a n t t o n o t e t h a t p l u n g e r t r a v e l i s g r e a t l y a f f e c t e d by t h e f l u i d l o a d . T h e p l u n g e r o v e r t r a v e l d e c r e a s e s a s t h e fluid load increases. Card No. Speed &. s t r o k e Plunger s t r o k e Overt r a v e l Time Fio IA-Anh
1 18-52-in. 62 i n . 10 in. 3:lO
2 18-52-in. 60 in.
8 in. 3:40
3
4
16-52-in. 18-52-111. 58.5 in. 57 i n . 6.5 i n . 5 in. 4:lO 4:40
5 18-52-in. 57 i n . 5 in. 5:lO
6
18-52-111. 53 in. 1 in. 5:40
603 0
Undertravel TOTAL DEPTH 4050 f t FOR MATION XD A I L Y P R O D ~ C T I O N :0 1 1 WATERSP. G R A V . S T R O K E LENGTH 64 in* S . P . M . 16 A.V.E. 3 7 DCt FLUID LEVEL: STATIC PUMPING B.H.P. W E L L C L A S S I FICATION: AGITATOR PUMPER DOES WELL POUND No CASING HEAD P R E S . V e n t e d Yes H2SC O R R O S I V E CONDITIONS: PITTING CASING SIZEL~LFEET TUBING SIZE 3 in. G A S ANCHOR No TUBING ANCHORED No 5/8" S U C K E R RODS: l''1000 f t 718"3/4" PUMPING UNIT Twin Crank MOTOR M u l t i - C v l i UNIT RATING: LOAD 25,000 lb PEAK T O R Q U E 285°00"dyi*-lb_ GEAR BOX: S I N G L E DOUBLE RATIO
Max l o a d Min l o a d Range Speed Stroke
P o l rod hp Time
hiax l o a d lviin l o a d Range Speed atroke P o l r o d hp
Time
0
Fig. 14-4O.c.
~ ~ ~ I a l-
W
Li4
W LL
m
0
!-
22000 l b 4700 l b 17300 l b
16 s?m
64 i n . 20.6 9:oo HiLl
23700 l b 4500 l b 19300 l b
1 7 spm 64 i n . 25
1o:oo
IUtI
604
Fluid Pound Max Load
Min Load Range
6000
Speod Stroke P o l Rod hp Time
I
Sucker Rods
lb lb lb spm 54 in* 10.1
14500 2800 11700 18.5
8:40 AM
--
1500 f t 7/8 in. 2000 f t 3/4 in. Pump 1 3/4 i n . bore RLB App vol eff 2 7 . 7 p c t
It i s g e n e r a l l y r e c o g n i z e d by most p r o d u c e r s t h a t d e t r i m e n t a l s t r e s s e s a r e imparted t o t h e mechanical equipment of a pumping w e l l when i t i s poundinq f l u i d . I n o r d e r t o v i s u a l i z e more e a s i l y what hapnens when a w e l l i s pounding, the above example Is g i v e n a s an i l l u s t r a t i o n . The p o l i s h e d rod t r a v e l s from "a" t o "b" b e f o r e t h e t r a v e l i n g v a l v e in t h e pump opens. This means t h e approximate t o p t h i r d of the pump b a r r e l i s f i l l e d w i t h gaseous f l u i d , which m u s t b e compressed t o a p r e s s u r e i n excess o f t h e f l u i d weight above t h e t r a v e l i n g v a l v e b e f o r e t h e t r a v e l i n g v a l v e w i l l open. I n t h i s c a s e (3500 f e e t pumping d e p t h ) t h e pressure amounts t o about 1400 p s i . A t p o i n t "b" t h e p r e s s u r e below t h e t r a v e l ing v a l v e must exceed 1400 p s i . When t h e t r a v e l i n g v a l v e i s opened t h e f l u i d l o a d i s i n s t a n t l y t r a n s f e r r e d from t h e sucker r o d s t r i n g t o t h e t u b i n g . The f l u i d weight i s now supported by t h e s t a n d i n g v a l v e i n s t e a d of t h e t r a v e l i n g valve. In this c a s e we have a sudden r e d u c t i o n i n l o a d amounting t o 8950 l b . Fig. 14-40.d.
enables determination of: (1) the load range, (2) average loads, (3) counterbalance, (4) torque; and (5) the polished rod horsepower. (API l l L 2 , 1969.) The procedure for the loads determination is shown on Fig. 14-41, where D,= maximum deflection (in.), D, = minimum deflection (in.), A , = lower area of card (in.*), A , = upper area of card (in.,), and L = length of the dynamometer card (in.). If the calibration constant of the dynamometer card is C (lb/in. of card height), then: Maximum load
=C
x D,
(14-71)
605
Max l o a d Min l o a d Range Speed Stroke Time
15,000 l b 6,000 l b 7,000 lb 18 spa 60 i n .
Max l o a d Min l o a d Range Speed Stroke Time
15,000 l b 4,000 lb 11,000 l b 18 spa
Max l o a d Min l o a d Range Speed Stroke Time
14,800 l b
8 : 4 5 Ai
GAS LOCK
60 in. 1 0 : 4 5 AM
GAS POUND
2,800 lb 12,000 lb 18 spm 60
in.
11:05 Ahil
PUMPING
These c a r d s show th e normal c y c l e , which i s o f t e n ex p eri en ced
when a well i s in t h e semi-flowing s t a g e . The shape of t h e card i s alm o st c o n t i n u o u s l y changing a s t h e well flows and pumps d u r i n g v a r i o u s i n t e r v a l s o f time. Fig. 14-40.e.
Minimum load
=CX
(14-72)
D,
Range of load = C( D,- D 2 ) Average upstroke load = C( A ,
(14-73)
+A , ) / L
(14-74)
606
Excessive Friction
& ;12ooo-
soon
6 o o o
Max l o a d Min l o a d
19,800 l b 2,000 lb
Range Speed dtroke P o l rod h p
17,800 l b 22 spm 44
in. 18.8
30 ring pump plunger
3000
18000
15000 I2000
9000 6000
Max l o a d tdin l o a d Range Speed Stroke P o l rod hp
17,000 l b 2,000 l b 15,000 l b 23 spm 44
in.
18.5
3000
15 r i n g pump p l u n g e r
I5000
12000
3000
Max load Min l o a d Range Speed Stroke P o l rod hp
14,800 lb 3,000 l b 11,800 l b
23 spm 44
in.
15.4
V
7 ring pump plunger
t h e s e examples a r e g i v e n t o show how e x c e s s i v e t h e pump can a f f e c t t h e Deak Dolished r o d l o a d r o d horsepower. I n t h i s - case- t h e only changes t h e number o f r i n g s used on t h e pump plunger. new i n each case. Fig. 14-40.f.
friction i n and p o l i s h e d made were in The pump w&s
607
Heavy Pumper O-I
TOTAL D E P T H 4009 ft FORMATION- Iiunton Lime 35 __ WATERSC SP. G R A V 4 0 2D A I L Y P RODUCT ION: 011 S T R O K E LENGTH 74 in* S.P M . 20 A.V E. 7 8 . 5 u c t FLUID LEVEL: STATIC PUMP I NG LOW B.H.P. I W E L L CLASS1 FICATION: AGITATOR PUMPER IQ DOES WELL POUND No CASING HEAD P R E S . W C O R R O S I V E CONDITIONS: PITTING No H2S No -0 4CASING S l Z E r l F E E T TUBING S I Z E 2; i n * -IW GAS ANCHOR No TUBING ANCHORED 90 __ w lL S U C K E R RODS: 1” 7 / 8 ” 3 9 5 0 ft 3/4“ 5/0”-TYPE STEEL Carbon .35; N i c k e l . 6 5 ; Chromium .40: Molybdepnrn; 2 PUMP 28 i n . X 2 f i n . X 1 6 . 5 ft PLGR. S I Z E 2% i n . a ffl 6 PUMPING UNIT Twin C r a n k MOTOR blcyl 3 UNIT RATING: LOAD 25.000 PEAK TORQUE 305.000 in.-&? 0 r GEAR BOX: SINGLEpp DOUBLE X RATIO 3 1 . 4 I-
I
OJY
8-
10-
Max Load Min L09d Range Speed Stroke P o l r o d hw Time
16,900
lb
4,000 lb 1 2 , 8 0 0 lb
17 spn 74 i n . 19.7 11:15 AFb
L”ax l o a d 19,000 l b Yin load 2,500 lb Range 16,200 lb Speed 20 spn 74 i n . Stroke Pol r o d h p 33.5
Time
~~
Fig. 14-40.g.
~
12:oo
FT!
608
Counter bolonce
Fig. 14-41.Dynamometer card showing areas and deflections needed for calculating loads. (Modified after Zaba, 1962; courtesy of Petroleum Publishing Co., Tulsa, Okla.)
Average downstroke load
= CA,/L
(14-75)
If the polished rod stroke length and the pumping speed are known, the polished rod horsepower can be calculated as follows: PRHP=C(A
L 2/
SXN 33,000(12)
(14-76)
where S = stroke length, in.; N = stroke per minute, SPM; A , = area of dynamometer card, in.2; and L = length of the card, in. (See Day and Byrd, 1980, p. 67.) The counterbalance effect from dynamometer cards is determined using the following procedure: (1) The counterbalance ( C B ) line should be drawn on the card at the position of maximum counterbalance effect, i.e., when the crank arm is horizontal on the upstroke at 13 equal to 90" and 270". The angle- is measured in the clockwise direction. (2) The ideal counterbalance effect, CB,, is approximately equal to: CB, =
PPRL + M P R L 2
(14-77)
(3) The actual counterbalance effect, CB,, is equal to: (14-78)
CB, = C X D,
(4) The correct counterbalance effect, CB,, can be determined from the following relationship: CBc = :(average upstroke load
+ average downstroke load)
(14-79)
If the counterbalance line is not drawn, the approximate correct counterbalance can be determined from the following equation: CB, = C ( A ,
+A2/2)/L
(14-80)
609
Fig. 14-42. Fagg's approximate method for calculating instantaneous torque. (After Fagg, 1950: courtesy of the Soc. Pet. Eng. of AIME: also in: Craft et al.. 1962, p. 344. fig. 5.31: courtesy of Prentice-Hall Inc.)
This relationship gives the same result as eq. 14-79 above. (See Brown et al., 1980, p. 67.) Instantaneous torques throughout the pumping cycle can also be determined from the dynamometer cards. An approximate method proposed by Fagg (1950) requires only information from a dynamometer card. In this method, simple harmonic motion is assumed for the rods, i.e., a uniform angular velocity for the crank. I t is also assumed that the pitman is vertical all the time. The geometry of the surface installation is ignored. This method is generally sufficiently accurate and can be easily used by the field engineer (Craft et al., 1962, p. 344). Figure 14-42 illustrates the approximate method for instantaneous torque determination (also see Fig. 14-43). The line of counterbalance effect and zero load are shown by C and D , respectively. Points A and B represent the beginning of the upstroke and the beginning of the downstroke, respectively. The angle 6 is between the crank and the vertical, and is measured clockwise from the crank position at the beginning of the upstroke. After points A and B are projected vertically onto the zero load line, a semicircle is drawn having diameter A B . Points on the dynamometer card corresponding to given crank angles are then determined as follows: (1) Radii are constructed in order to divide the semicircle into equal segments ( e g , 15" each). (2) The intersections of the radii with the semicircle are projected vertically. (3) These projections intersect the load curve at instantaneous values of polished rod load, W , at various crank angles 8. The instantaneous torque can be calculated from the following equation (see Craft et al., 1962, p. 345): T = ( W - C ) ( S / 2 ) sin B
(14-81)
I t should be emphasized that on the upstroke (point A to point B ) crank angle varies from 0 to 180°, whereas on the downstroke (point B to point A ) crank angle varies from 180-360" (Craft et al., 1962, p. 345).
610 0.
Counterbalancing TOTALDEPTH 4325 f t FORMATION W i l C O X DAILY PRODUCTION: 011 l2 WATER= SP. GRAV.& STROKE LENGTH 64 S.P.M A.V.E. 76 Dct FLUID LEVEL: STATIC L O W PUMPING L O W B.H.P. X W E L L CLASSIFICATION: AGITATOR PUMPER DOES W E L L POUND No CASING HEAD PRES. &PO CORROSIVE CONDITIONS: PITTING H2S-: CASING SIZE-FEET TUBING SlZE&Li.LGAS ANCHOR No TUBING ANCHORED No SUCKER RODS: 1" 718'' 2000 3/4''2300 518" TYPE STEEL Carbon .35; Nickel .S5; Chromium .40; Kolgbdeny
. L
PUMP 2 1/2 in. PUMPING U N I T m d UNIT RATING: LOAD GEAR BOX: SINGLE
PLGR. SIZE 2 MOTORS-r PEAKTORQUE DOUBLE RATIO
in*
O V
frant
I l-
e
W
w W
Lc v)
4
5 6 0 I I-
1c
I3000 10500 8000
18Oo15O0 18O021O0
120° 240°
900 60' 270. 300. CRANK ANGLE
COUNTERBALANCE EFFECT A T POLISHED ROD
30. 0 330. 360.
APPROXIMATE METHOD I n order t o b e t t e r i l l u s t r a t e how chanqes i n counterbalance v a r i e s t h e peak torque, t h e same dynamometer card was used t o p l o t torque curves based on 8,000 l b , 13,000 l b and 10,500 l b e f f e c t i v e counterbalance a t t h e polished rod, A counterbalance e f f e c t of 8,000 l b shows the well i s undercounterbalanced and has a peak torque of 300,000 i n . - l b on the upstroke and 150,000 i n . - l b on the downstroke. Negative torque occurs a t each end of t h e s t r o k e . With a counterbalance e f f e c t of 13,000 i n . - l b the well i s overcounterbalanced and has a peak torque of 300,000 i n . - l b on t h e downstroke and 150,000 i n . - l b on t h e upstroke. Negative torque occurs a t t h e end of the upstroke and the beginning of t h e downstroke. A counterbalance e f f e c t of 10,500 l b shows t h e well i s c o r r e c t l y counterbalanced having equal peak torque on t h e up and domstroke. This i s a l i t t l e under 250,000 in.-lb. Fig. 14-43. Approximate method of counterbalancing. (Courtesy of Bethlehem Steel Co.)
611
Pumping efficiency determination
In any pumping unit installation, the pumping equipment performance must be evaluated. A detailed analysis of the dynamometer card is of great help in this evaluation (Marsh and Watts, 1938). The required measurements are: (a) area between load loop and zero load line; (b) area within the load loop; (c) distance above the zero load line of points on the loop, showing maximum and minimum loads (Fig. 14-41); and (d) length of the load loop. Upon establishing the instrument constant for the dynamometer, one can determine the load range, average loads, and correct counterbalance of the unit. In addition, polished rod horsepower can be calculated if the polished rod stroke length and the pumping speed are known. The surface efficiency is calculated from the polished rod horsepower and the power input to the prime mover. The subsurface efficiency is determined from the polished rod horsepower and hydraulic horsepower, which is calculated from the net lift ( L " ) ,production rate (Q), and specific gravity of the fluid (G).
PROBLEM WELL TESTING
Determination of an oilwell productivity is indeed a complicated task. Problems related to tubing, rods, casing, formation, and subsurface pump could all cause reduction in productivity. In such a case, the proper use of a dynamometer card to evaluate the well performance and obtain basic load-time relation can aid in resolving the problem. Merryman and Lawrence (1958) and API (1977) have presented an excellent classification of problem wells and methods of isolating the cause of an unsatisfactory performance. The API recommended procedures for problem well analysis are presented in Fig. 14-44.a,b,c,d,e,f,g. ENERGY OPTIMIZATION
With continually increasing demand for petroleum, there has been an increased interest in optimizing lifting costs in sucker-rod pumping systems. Only a fraction of the power that is consumed by the prime mover of a sucker-rod system, is expended in lifting fluid (hydraulic horsepower). Mathematically, this can be presented by the following equations: P,
= PRHP X
(14-82)
E,
where P, = prime mover power consumption (HP), P R H P power, and E, = surface efficiency.
= polished
(14-83)
P R H P = H , t- H f
where H ,
= hydraulic
rod horse-
horsepower and H , = frictional horsepower.
Information to be collected prior to weighing operations:
m N
1. Production: daily oil, wtr, gas, allowable. 2. Pump: size and type. 3. Rods: size and type, length of each string. 4. Tubing: size, type and seat. 5. Mud anchor: size and type. 6. Gas anchor: size and type. 7. Producing interval and TD or PBTD. 8. Motor or engine: size. 9. Fluid: specific gravity. 10. Auxiliary equipment. 11. LS and SPM. 12. Pertinent well treating data. 13. Daily pumping time and schedule. 14. Power consumption. 15. Calculations: Rod weight in air. Rod weight in fluid. Fluid weight on pump (pounds) Volumetric pump capacity (bbls/day).
Weigh well, record load diagram, and make traveling valve and standing valve tests.
u Classify card.
Valves good fluid weight satisfactory
Page A
Valves good, fluid weight satisfactory Page B
Indicated Standing valve (SV) or Traveling valve (TV) Leak Page C
1-1
SV recorded
Page D
PI Page E
Fig. 14-44. a. A systematic approach to the problem-well tests. (After API, 1983, pp. 9-15, figs. 5-1 through 5-7; courtesy of the American Petroleum Institute.) page A: p. 613; B: p. 614; C: p. 615; D : p. 616; E: p. 617.
I
I
VALVES GOOD YI.UID WEIGHT SATlSFACTOItY
d
FPounding fluid.
Test TV & SV several times at one position. I
Check rod, tubing, pump design, Calculate pump efficiency. If TV indicates leak at one position and does not a t another, then the barrel has a worn spot.
Intermittent leak indicates that ball is Mechanical design
Mechanical design
valve and pres-
design (LS, SPM, and pumping time) a s economics justify.
equipment as economics justify.
Will build up pressure and hold. Tubing leak.
immediately.
I
I 1-1
Observe card for a short time to prevent misinterpreting severe gas compression as a fluid
Operational design could be improved.
Gas compression. Card shape will verify.
I Pump all load oil before pump pounds fluid.
to pound fluid.
Well producing
I Well producing a t capacity.
IBridee before tubing 1
J perforations High pressure tubing leak. Additional pressure tests will verify. i
pound fluid in relation to shut-in time.
inlet.
Gas anchor or pump inlet par-
o r mud anchor partly plugged.
Tubing perf01 itions bridged.
1 m c)
Fig. 14-44.b.
W
I
VALVES GOOD FLUID WEIGHT LESS THAN SATISFACTORY
I
I Fluid oound test. I b
Not pounding fluid. I
I
I
design if additional production is required.
Tubing partially loaded from intermittent flow.
%-
Verify by retesting after time lapse to check load increase.
i
I
L
I
orpGmpinlet.
is result of fluid pound beating gas out of the fluid.
fluid pound section
Will not pressure up. Load tubing from outside source to verify.
Will pressure up and hold.
U No tubing leak.
Fig. 14-44.c.
after time lapse to check load increase.
;-.L
Shut wing valve and pressure up tubing.
lapse to evaluate mechanical design.
to circulation from tubing to casing through wellhead or flow connections.
I
I -
I
I
INDICATED VALVE LEAKS TV or SV
1
I
Standing valve
Traveling valve leak indicated.
t
I
Special design large fluid slippage Pump.
-
1. Leaking traveling valve.
I
I Tubing leak.
2. Worn plunger or barrel. 3. Leaking component parts
Close wing valve. Re-weigh after a short time.
calculations, improve operating design.
Close flowline valve and pressure-up tubing.
Re-test well when efficiency justifies.
Run fluid pound test.
Pull pump when
I
Will not pressure up.
Will pressure up and hold.
I
Pressure tubing with outside source.
I
Will pressure Re-pressure several times to eliminate possibility of high pressure or intermittent leak.
Pull and inspect pump. If no failure is visible, drop SV and pressure up from outside source.
Pull and repair pump. Tubing leak. I
Pull and inspect tubing.
No tubing leak, no pump seat leak. Probable valve, plunger or barrel leak not visible in field. Fig. 14-44.d.
I
I
ONLY SV OR TV RECORDED
I
I
SV or rod weight only.
plus fluid weight.
I
b
Special pump only rod plus fluid weight recorded.
Well flowing through pump.
1
I
Tubing parted low. Tubing weight approximaiely equal t o calculated fluid weight and supported by pump.
I
necessary.
below pump.
Rods parted near pump or pump parted.
be relieved by lowering pump and tapping bottom.
Pressure tubing from outside source to verify.
I
If necessary shut well in for short time and re-weigh
I"""i]
pressure.
perforations.
t o pump, leak in pump seat, or
Leak in component parts of various
Situation can often be relieved by lowering pump and tapping bottom.
Pressure tubing with
I outside source.
the pump can often be seen on the dynar mometer a s the casing
with small volume of fluid and pump
I
o r hold pressure.
[""i""'
~~~
[Leak in pump above standing p l v e orlower rods parted.
Bad seating
joint of tubing Pump seating in tubing above the seating nipple.
I
1
1
Probable pump failure t h a t is not visible in field. Possible high pressure split, damaged or improper sized seat CUDS. etc.
__
I
*
Fig. 14-44.e.
I
Will build up and hold pressure.
I
Will not pressure up.
I I Pull tubing.
1
I
11fnot.1
leak, pump is seated and no leak exists in pump at o r helnw the valve. - . etandine . ~ -o~ ~ ~ ~
standing valve and pressure up tuhinc with an outside Ronrce.
Low tubing
Special pump, only rod weight recorded.
Note:
Bent joint of tubing ahove the spdting nipple may not pass a pump, but it may pass a short staiidinE valve and permit the tubing to hold fluid and pressure.
I1
ABNORMAL LOAD INDICATED BY VALVE MEASUREMENTS
j
I
r
Severe plunger restriction such as collapsed or pinched barrel, sand. scale, etc.
I
I
-
-
Distinct impact a t bottom of stroke.
-
I
8
Pump stuck in top portion of stroke.
Too many rods in the hole.
J
Static weight will measure same a t any stroke position.
I
Distinct impact
Maximum weight reachesorexceeds weight of rods in fluid near top of stroke.
Loosen stuffing
I
1
L Raise rods to verify. 1 ~
greater than rods plus fluid verifies
Fig. 14-44.f
J
I
Pump length shorter than stroke.
*
calculated fluid weight.
I
ABNORMAL LOAD INDICATED BY CARD SHAPE
I
equal to rods plus fluid verifies too
5
4
I
A SYSTEMATIC APPROACH TO PROBLEM WELL TESTS WITHOUT WEIGHT MEASUREMENTS
1
I
I
Check casing pressure. Bleed and maintain at minimum permissible
Pressure up and bleed several time to eliminate chance of high pressure o r intermittent
Pull the rods and pump and inspect the pump. If no failure is visible, drop a SV and pressure the tubing.
pull pump. If the rods and pump are in good condition and the gas anchor is clean, then the tubing perforations are plugged or the
the bottom hole conditions.
pass a free SV snd permit pressuring up of the tubing, but may not pass a pump.
Fig. 14-44.g
619
Examination of eqs. 14-82 and 14-83 shows that energy losses can occur either as a result of surface inefficiencies or frictional losses downhole. Surface efficiency
Both the prime mover and the pumping unit determine the surface efficiency of a sucker-rod pumping system. The overall surface efficiency is a function of both the average (constant torque) efficiency of the prime mover and the variation in the torque which are placed upon the prime mover. The torque load is directly related to the current. The surface inefficiencies increase with increasing span between the IRMS and the average current, i.e., with increasing torque range (Eickmeier, 1973). The torque range of the prime mover is influenced by both the prime mover and the geometry of the pumping unit. In general, with the increasing slip of the prime mover, the prime mover is more evenly loaded. Air-balanced and Mark type pumping units tend to produce more even loading than do conventional units. The optimization of surface equipment is generally a step that should be taken when the equipment is originally installed, because the power savings that can be incurred by replacing a prime mover or a pumping unit will not justify the expense of the new equipment. Subsurface efficiency
The hydraulic horsepower in eq. 14-83 can be determined using the following equation (Day and Byrd, 1980, p. 43): H,
= 7.36 X
10-6QGL,
(14-84)
The frictional horsepower can be determined empirically using eq. 14-57 (Day and Byrd, 1980, p. 44): H , = 6.31 X lO-’W,SN
(14-85)
Equations 14-84 and 14-85 indicate that if a well is pounding fluid, resulting in a low pump fillage, then the amount of frictional horsepower per barrel of fluid lifted will be greater than in the case of = 100% pump fillage. By combining eqs. 14-82 and 14-83, a power savings that can be incurred by volumetrically downsizing, can be determined as follows: PRHP, ps=--ES,
PRHP, ES 2
(14-86)
where Ps = power savings (HP), and the subscripts 1 and 2 indicate initial and final operating conditions, respectively. It is assumed that when a unit is volumetrically downsized to increase pump fillage and reduce lifting costs, the surface efficiency
100
80
,
I
,
I
,
-
,
,
,
,
I
,
,
,
,
,
1
,
k------------ -=WITH ANTI-FRICTION BEARINGS
\WITH
BRONZE BUSHINGS -I
-
-
Fig. 14-45. Relationship between speed reducer efficiency and horsepower loading. (Courtesy of Lufkin Industries.. Inc.), T X N Nominal H.P. = 63,000 X 1.57 API torque rating in Ib
Nominal horsepower rating at 20 SPM
API torque rating in.-lb
Nominal horsepower rating at 20 SPM
40,000 57,000 114,000 160,000 228,000 320,000
8 11 25 33 46 65
456,000 640,000 912,000 1,280,000 1,824,000 2,560,000
93 130 185 260 370 526
and the total fluid lifted will remain constant. Thus, eq. 14-86 can be rewritten as follows:
(14-87)
The value for the surface efficiency, E,, is either known from field experience or can be estimated from Fig. 14-45. The power savings, P,,predicted by this equation and those actually measured in field tests were in close agreement in the case of wells where only stroke lengths and stroke rates were changed. In the wells where pumps were changed to smaller ones, the power savings were greater than those predicted by eq. 14-87, because this equation does not take into consideration power savings by changing to a smaller pump (see Byrd and Beasly, 1974; Byrd, 1971). In order to predict power savings in wells in which pumps are changed, another method of determining polished-rod horsepower is required. It should be kept in mind that the API (1977) RP 11L method will not work in these instances because its correlations are based on 100%pump fillage (see Bommer, 1981).
621
Dynamometers can be used in order to determine the polished-rod horsepower in pumped-off wells. These measured polished-rod horsepowers can then be correlated in order to estimate the polished-rod horsepowers of similar wells in the field. Testing In the case volumetric downsizing to optimize pumping efficiency is going to be undertaken on numerous wells in the field, then it may be desirable to conduct power consumption tests in order to determine the amount of power that is actually being saved. In such a case, two test methods are available. One method of testing involves the use of kilowatt-hour meters, which are similar to those used in residences for electrical billing purposes. The meter is first attached to the pumping unit prior to any downsizing operation. The unit is then allowed to pump for a period of time (for about a week) in order to determine the initial power consumption. After the pumping system is ;ohmetrically downsized, once again it is allowed to pump for a period of time to determine the final power consumption. These kilowatt-hour meters are usually detented, i.e., they are racheted and will respond only to current flowing toward the prime mover. In some instances, when a unit is improperly balanced, the prime mover may actually be generating power during part of the stroke. As a result, the meter will indicate more power consumption than is actually occurring. The second test method involves the use of (1) a current (ampere) probe, (2) a voltage probe, (3) a power-factor transducer, and (4) a computer with graphical capabilities. The current and voltage, which are measured throughout an entire stroke, are plotted by the computer. The power-factor (which is the cosine of the phase angle between the current vector and the voltage vector) is measured by the power-factor transducer and is then plotted versus position in the stroke. The real power consumption at any position in the stroke can be determined by the computer by multiplying the instantaneous values of power-factor, current (amperes), and voltage (volts). Real power can then be plotted versus the position in the stroke. Thus, an average power consumption can be determined. Pump-off controls and timers In volumetrically oversized systems that are pounding fluid, it may be more desirable to install a pump-off control or a timer than it is volumetrically downsize the system by decreasing (1) the stroke length, (2) the stroke rate, and/or (3) the pump size. A pump-off control is a device that will shut down the pumping unit for a given period of time once a pumped-off condition is sensed by the controller. On the other hand, a timer will cause the unit to pump for a preset period of time and then be shut-off for another preset period of time. Some of the available pump-off controllers lend themselves quite well to field-wide automation. Many controllers may result in greater production due to their dynagraph generating capacity, which enables early detection of downhole problems.
622
SELECTION OF MATERIALS
Several basic factors must be considered in the selection of materials: strength, abrasion, corrosion resistance, and cost. Corrosion
The most important factor controlling a corrosion cell is the environment. If the environment is neutral, corrosion will not occur in subsurface pumps to any appreciable degree. The most common types of corrosion encountered in subsurface pumps are: (1) pitting corrosion, ( 2 ) erosion corrosion, (3) stress corrosion, (4) galvanic corrosion, (5) sulphide stress cracking corrosion, (6) corrosion fatigue, and ( 7 ) hydrogen embrittlement. (See Axelson, 1982, pp. 37-44.) ( 1 ) Pitting and concentrated cell corrosion Pitting corrosion and concentrated cell corrosion both result in pitting of metals. Pitting corrosion is the loss of metal at a localized area rather than over the entire surface. It may be caused by many different conditions in a subsurface pump. Pitting sometimes occurs in subsurface pumps because of the breakdown of (a) protective films, (b) coatings, (c) platings, (d) scale, and (e) non-metallic deposits, or by ineffective inhibitor treatment. The introduction of oxygen into a moist H,S environment promotes and accelerates this condition. Pitting may occur beneath non-metallic deposits, in the form of concentrated cell corrosion, or may appear in the form of crevice type corrosion at thread reliefs or subsurface pump thread connections. This type of corrosion can be retarded or checked by the proper selection of materials or by use of durable coatings or platings and effective inhibitor program. If left unchecked, however, this type of corrosion can be extremely troublesome, because it takes place in limited areas with accelerated rate of penetration.
(2) Erosion corrosion Erosion corrosion starts when damage is caused to the built-in defense mechanism of the material. Most materials employed in the manufacture of subsurface pumps have a built-in defense mechanism. They tend to form a protective film that retards or checks the advance of corrosion. Damage to this protective film initiates or accelerates corrosion. High-velocity impingement of corrosive-abrasive-laden fluid is responsible for many such failures in subsurface pumps. This type of corrosion failure can best be avoided by selecting the type of sucker-rod pump that will minimize high velocities and by employing abrasive-resistant platings or coatings in critical areas. (3) Stress corrosion Stress corrosion failure is the result of an interaction between a corrosive media and imposed stresses. Stresses may be applied, residual, or a combination of both.
623
This type of failure may occur in subsurface pump threaded connections where a combination of applied and residual stresses are high. To avoid this type of failure, one must select subsurface pumps that will minimize the stress at critical dynamic points and eliminate overtorque of threaded connections. (4) Galvanic corrosion
Galvanic corrosion takes place when two dissimilar metals are in electrical contact with each other in an electrolyte. The less noble of the two metals is attacked to a greater degree than if it were exposed alone. This type of attack is known as galvanic corrosion because the entire system acts as a galvanic cell. By referring to the galvanic series, which gives an indication of the rate of corrosion between different metals or alloys when they are in contact in an electrolyte, one can determine the possibility of galvanic corrosion. The metal close to the active end of the galvanic series will act as a cathode and will be protected. As an example, an aluminum fitting in a brass or monel part will corrode while the monel part will be unaffected. This type of corrosion can be prevented by not coupling dissimilar metals or by using a nonconductive couple.
(5) Sulfide stress cracking Sulfide stress cracking is usually associated with a spontaneous brittle failure of a ductile material which occurs as a result of an interaction between a moist hydrogen sulfide environment and an applied stress. Both ingredients must be present for such failures to occur. Materials heat-treated to high yield strengths (above 90,000 psi) with Rockwell hardness greater than 23 R, are more prone to suffer this form of corrosion failure. In order to avoid this type of failure, one must select materials with low hardness and minimize the applied stresses wherever possible. (6) Corrosion fatigue
Corrosion fatigue is more common in metals having high hardness values. Hydrogen embrittlement is a phenomenon that usually occurs in carbon steels and low-alloy steels, heat-treated to high hardness levels above R, 23. Such steels are more prone to allow diffusion of atomic hydrogen into the steel lattice, which results in brittlement and leads to a brittle type of failure. This type of failure can best be avoided by selection of a corrosion-resistant material at low hardness levels and proper handling of components during the electroplating process. (7) Material strength One of the most important considerations in the selection of the material is its strength. Materials are usually selected for their corrosion and/or abrasion resistance, while ignoring the strength of the material. The yield strength of the basic materials used in the manufacture of subsurface pumps are as follows: (1) 300 series stainless steel: 35,000 psi minimum, ( 2 ) 400 series stainless steel: 40,000 psi minimum, (3) yellow brass: 50,000 psi minimum, (4) carbon steel: 60,000 psi minimum, (5) monel: 70,000 psi minimum, and (6) alloy steel: 90,000 psi minimum.
624
There are variations of these strengths within each common classification above. As an example, some barrel tubes, which are manufactured from an Admiralty type brass, have strengths comparable to monel. Maximum setting depth for barrel tubes based on their strength of material are presented by Axelson (1982). They can be used as a reference for making proper selection of materials knowing the yield strength. Carbon steels and low-alloy steels are used in non-corrosive or very mildly corrosive and non-abrasive environments. In selection of carbon steel, the use of free-machning steels with a high sulfur content should be avoided. The nickel-copper alloys (monel) and the 7-30 brass, or yellow brass and Admiralty brass are basically recommended for use in H,S and brine environments. Monel is usually the best material in the case of high H,S concentration. The 300-series stainless steels are usually selected for environments containing high percentages of CO, in brine. Brass has been employed successfully in this type of environment. If both CO, and H,S are present, monel is about the only material which can satisfactorily resist corrosion. Common materials used in manufacture of pump components are briefly described here: Carbon Steel: Regular-”on-corrosive, non-abrasive environment. Hardened-”on-corrosive, moderately abrasive environment. Chrome-plated-Non-corrosive, severely abrasive environment. Metal spray-Non-corrosive, moderately to severely corrosive environment. Metal spray with protected threads in 1.D.-Severely corrosive and moderately to severely abrasive environment. Alloy Steel: Regular- Moderately brine corrosive, non-abrasive environment. Hardened- Moderately corrosive, moderately to severely abrasive environment. Nitralloy N hardened- Moderately corrosive, moderately to severely abrasive environment. Stainless Steel: 300 series- Moderately to severely corrosive, non-abrasive environment. 4/6 chrome regular- Moderately brine corrosive, non-abrasive environment. 4/6 chrome hardened-Moderately brine corrosive, moderately abrasive environment. 440 series-Mildly corrosive and abrasive environment. Brass : 70/30 (yellow)- Moderately corrosive, non-abrasive environment. 85/15 (red)-Moderately to severely brine corrosive, non-abrasive environment. Admiralty brass-Moderately corrosive, non-abrasive environment. Admiralty brass, chrome-plated- Moderately corrosive, severely abrasive environment. Monel: R-400 regular -Severely corrosive, non-abrasive environment. Chrome-plated-Severely corrosive and severely abrasive environment.
625
K-500-Non-abrasive, severely corrosive environment. Cast Sections Grey-white iron, regular-Non-corrosive, non-abrasive environment. Grey-white iron, hardened-Non-corrosive, non-abrasive environment. DiHard, eutectic, etc.-Moderately to severely corrosive and abrasive environment. Cast Materials; Cobalt alloy (Rexalloy, Stoody, Stel1ite)-Severely corrosive, moderately to severely abrasive environment. Carbide, tungsten-Severely corrosive and abrasive environment. Carbide, titanium or chrome- Severely corrosive and abrasive environment. T h s applies to the ball only. The seat always must be made from tungsten-carbide. According to Axelson (1982, p. 42), balls and seats used in subsurface pumps are of the flat type. Although the ball and seat is a relatively small item, it is the heart of the sucker-rod pump. All balls are purchased from one or two ball manufacturers in the U.S.A. Inhibitors Basically, a corrosion inhibitor may be any chemical which, when added to a corroding system, minimizes or prevents metal loss. The most common types of inhibitor used in pumping wells are either oil-soluble or water-soluble. Inhibitors minimize equipment failure and reduce the cost of operation of a particular installation by decreasing the number of times that the particular pump must be pulled for replacement of parts which have been corroded. This includes the sucker-rod string as well. There are two basic methods used to inject the inhibitor into a producing well: (1) continuous treatment and ( 2 ) batch treatment. In the continuous treatment method, the inhibitor is pumped into the system continuously by means of a pump, which may be electrical or gas-driven, or may be operated by the beam unit itself. In the continuous treatment method, the inhibitor must be flushed down the annulus in order to comingle with the produced fluid and return through the production column. Batch treatment involves addition of an inhibitor on a periodic basis for the purpose of corrosion control. The most successful batch treatment is achieved when an inhibitor which is introduced into the well slowly feeds into the production column. One advantage of the batch type treatment is that the tubing and the pump are exposed to a high concentration level of inhibitor and, therefore, have the best chance of being covered by the protective film of the inhibitor. Another type of batch treatment is to introduce the inhibitor into the well, flush it around the bottom, and then circulate out through the production column. Laboratory tests can be conducted to determine the best type of inhibitor to be used in a particular environment. The actual field service, however, is' the best method of establishing effectiveness of an inhibitor. It is also of utmost importance to determine whether the inhibitor will change the wettability of rocks or not.
626
INSTALLATION A N D OPERATION
As pointed out by Axelson (1982, p. 43), all of the care used in the manufacture and assembly of the pump is wasted unless it is carefully transported to the well and installed free of dirt, foreign matter, and mechanical damage. When transporting the pump, it is important to protect it from bending, dropping, or other damage. Also, a pump should not be disassembled at the wellsite and should not be laid on the ground prior to installation. If the pump is laid down at the wellsite, it should be placed off the ground on some type of support and be adequately supported at least every six feet. Prior to running the pump, it should be checked for freedom of plunger action. In addition, a final inspection should be made to insure that all caps, plugs, and protective wrappings are removed. All necessary tools and equipment should be in place when preparing to run the pump. Extreme care should be exercised in picking up the pump. The use of a short pony rod is advisable. Pumps over 20-ft in length require special handling. A proper wrench should be carried on the service unit for connecting the pony rod to the top of the pump. In tubing-pump installation, the barrel, which is already in place, is run with the tubing, whereas the plunger is run with the sucker-rod string. This requires extreme care so that the plunger is not damaged when running to bottom. When installing insert type pumps, extreme care should be used in handling the pump, particularly long pumps. They should be hoisted in such a manner as not to bind the plunger in the barrel tube or produce a permanent kink or bend in the barrel tube. When the pump is lowered into the well, extreme care should be exercised when coming close to the fluid level. If the pump is lowered too rapidly into the fluid, the jarring action can cause the plunger to jam into the barrel or cause some other damage. The pump must be lowered very slowly into the fluid. Once the pump is below the fluid level, it will not fall as rapidly in the fluid as it did in the open tubing. The speed of lowering, therefore, must be slowed down until the pump is seated in the seating nipple. It is necessary to approach the seating nipple very slowly. After the pump is settled in the seating nipple, the pump must be jarred two or three times to insure that the hold-down is engaged. Great care must be exercised not to jar the pump so hard as to cause mechanical damage. During initial operation of the pump, it is necessary to watch and listen for any unusual signs of improper operation. It must be spaced properly and should not pound either at the top or the bottom. It should also be checked periodically to insure that the travelling and standing valves are spaced as close as possible without bumping. If production begins to fall off, the pump should be pulled as soon as possible.
SAMPLE PROBLEMS AND QUESTIONS
(1) Outline the minimum amount of information needed for a sucker-rod pumping unit installation design.
627
(2) Describe in detail the importance of counterbalance effect and the reasons for counterbalancing a pumping unit system. (3) Explain the torque factor for a beam pumping system and how one can obtain torque from a dynamometer card analysis. (4) List the typical problems in a downhole pump performance, which could be observed (or detected) by analyzing a dynamometer card profile. ( 5 ) Describz briefly the following: net lift, effective plunger stroke, plunger overtravel, cyclic load factor, dynagraph, fluid pound, and gas lock. (6) A pump having If-in. plunger is to be set at 6550 ft using a two-way tapered (API Rod No. 76) rod string. The stroke length is 64 in. and the average specific gravity of the well fluid is 0.83. The tubing is anchored and the pump is to be set at the fluid level. Determine the following: (a) Length of each system of the tapered rod string. (b) The pumping speed at which polished rod stroke length and effective plunger stroke length will be equal. (7) An airbalanced pumping unit is producing 750 bbl/D of fluid (550 bbl/D of oil and 200 bbl/D of water). The oil specific gravity is equal to 0.80. The well depth is 6500 f t and the pump is set at a depth of 5600 ft. The plunger, having a diameter of 2.5 in., is to be set in a 3.5-in. anchored tubing. The rod utilizes a rod string consisting of li-in., 1-in. and ;-in. rods and operates at 10 spm with a surface stroke of 168 in. Calculate: (a) rod stretch, (b) tubing stretch, (c) plunger overtravel, and (d) effective plunger stroke. (8) A certain pumping unit is producing 600 bbl/D of a fluid having a specific gravity of 0.90. Pump setting depth is 5100 ft and pumping fluid level is 3200 ft. The tubing pressure is 60 psi. Determine: (a) the net lift of the fluid, and (b) the overall efficiency of the pumping system, if the fuel requirement for the prime mover is 4 Mcf/D of a gas having a net heating value of 1000 Btu/scf. (9) Design a sucker-rod pumping system given the following information: Fluid level = 5000 f t Pump depth = 5000 f t Pumping speed = 16.3 spm Stroke length = 64 in. Plunger diameter = 1: in. Fluid sp. gr. = 0.86 API rod no. = 64 Tubing size (unanchored) = 2 in. Production = 200 bbl/D Maximum allowable stress = 26,000 psi
628
APPENDIX 14.1- -USEFUL FORMULAS (COURTESY O F LUFKIN INDUSTRIES. INC.) BELT LENGTH
STROKES PER MINUTE S M pX d= & !?
R
PL=ZCD-l57(D*d)--
D
Example:
Example
RPM = 1170 Revolutions per minute of prime mover R = 30.12 (320D Gear Reducer) d = 12" Pitch Diameter of Prime Mover Sheave D = 47" Pitch Diameter of Gear Reducer Sheave SPM =
~
4 X CD
CD = 65 43 Inch Center Distance of Shafts D = 47 Inch Pitch Diameter of Gear Reducer Sheave d = 14 5 Inch Pitch Diameter of Prime Mover Sheave
12 1170 X- = 9.9 30.12 47
PL = 2 X 65.43 I 1.57 (47 -- 14.5)
PRIME MOVER SHEAVE DIAMETER
-
PL = 231 45 Inches
d = SPM X R X D RPM
use C225 or D225 Belts Depending on Sheaves Selected
Example: HORSEPOWER OF PRIME MOVER
SPM = 12 Strokes Per Minute R = 30.12 Ratio (320D Gear Reducer) D = 47" Pitch Diameter of Gear Reducer Sheave RPM = 1170 Revolutions Per Minute of Prime Mover
Foi High Slip Electric Motors and Slow Speed Engines H P = BPD X Depth 56000 t o r Normal Slip Electric Motors and Multi cylinder Engines
Use nearest size available depending upon belt section and number of grooves in sheave
H P = BPD X Depth 45000
BELT VELOCITY .i;
V =
Multiply HP by 0 8 for Mark II Units
X d X RPM
.12_
Example'
Limit Between 2000 and 5000 feet per min Belt Velocity less than 2000 FPM results in poor belt life Belt Velocity greater than 5000 FPM requires dynamically balanced sheaves
o efficiency BPD = 217 @ 1 0 0 ~ pump Depth = 5600 Feet pump setting Assume High Slip (Nema D) Motor)
Example:
Maximum Strokes Per Minute Based on the Free Fall Speed of the Rod
v = 3.1416 X 14.5 X 1170 =4441 FPM
12
Conventional Units
CENTER DISTANCE CD = also =
J( U +p)'
J(
UU
SPM = . 7 T
(AB - b)'
d q
Air Balanced Units
- q)' + (AA - b)'
SPM= 63 Mark II Units
Example: Assume Hi.Prime Electric Motor Driven C.320D.256 100 Conventional Unit
SPM= 56
UU = 30.375 (See General Dimensions) W = 34.25 (See General Dimensions) (See General Dimensions) AA = 5 3 b=8 (Assume 25-HP, Frame 324T Motor) CD =
= 21.7. use 25 HP Motoi
HP =
d = 14.5 Inch Pitch Diameter RPM = 1170 Revolutions per minute of Prime Mover
d(
30.375
JF __ d F
Example Assume C 320D 256 100 Unit
+ y ) '(53 - 8)'
SPM = 7
Jg
= 17 15 SPM Maximum
C D = 65.43 Inches DEFINITION OF SYMBOLS IJSED SPM = RPM = R= D= d=
Strokes Per Minute Revolutions Per Minute of Prime Mover Gear Reducer Ratio Gear Reducer Sheave Pitch Diameter, Inches Prime Mover Sheave Pitch Diameter. Inches v = Belt Velocity, Feet per Minute
-. = . ~ -2 . 1-A. l-f i-
fPil
PL = Belt P i t i h 'Length, Inches CD = Shaft Center Distance, Inches U = See General Dimensions
V= AB = UU = W= AA = b=
See General Dimensions
See General Dimensions Sec General Dimensions See General Dimensions See General Dimensions Prime Mover Backing (Vertical Distance from Mounting Feet to Center t o Shaft). In HP = Horsepower BPD = Barrels Per Day at 100% Pump Efficiency Depth = Pump Setting. Feet L = Stroke Length, Inches
629
UNIT DESIGN CALCULATIOKS (COURTESY OF LUFKIN IN-
APPENDIX 14.11-PUMPING DUSTRIES, INC.)
PUMPING UNIT DESIGN CALCULATIONS
FO88-C
Ro.1
Well Name:
Comp.nv:
h e :
Count+
Fielw Ro -in-:
BBL'S/Dw
I"4
Plvnpr um,:
..Fluid Grivity 1.0
- Tubing Size:
Inches
Inches
-
state:
8.650
Pump Dmpth
06
Rod Sirs:
-
Ft. -Stroke Length 1 6 8 l n d m
7.6
Pumping Speed
SPM
A L L TYPES OF UNITS 1.
F
2
SKR = 1000 xStrokeilEr.Table2 x Depth) = 10W
3.
FolSKR =
4.
N/No = SPM x Depth + 245000
5.
N/No'=
6.
BPD (100% sff.1
7.
WRF
8.
WRFISKR
9.
o
-
~
p
~
h
~
~
~
I
9.005
lN/Nd+Fe. Table 2 =
-
G
-
"
.+
=
i
d
L
o
~ x,
160
x
27.746 7.6
.260
i
-
1.041 ~ a b I ~ l 9.005 - ~ ~ +I.OOO~1-x
.325
=
0.650 1-230 ,357 x 7.6
Pump Conrt. Table 1 x SPM x Stroke x SP. Table 3
-
168
x
x
E4.128 x GI 2.186 x 8.650. ~ . ( . 1 2 8x L d 16.481 + 27.746 = .594 T A = l + F , T n b i s 7 x l ~ - . 3 1 x 1 6 j = l ~ - . x~( 7.594 5 -.3)x 101.978 Rod Wniaht. Tabla 2 x Depth x
-
=
+ 245,000
x
I. I64
~
- 27.746
I
.268
,771 . = 351 16,481
CONVENTIONAL UNITS 10
PPRL = WRF + IF(, Table 4 x S K R l = _ 1 6 . 4 8 1
11
MPRL = WRF-
12
C8L = 1.06 x IWRF
13
PT = T. Table6x SKR x Stroke12 x T A =
14
Rod Stress
15
PPRL = WRF + Fo
-
+I'
16.481 F d 2 I = 1.06 x ( 16.481
(F2,Table5xSKRI =
+
PPRL-Area. Table 8 =
.
-1
x
27.746
I
x
27.746
I=
I
t
x
30, 270
.497
.177
+
27.746 .785
x
=
=
=
30.270 11.570
22.243 -978 x
- 793,200
38.561
AIR BALANCED UNITS
-
+
.85x(Fl,TabIe4 x S K R - F d = M + % ! ? &
.85 x
l&x21.746-9.0061=
29.553
* I77 I x 27.746 = 10.652 10.852 l+2= 21.415 Stroke12 x T A x .96 = .340 x 27.746 x 04 x .978 x .96= 761. 500 Table 8 = 29.353 + .705 37.647
PPRL- IF1, Table 4 t FZ. Table 51 x SKR =
16
MPRL
17
CBL=l.06xlPPRLtMPRLl-2=1.06
18
PT=T.Table6 x SKR x
19
Rod Srrerr = PPRL-Area,
x I
29.553-@l+
29.553
+
-
MARK II UNITS
24
16.481 + 9.00s + . 7 5 X ( ~497x Z ? ; H ~ % O O ~ I -29.075 29.075 - 1 -497 + .I17 1 x 27.746 = 10.374 CBL=1.WxlPPRL+1.25xMPRLI12=1.Wxl 29.075+ 1.25 x 10.374 I + 2 = 21.862 612. 800 PT (PPRL x .93-MPRLx1.21 x S t r o k e i 4 = l Z9.07Sx.93 - 10.374 x1.2) x & t 4 = Rodsirerr = PPRL+Area,TableS = 29.075 i -785 = 37.038
25.
NOTE
20
I21
22 23
PPRL = W R F t F o t . 7 5 x l F l , T a b l e 4 x S K R - F o f = MPRL=PPRL--IFl,Table4tFZ.Table5IxSKR=
-
00 Not Use Lerr Than One Sire Smaller Reducer Than Requlred For Conventional Unit
630 BRAKE HORSEPOWER REOUIRED BASED ON 100% VOLUMETRIC EFFICIENCY:
Pap. 2
Conventional and Air Balanwd Unitr.
Mark II Units
For slow speed engines E high slip elnctric motors
For slow s p o d engines E high slip nlnctric motors
56,000
56.000
For multi-cylinder engines E normal slip nlectric motors
For multi-cylinder engines & normal slip electric motors
Depth 8650 Ft. x EPD
351
=
68
BHP
Dnpth
8.650
4 5 . m
Ft. x EPD
35I
x .8 -&BHP
45.000 TABLE 1 Fluid Lord Ib. per ft.
EXPLANATION OF SYMBOLS Fo = Fluid Load on Full Plunger A r w Plunger SKR = Load r e q u i r d to strntch thn rod string t o an amount equal to the stroke Iongth Din. FolSKR Percent of the rtroka length which thm fluid load wlll stretch tho rod string 1.1116 NINo Ratio of SPM t o natural frequancy of straight rod string 1.114 N/No' Ratio of SPM t o natural frnqwncy of tapered rod string BPD = Barrels per day production a t 100% volumetric dficisncy 1.112 --C 1-314 WRF = Weight of rod string i n fluid 2 TA Torque adjustmnnt for pnek torqua for values of WRFISKR other then .3 2.114 PPRL = Pnak polishd r o d load, poundr 2-112 MPRL = Minimum polished rod lord, pounds 2-3/4 CEL Counterbalanm required, pounds 33/4 PT = Pnek reducer torque, inch pounds 4314 Wr Average Weight of rods i n air, pounds par foot ROD AND PUMP DATA G Spncific Gravity of p r o d u d fluid TABLE 2
-- -
0.384
Rod Wt. Ib. per ft. Wr
Elastic Constant Er
Frequency Factor Fe
44
All
0.726
.oar99
1.Ooo
54 54 54 54 54
1.06 1.25 1.50 1.75 2.00
0.908 0.929 0.657 0.990 1.027
,00167 ,00163 .00156 .00153 .00146
1.138 1.140 1.137 1.122 1.095
55
All
1.135
.00127
1.000
64 64 64 64
1.06 1.25 1.50 1.75
1.164 1.211 1.275 1.341
,00138 .00132 DO123 ,00114
1.229 1.215 1.184 1.145
33.3 37.2 42.3 47.4
33.1 95.S 40.4 45.2
65
65 65 65 65
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75
1.307 1.321 1.343 1.369 1.394 1.426 1.460 1.497
,00114 .00113 .00111 .w109 .00107 .00105 .00102
.oO099
1.098 1.104 1.110 1.114 1.114 1.110 1.098 1.082
34.4 37.3 41.8 46.9 52.0 58.4 86.2 72.5
65.6 62.7 58.2 53.1 48.0 41.6 34.8 27.5
66
All
1.634
.M)o88
1
1
1.566 1.604 1.664 1.732 1.603 1.875
.00100 .00087 .00094 .00089 ,00085 .O0080
1.191 1.193
76 76 76 76 76 76 76 76 76 76
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.25 3.75
1.602 1.614 1a 3 3 1.855 1.880 -1.808 1.934 1.967 2.039 2.119
.00082 ,00061 .OM)80 .OWSO .00079 ,00077 .OOO76 .w75 ,00072 .OOO69
1.072 1.077 1.082 1.088 1.093 1.096 1.097 1.094 1.078 1. w 7
i.im
718
3/4
518
1I 2 100.0
44.6 49.5 56.4 64.6 73.7
66.4 50.5 43.6 95.4 26.3
100.0
.m
1.06 1.25 1.50 1.75 2.00 2.25
0.466 0.590 0.728 0.881 1.840 2.630
Rod String, % of Each Sire
Plunger Die.
75 75 75 75 75 75
&
1.360 1.721 2.126 2.571 4.781 7.671
Rod No.
66 65
0.19 0.162
g
--
65
Pump Constant
I
----..I
1.174 1.151 1.121 --.-I
33.5 28.9 11.3 7.4
100.0
27.0 29.4 33.3 37.8 42.4 46.9
27.4 29.8 33.3 37.0 41.3 45.8
28.5 30.6 33.8 37.5 41.7 46.5 €4.6 66.5 88.7 82.3
71.6 69.4 66.2 62.5 58.3 53.5 49.2 43.5 31.3 17.7
46.6 40.8 33.3 25.1 16.3 7.2
-I
I
I -
I
I -
631 TABLE 2 (Continudl
Rod No.
Plunger Din.
77
All
2.224
,00065
1.000
85
85 85 85
1.06 1.25 1.50 1.75
1.883 1.943 2.039 2.138
,00087 .00084 ,00079 ,00074
1.261 1.253 1.232 1.201
_........
96 86 86 86 86 86 86 86
1.06 1.25 1.50 1.75 2.00 2.25 2.50 2.75
1.151 1.156
-........ ..........
87 87 87 87 87 87 87 87
+
Elastic Constant E,
Frequency Factor Fe
1-114
....
>is 2.658
,00074
paw 3
Rod String. % of Each Sizm
Rod Wt. Ib.per ft. W,
..........
1.118
....
..........
1
718
314
100.0
--
22.2 23.8 26.7 29.6
22.4 24.2 27.4 30.4
22.4 24.3 28.8 29.5
22.6 24.3 26.8 29.4 32.8 36.9 40.6 44.5
23.0 24.5 27.0 30.0 33.2 39.7 43.3
54.3 61.2 46.3 40.6 33.8 27.1 19.7 12.2
27.7 30.3 33.2 36.4 39.9 43.9 61.2 83.6
72.3 68.7 66.8 63.6 60.1 56.1 38.8 16.4
2.247
.00068
2.315 2.385 2.455
,00066 ,00063 ,00061
1.161 1.153 1.138 1.119
1S O 1.75 2.00 2.25 2.50 2.75 3.75 4.75
2.413 2.430 2.450 2.472 2.496 2.523 2.641 2.793
,00061 .00060 .00060 ,00059 ,00059 ,00058 .00056 ,00052
1.062 1.066 1.071 1.075 1.079 1.082 1.078 1.038
88
All
2.904
.00050
1.000
..........
96 96 96 96 96 96
1.06 1.25 1.50 1.75 2.00 2.25
2.382 2.435 2.511 2.507 2.703 2.806
.00067 ,00066 .00063 ,00061 ,00058 .00055
1.222 1.224 1.223 1.213 1.196 1.172
.......... ........ ........
19.1 20.5 22.4 24.8 27.1 29.6
18.2 20.6 22.5 25.1 27.9 30.7
18.5 20.7 22.8 25.1 27.4 29.8
97 97 97 97 97 97
1.50 1.75 2.00 2.25 2.50 2.75
2.707 2.751 2.801 2.856 2.921 2.988
.OW58 ,00055 ,00054 ,00053 ,00052 ,00050
1.131 1.137 1.141 1.143 1.141 1.135
........ .........
22.5 24.5 28.8 28.4 32.5 38.1
23.0 25.0 27.4 30.2 33.1
54.5 50.4 46.7 40.4 34.4 28.6
98 98 98 98 98 98 98
1.75 2.00 2.25 2.50 2.75 3.75 4.75
3.103 3.118 3.137 3.157 3.180 3.289 3.412
,00047 ,00047 ,00047 ,00046 .00046 ,00045 .OW43
1.051 1.065 1.058 1.062 1.066 1.074 1.064
..........
25.7 27.7 30.1 32.7 35.8 49.7 65.7
74.3 72.3 68.9 67.3 84.4 50.3 34.3
99
All
3.676
.wa39
1.OOo
....
100.0
107 107 107 107 107 107
1.50 1.75 2.00 2.25 2.50 2.75
3.085 3.158 3.238 3.338 3.435 3.537
.00051 ,00049 .00048 ,00046 ,00045 ,00043
1.195 1.197 1.195 1.187 1.174 1.156
19.4 21.0 22.7 25.0 26.9 29.1
19.2 21.0 22.8 25.0 27.7 30.2
19.5 21.2 23.1 25.0 27.1 29.3
41.9 36.9 31.4 25.0 18.2 11.3
108 108 108 109 108 108
1.75 2.00 2.25 2.50 2.75 3.75
3.41 1 3.452 3.498 3.548 3.603 3.873
.wo44
1.111 1.117 1.121 1.124 1.126 1.108
20.9 22.6 24.5 26.5 28.7 40.6
21.4 23.0 26.0 27.2 29.6 39.5
57.7 54.3 5a5 46.3 41.6 19.8
--
109 109 109 109
2.50 2.75 .3.75 4.75
3.91 1 3.930 4.020 '4.120
,W037 .OW37
1.048 1.051 1.063 1.066
27.2 29.4 39.9 51.5
72.8 70.6 60.1 48.5
,00043 ,00043 ,00042 ,00042 ,00038
.00036 ,00035
........
_ ........
..........
..........
..........
36.0
33.0 27.6 19.2 10.5
I -
--.
-.
100.0
36.3
-.--
518
42.3 38.3 32.3 25.1 17.6 9.8
--
-
I -
--
I -
632 -4
TABLE 5. FZ, MIN. POLISHED ROD L O A D FolSKR
N!No
128
.3 n. ,019 ,045 08 ,125 154 ,165
.4 ,5 n o . . 015 ,022 ,039 .05 073 083 12 ,119 ,161 ,158
.20(>.r
TABLE 8.
.6 o ,025 ,055
,086
+@
,195-
,241 235 ,235 275 .27 ,263 2 7 3 ,306 .3w 29 364 .35 ,339 .34 433 ,413 ,384 .38 ,415 .45 .42 .41
Area
RodSile 1 12 518 314
-
718
A18 1.114
SS 1; 0 196 0 307 0 442 0.601
cz? 1.227
633 REFERENCES American Petroleum Institute, 1977. Design Calculations for Sucker Rod Pumping Systems. A.P.I., Dallas, Tex., RP 11L, 24 pp. American Petroleum Institute, 1983. A. P.I. Recommended Practice for Care and Use of Subsurface Pumps. A.P.I., Dallas, Tex., RP l l A R , 41 pp. American Petroleum Institute, 1979. A. P.I. Specifications for Subsurface Sucker Rod Pumps and Fittings. A.P.I., Dallas, Tex., Spec IlAX, 62 pp. American Petroleum Institute, 1969. A.P.I. Catalog of Analog Computer Dynamometer C a r d . A.P.I., Dallas, Tex., Bull., 11L2, 77 pp. American Petroleum Institute, 1970. A.P.I. Bulletin of Sucker Rod Pumping System Design Book. A.P.I., Dallas, Tex., Bull., l l L 3 , 575 pp. Atlantic Richfield Company, 1984. Artificial Lift Sucker Rod Pumping Manual (A collection of notes and discussion). ARCO, Houston, Tex., 437 pp. Axelson, Inc., 1980, 1982. Pump and Rod Engineering Manual. Manual M-63. U S . Industries, Longview, Tex., 92 pp. Bethlehem Steel Co., 1950. Sucker Rod Handbook. Handbook 282, 213 pp. B o w e r , P.M., 1981. Sucker rod pumping systems design -another look. 56th Annu. Fall Tech. Conf. Exhibition, SOC.Pet. Eng. of AIME, San Antonio, Tex., Oct. 5-7, 6 pp. Brown, K.E., Day, J.J. Byrd, J.P. and Mach, J., 1980. The Technology of Artificial Lift Methods, Vol. 2a. PennWell, Tulsa, Okla., 720 pp. Byrd, J.P. and Beasly, W.L., 1974. Predicting prime mover requirements, power costs, and electrical demand for beam pumping units. 25th Annu. Fall Meet., Pet. SOC.of C.I.M., Calgary, PSCIM 374035. Byrd, J.P., 1977. Pumping deep wells with a beam and sucker rod system. In: Deep Drilling and Production Symp. Soc. Pet. Eng. of A.I.M.E., Amarillo, Tex., SF'E 6436, 37 pp. Byrd, J.P., 1966. Polished rod dynamometers and the construction of a theoretical dynamometer card. Presented at the Deep Well Drilling, Production and Maintenance Institute, sponsored by the Univ. of Kansas, April 19, 28 pp. Byrd, J.P., 1971. A technique for measuring and evaluating the performance of a beam pumping system. 43rd Annu. Conf. Pet. Electr. Power Assoc., Houston, Tex., 6 pp. Byrd, J.P., 1983. The Beam and Sucker Rod Pumping System. (A collection of memos and papers dealing with certain aspects of beam driven, sucker rod pumping system.) Lufkin Industries, Tex., 203 pp. Craft, B.C., Holden,,W.R. and Graves Jr., E.D., 1962. Well Design: Drilling and Production. Prentice-Hall, Englewood Cliffs, NJ, pp. 286-368. Day, J.J. and Byrd, J.P., 1980. Beam pumping: design and analysis. In: K.E. Brown (Editor), The Technology of Artificial Lift Methoak, Vol. 2a. PennWell, Tulsa, Okla., pp. 9-94. Doty, D.R. and Schmidt, Z., 1981. An improved model for sucker rod pumping. 56th Annu. Fall Tech. Conf. and Exhibition, Soc. Pet. Eng. of A.I.M.E., San Antonio, Tex., Oct. 5-7, 7 pp. Eickmeier, J.R., 1973. Pumping well optimization techniques. Canad. J. Pet. Tech., April-June: 53-63. Eickmeier, J.R., 1967. Diagnostic analysis of dynamometer cards. J. Pet. Tech., 19(1): 97-106. Eubanks, J.M., Franks, B.L., Lawrence, D.K., Maxwell, T.E. and Merryman, C.J., 1958. Pumping Well Problem Analysis. Chastain, Midland, Tex., pp. 155-181. Fagg, L.W., 1950. Dynamometer cards and well weighing. Trans. Soc. Pet. Eng. of A.I.M.E., 189: 165-174. Gibbs, $G., 1977. A general method for predicting rod pumping system performance. 52nd Annu. Fall Meet., Soc. Pet. Eng. of A.I.M.E., Denver, Colo., Oct. 8-11, SPE 6850, 19 pp. Gibbs, S.G., 1963. Predicting the behavior of sucker-rod pumping systems. J . Pet. Tech., 15 (7): 769-778. Griffin, F.D., 1975. A n update on pumping unit sizing as recommended by API-RP-IIL. 26th Annu. Meet. Pet. SOC. C.I.M., Banff, June 11-13, 15 pp. Kelly, H.L. and Willis, R.M., 1954. Manual for selection of beam pumping equipment - part 3. Pet. Eng., Sept.: B77-B97. Lufkin Industries, 1983. Lufkin Pumping Units Catalog No. 82-83. Lufkin Industries, Lufkin, Tex., pp. 25-28.
634 Marsh, H.N. and Watts, E.V., 1938. Practical dynamometer tests. Drill. Prod. Pract. A P I , pp. 162-182. Mills, K.N., 1939. Factors influencing well loads combined in a new formula. Pet. Eng., Apr. Mills, K.N., 1943a. Standardization committee report on pumping equipment and engines (Exhibit A, well-load formulas). Prod. Bull. API., 230:442-462. Mills, K.N., 1943b. Counterbalancing and determining peak loads on pumping units. World Oil, 14: 176-186. Neely, B., Gipson, F., Clegg, J., Capps, B. and Wilson, P., 1981. Selection of artificial lift method. 56th Annu. Fall Tech. Conf. and Exhibition, SOC.Pet. Eng. of A.I.M.E., Oct. 5-7, San Antonio, Tex., SPE 10337, 6 pp. Russel Jr., J.H., 1953. Interpretation of dynamometer cards. World Oil, 24: 175-188. Szilas, A.P., 1975. Production and Transport of Oil and Gas (Developments in Petroleum Science, 3.). Elsevier, Amsterdam, 630 pp. Treadway, R.B. and Focazio, K.R., 1981. Fiberglass sucker r o d - a fururistic solution to today’s problem wells. 56th AMU. Fall Tech. Conf. and Exhibition, SOC.Pet. Eng. of A.I.M.E., Oct. 5-7, San Antonio, Tex., SPE 10251, 4 pp. Zaba, J., 1943. Oil well pumping practices. No. 3-Sucker-rod pumping system. Oil Gas J . , 42: 67-78. Zaba, J., 1943. Oil-well pumping practices. No. 4-Sucker-rod load measurement. Oil Gas J . , 42: 34-39. Zaba, J., 1962. Modern Oil Well Pumping. Petroleum Publishing, Tulsa, Okla., 818 pp. Zaba, J. and Doherty, W.T., 1956. Practical Petroleum Engineers Handbook. Gulf Publishing, Houston, Tex., 4th Ed., pp. 487-542.
635
Chapter 15
HYDRAULIC SUBSURFACE PUMPS
PHIL WILSON and GEORGE V. CHILINGARIAN (EDITOR) INTRODUCTION
Subsurface hydraulic pumps are downhole pumps powered by high-pressure fluid directed to them from the surface. There are two types in common use: (1) the positive displacement piston pump, driven by a hydraulic reciprocating subsurface
-FLUID LEVEL
s
-
ENGINE
PUMP
Fig. 15-1. Schematic diagram of piston pump, free type
636
SHUT OFF AND BLEED
PUMP IN
fi
0PERAT E
A
PUMP OUT
fi
I
FLOW LINE
J t STANDING IA LV E :LOSED
STANDING VALVE
CLOSED
Fig. 15-2. Schematic diagrams showing operation of parallel free pump.
engine directly connected to the pump, and (2) the jet pump. Only the piston-type hydraulic pump is discussed here. Figure 15-1 shows a typical hydraulic subsurface pump installation. Two strings of tubing are shown: (1) for directing the high-pressure power fluid to the engine, and (2) for conducting the return power fluid plus produced fluid to the surface. In this type of installation, the pump is free to be circulated in and out of the well. T h s “free pump” principle is a unique feature of subsurface hydraulic pumps and is further illustrated in Fig. 15-2. The “shut-off and bleed” function of the “free pump” system is shown at the left in Fig. 15-2. The U tube is full of low-pressure power fluid, which is retained in the tubing by the closed standing valve. The power fluid is shut-off at the surface and the large tubing is opened at the top allowing the “free pump” to be inserted. The next function shown is “pump in”, whereby the pump, with its packer and nose assembly, is being circulated to bottom. The large tubing is closed at the top by the insertion of the pump catcher which locks in place. The standing valve is still closed and the 4-way valve at the surface is positioned to direct power fluid into the large tubing and the return fluid into the flowline. The flow of power fluid down the large tubing carries the pump to the bottom.
631
6
8 Tubing
Fig. 15-3.Complete hydraulic pumping system.
-
638
The “operate” function is shown next: The pump is seating at the bottom, metal-to-metal, on the standing valve and at the top by an O-ring in a seal collar. The flow of power fluid down the large tubing is operating the engine, while the production, mixed with exhausted power fluid, is returning up the small tubing. The standing valve is now open and its ball is being held off seat by a magnet. The “pump out” function is shown at the right in Fig. 15-2: The power fluid is directed down the small tubing, by the surface 4-way valve, thereby forcing the pump off of the standing valve. When the pump unseats, the rush of fluid trying to exit forces the ball off of the magnet and on the seat. With the standing valve now closed, the U tube necessary to surface the pump is provided. The swab cups of the packer and nose assembly provide the efficiency necessary to circulate the pump to the surface, where the fishing neck of the assembly engages the pump catcher. With the pump latched, the tubing strings can be bled off and the catcher used to remove the free pump. A complete hydraulic “free pump” system is shown in Fig. 15-3. The various elements of this system are discussed in the order shown from A through F.
SUBSURFACE PUMPS-PISTON TYPE
Originally, the subsurface pump was designated “production unit”, but in practice it was usually referred to as a pump. This caused confusion because the unit consists of an engine as well as a pump, and it was continuously necessary to refer to the operation of the engine separately from the operation of the pump. The problem was solved by identifying the engine as the “engine-end’’ of the pump and the pump as the “pump-end’’ of the pump. These commonly accepted terms are used in the following discussion. The engine-end of a Kobe type A pump is shown in Fig. 15-4. High-pressure power fluid enters from the top and is shown being directed to the top of the piston. Exhausted power fluid from the lower side of the piston is being directed by the relieved area of the engine valve to the discharge port. When the piston reaches the end of the downstroke, the reduced diameter at the top of the valve rod allows high-pressure fluid to enter under the valve as shown in Fig. 15-5. Inasmuch as the valve has a larger area at its bottom than at its top, the high pressure causes it to move upwards. With the engine valve in the up position as shown in Fig. 15-6, the flow paths to the piston are reversed. The piston, therefore, begins the upstroke. When the piston reaches the end of the upstroke, as shown in Fig. 15-7, the reduced diameter near the lower end of the valve rod connects the area under the valve to the discharge, or low-pressure, side of the engine. With high pressure on the top of the valve and only exhaust pressure at the bottom, the valve will move to its down position and the cycle will be repeated. The rate of flow of power fluid determines the pump speed. Pump speed is measured in cycles per minute and is referred to as “strokes per minute”.
639 DOWNSTROKE
END OF DOWNSTROKE
ENGINE VALVE,
-
VALVE ROD
ENGINE PISTON
-
-
MIDDLE ROD CONNECTS TO PUMP PISTON
Fig. 15-4.Cross-section of engine-end: downstroke. Fig. 15-5. Cross-section of engine-end; end of downstroke. START OF UPSTROKE
Fig. 15-6. Cross-section of engine-end: start of upstroke. Fig. 15-7. Cross-section of engine-end; end of upstroke.
END OF UPSTROKE
640
DOWNSTROKE
EXHAUST ,VALVES
Fig. 15-8.Cross-section of pump-end; downstroke. UPSTROKE
Fig. 15-9. Cross-section of pump-end; upstroke.
641 DOWN STROKE
UP STROKE
-
ENGINE PISTON-
-
PUMP PISTON-
Fig. 15-10. Cross-section of complete pump: downstroke and upstroke.
The pump-end of the unit is shown in Fig. 15-8 making a downstroke. This pump is double-acting, i.e., it pumps on the upstroke and on the downstroke. The arrows show that well fluid is entering on the left and filling the upper part of the cylinder, while the well fluid below the piston is being discharged through the ball check valve at the lower right. On the upstroke, shown in Fig. 15-9, well fluid is filling the lower part of the cylinder and is being discharged from the upper end. The complete pump is shown in Fig. 15-10. The hollow rods serve two purposes: (1) to balance the areas, and hence the forces, on the upstroke and downstroke; and (2) to provide clean high pressure power fluid to the pump piston for lubrication and for washing sand out of the piston-cylinder fit. The Kobe type A pump is also shown in Fig. 15-11 along with a pump specification table. Figures 15-12 through 15-24 show other pumps currently avail-
642
able from the various manufacturers. Some of these pump schematics are difficult to analyze because of the various internal passageways and engine valve designs. A simplified approach is shown in Fig. 15-25. T h s is the Kobe type A pump previously covered. The engine valve, pump valves, and fluid passageways are all shown external to the piston and cylinder. As a further simplification, the engine valve is schematically illustrated as a rotating valve. Figure 15-26 shows all of the various engine-ends and pump-ends in this simplified format. The engine-ends at A , B , and C use a 4-way, 2-position, engine valve, whereas those at D, E , F, and G use a 3-way, 2-position, engine valve. A 3-way valve requires the pressure in one end of the cylinder to be the same on the upstroke and on the downstroke. The pressures for the upstroke are shown in the left half of the cylinders and the pressures for the downstroke are shown in the right half of the cylinders. In some of the pump-ends, the pump valves must be shown internally. The Kobe type A single pump-end pump (Fig. 15-11) is illustrated at A and H in Fig. 15-26. The engine valve is at the top of the pump and is shifted hydraulically. The top rod (called valve rod) acts as a pilot valve for shifting the main valve. This pump is double acting, i.e., it pumps on the upstroke and on the downstroke. The Kobe type A double pump-end pump (Fig. 15-12) is illustrated at A and I in Fig. 15-26. It is similar to the pump in Fig. 15-11 except it has an added pump piston. This piston is in parallel hydraulically with the other pump-end piston so it doubles the pump-end volume capacity. The Kobe type B single pump-end pump (Fig. 15-13) is illustrated at A and H in Fig. 15-26. It is similar to the type A pump except some passageways that are internal in the type A pump are external to the type B pump. This allows the type B cylinders to be larger than the type A cylinders. Hence the type B pump has greater capacity. The Kobe type B double pump-end pump (Fig. 15-14) is illustrated at A and I in Fig. 15-26. The comments regarding Figs. 15-11, 15-12, and 15-13 are applicable to this pump also. The Kobe type D single pump-end pump (Fig. 15-15) is illustrated at B and H in Fig. 15-26. It has two engine pistons (in parallel hydraulically) for greater engine-end capacity. The previous comments are applicable to this pump. The Kobe type D double pump-end pump (Fig. 15-16) is illustrated at B and I in Fig. 15-26. The previous comments are applicable to this pump also. The Kobe type E pump (Fig. 15-17) is illustrated at C and J in Fig. 15-26. The engine valve is in the middle of this pump and is shifted hydraulically. The rod acts as a pilot valve for the main engine valve. It is somewhat awkward to refer to the engine-end and the pump-end of this pump because the engine is in the middle. The inboard ends of the pistons are the engine and the outboard ends are the pump. More descriptive terms would be “upper and lower engines” and “upper and lower pumps” or “ upper and lower pump-ends”. The Guiberson Powerlift I1 pump (Fig. 15-18) is illustrated at C and J in Fig. 15-26 and is similar to the Kobe type E pump.
643
SPECIFICATIONSKOBE TYPE A PUMP-SINGLE ENGINE, SINGLE PUMP END DI
,............... 2x1-llle 2x11~-1
.............' . . . . . . . . . . . . .I
1546 647
LACEMENT EPD Der SPM
Eng.
Pump
Max. Rated Speed (SPW
139 254 393 254 393
2 15 2 15 2 15 3 30 3 30
115 2 10 3 25 2 10 3 25
121 121 121 121 121
256 367 492 703 492 703
5 02 5 02 5 02 5 02 7 13 7 13
2 56 3 67 4 92 7 03 4 92 7 03
100 100 100 100 100 100
486 646 821 1218 821 1218
9 61 9.61 9 61 9 61 14 17 14.17
5 49 7 43 9 44 14 00 9 44 14 00
87 87 87 87 87 87
1108 1617 2502 1617 2502
21 44 21 44 21 44 32 94 32 94
14 40 21 00 32 50 21 00 32 50
77 77 77 77 77
At Rated Speed
Fig. 15-11. Specifications and cross-section of a Kobe type A pump; single-engine, single-pump-end,
644 ..j:j
:j.
:;
@fl :j :.: :.: (I
....
':
,
4
PUMP SIZE
At
DESCRIPTION
. . .. . . .. . . ........ ......
2 x 1%. -1 x 1 2xl%.-l%. X l z ~ i ~ ~ - -x li hv. ~ ~
.. . . . . , ,
3 x 134-1 % x 1% 3 x 1%--1% x 1% , , .:.;.:
.. . .., ........ . ....... 4 x 2'18-2 x 1% . . . . . . . . . . I 4 x 2W-2 x 2 . . . . . . . . . . . . 4 x 2%-2H x 2 . . . . . . . . . . 4x2%-2#x2h ........ 3xl%-l%xl'h 31 l%-l% x 1%
~
1
DISPLACEMENT BPD per SPM
RATIO
Speed
1.290 1.647 2.000
508 647 786
3 30 3.30
972
14.17 14.17 114.17 14 17
,800 1.351 1.675 2 000 1 094 1.299 1.850 2.000
~
lzli
2436
I
2725 3234 4119 5005
Max. Rated
Eng.
3.30
, 1
32 94 32 94 32.94 32 94
11.18 18.88 23.44 28.00 35.40 42.00 53.50 65.00
Fig. 15-12. Specifications and cross-section of a Kobe type A pump; single-engine, double-pump-end.
645
I
SPECIFICATIONSKOBE TYPE B PUMP-SINGLE ENGINE, SINGLE PUMP END
i
PUMPSIZE DESCRIPTION
I RATIO
At Raled Speed
DISPLACEMENT 1 BPDperSPM
544
4 54
4.50
744 1086
10 96
7 44
i
Max.
Rated Speed
100 100
Fig. 15-13. Specifications and cross-section of a Kobe type B pump: single-engine, single-pump-end.
1
646
SPECIFICATIONSKOBE TYPE B PUMP-SINGLE ENGINE, DOUBLE PUMP END DISPLACEMENT BPD p e r SPM
Max.
Rated
PIE RATIO
At Ribd Speed
...... ,....... ........ . . .. . . ... ...
.......
1.380 1.680 1.980 1.336 1.652 1.957
751 913 1076 1452 1794 2136
4.54 4.54 4 54 10.96 1096 10.96
14.52 17.94 21.36
121 121 121 100 100 100
.. . , . , , ........
1.454 1714 1.974
2726 3213 3700
21.75 21.75 21 75
31.34 36.94 42.53
87 87 a7
PUMP SIZE OR DESCRIPTION 2 x 1%-13% x 11%~ 2xl%-l~hxl~h 2X1~l~-I~!~x1"~ 2% x 1%-1% x 1 % . 2 % x 1%-1 114 x 1 'h , 2%x1~f4-lhxl~l, 3 x Z'h-178 x 1'18 , 3x2'4-2'hxIh 3~2'/~-2'hx2'/1
........
Eng.
Pump 6.21 7.55
8.W
Speed (SPM)
Fig. 15-14. Specifications and cross-section of a Kobe type B pump; single-engine, double-pump-end.
641
PUMP SIZE OR DESCRIPTION 2 x 1%. x l 3 h - l a ~ e Z x l S i i xl%--l'h
2'hXl'li. X l ' A - l ' h 2% x 1',e x I a A - I a A 3xl%x2%-1% 3 x 1% x 2%--2'h
... . .
....... .....
. .. . . ........ .. . ... ..
PIE RATIO
At Rated Sped
DISPLACEMENT BPD por SPM En@.
Pump
Max. Rated Speed (SPM)
407 581
381 544
7 79 7 79
3 15 4 50
121 121
411 608
744 1086
17 99 1799
7 44 1086
100 100
449 606
1357 1874
35 74 35 74
15 96 21 55
87 87
-
Fig. 15-15. Specifications and cross-section of a Kobe type D pump; double-engine, single-pump-end.
648
SPECIFICATIONSKOBE TYPE D PUMP-DOUBLE ENGINE, DOUBLE PUMP END PUMP SIZE OR DESCRIPTION
PIE RATIO
At Rated Speed
DISPLACEMENT BPD per SPM Eng.
Pump
Max. Rated Speed (SPM)
2 X 1 % 1 xI%-l%. Xl'h. 2 X I % # xl%-1% XlYw 2x1'111 x l % - l W x l %
,802 ,976 1,150
751 913 1076
7.79 7.79 7.79
6.21 7.55 8.90
121 121 121
2%X1%1 XI%-l%XlW. 2%xl%. x l U - l % x l ' l ~ , 2112 x IV?, x 1v+1*1, x 1 31,
,813 ,976 1.196
1452 1794 2136
17.99 17.99 17.99
14.52 17.94 21.36
100
3x1~/4x2H-I%x1% 3 x 1 % x 2'/1-2% x 1% 3x1Ux2%-2~/~x2%
,882 1.039 1.197
2726 3213 3700
35.74 35.74 35.74
31.34 36.94 42.53
87 07 07
. ..
. ... . .. ...
100 100
Fig. 15-16. Specifications and cross-section of a Kobe type D pump; double-engine, double-pump-end.
649
I 1
I i
SPECIFICATIONSK 0 6 E TYPE E PUMP DISPLACEMENT EPD per SPM
PUMP SIZE
Rated DESCRIPTION
RATIO 2397 4015 ICOUII~LY Kobe Inc --Subsidiary 01 Baker lnlernslional Corp )
Fig. 15-17. Specifications and cross-section of a Kobe type E pump.
Rated Speed
650
1
SPECIFICATIONS-
PUMP SIZE OR DESCRIPTION 2 Il%e.. 2 X f % . .................. 2x 2% X l'h ................. 2'h I l ' h 2% I 1% .................
................ .................. .................
PIE RATIO ,524 ,725 1.147 ,725 503 1 146
At Rated Speed 597 826 1560 1322 918 2500
Fig. 15-18. Specifications and cross-section of a Powerlift JI pump.
DISPLACEMENT BPD per SPM Eng.
Pump
Max. Rated Speed (SPM)
12.10 12 10 26.35 17.69 1769 43 97
5.53 7.65 30.00 12.59 8.74 50
108 106 52 105 105 50
.
651
P
VERLIFT I PUMP
PUMP SIZE OR
I
I
............... .............
2 x 14 x 1 2 x 1% x 1 ' 2
2'2 x 2 x 1"l
............. .............
............. ............. I ............
3 3 3 3
x x x x
272 x 1% 2% x 2
............. 2h x ............ 2 % x 2'h ............ 1 21.4
~
4 ~ 2 x"2 ~. . . ~. . . . . . . . . I 4 ~ 2 " ~ a x 2 h 4X2"/?8 X2'h 4
~
SPECIFICATIONS. SINGLE ENGINE. SINGLE PUMP END
P/E RATIO
ii
1 121
::
~
~
++J
I
j
Eng
. 645
35 35
1508
35
1
30 80 30 80 3080 30 80 30 80
3080
I'
43 71 4371 4371 4371
2142 2798 3541 4371
1% 1311
60 35 6035 60 35
1810
60 35
2798 3541 4371 5290 8035
264 467 547 637 831
1711
.
.
Max Rated
DISPLACEMENT I BPD Der SPM
1508 15 08 1508
312 450 528
643 840 1062
98 121
.57
At Rated ..
Speed
80 93
............ 72 ........... , .89 2% ............
........
1
~
!
' I
I
rCOurtesy Guibernon Division . D i e ~ ~ Industries ei
30 30 30 ~
30
Inc I
Fig . 15.19 . Specifications and cross-section of a Powerlift I pump; single.engine. single.pump.end .
652
SPECIFICATIONSOILMASTER F, FE 6 FEE PUMPS I
j C o ~ r I a l yARMCO-National Production Systems)
Fig. 15-20. Specifications and cross-section of an Oilmaster F, FE, and FEB pumps.
653
SPECIFICATIONSOILMASTER VFR PUMP SINGLE ENGINE, SINGLE PUMP END
-
PUMPSIZE
OR DESCRIPTION VFR201611 VFR201613 VFR 201616 VFR 252015 VFR 252017 VFR 252020 VFR 302424
............ ............ ............
............
............ ............ ............
PIE RATIO
At RIM S W
DISPLACEMENT EPD pmr SPM ~ng.
Max. Rated
Pump
SpHd (SPM)
62 87 1 32
318 444 673
4 24 4 24 4 24
2 12 2 96 4 49
150 150 150
74 1 00 1 32
6% 858 1119
8 89 8 89 8 89
5 25 7 15 9 33
120 120 120
1 28
1612
1299
1344
120
(Courtesy A R M C O - N ~ ~ I Production O~~I System1
Fig. 15-21, Specifications and cross-section of an Oilmaster VFR pump; single-engine, single-pump-end.
654
PIE RATIO
Speed
Eng.
Pump
Max. Rated Speed (SPM)
54 81
444 673
686 6 86
296 449
150 150
41 56 73
630 858 Ill9
I5 16 1516 15 16
525 715 933
120 120 120
PUMP SIZE
OR DESCRIPTION VFR20181813 VFR 20181818 VFR25202015 VFR 25202017 VFR 25202020
.......... .......... .......... .......... ..........
At
DISPLACEMENT BPD per spm
Wed
Fig. 15-22. Splecifications and cross-section of an Oilmaster VFR pump; tandem-engine, single-pump-end,
655
SPECIFICATIONS
-
OILMASTER V-11 PUMP-SINGLE ENGINE, SINGLE PUMP END
PUMP SIZE OR DESCRIPTION V-25-11-118 V-25-11-095 V-25-11-076 V-25-11-061
................... ................... , , ,. . , , , , , , , , ,. , , , ,
...................
P/E RATIO
At Rated Speed
1.18 .95 76 .El
1419 1299 550 550
DISPLACEMENT BPD per SPM
I
Max.
Eng.
Pump
Rated Speed (SPM)
5.33 6.66 5.33 6.66
6.31 6.31 3 93 3.93
225 206 140 140
656
The Guiberson Powerlift I pump (Fig. 15-19) is illustrated at E and K in Fig. 15-26. The engine valve is in the engine plunger and is shifted mechanically at each end of the cylinder. This pump does not use the upper rod (illustrated by dashed lines at E ) . Power fluid is discharged through the (middle) rod. This pump is single acting. The Oilmaster F, F E and FEB pumps (Fig. 15-20) are illustrated at E and L in Fig. 15-26. These pumps also do not use the upper rod. The engine valve is in the engine plunger and is slufted hydraulically when the valve senses “no-flow’’ at the end of each stroke. This pump is double acting. The Oilmaster VFR single engine pump (Fig. 15-21) is illustrated at E and K in Fig. 15-26. The engine valve is at the top of the pump and is shifted hydraulically. The upper rod is the pilot valve for the main valve. This pump is single acting. The Oilmaster VFR tandem engine pump (Fig. 15-22) is illustrated at F and K in Fig. 15-26. This pump is similar to that illustrated in Fig. 15-21 except it has two engine plungers. These are in parallel hydraulically for doubled engine capacity. Comments regarding Fig. 15-21 apply to this pump. The Oilmaster V11 pump (Fig. 15-23) is shown at E and M in Fig. 15-26. The engine valve is at the top of the pump and is shifted hydraulically. The upper rod is the pilot valve for the main valve. This pump is double acting. The Oilmaster V21 tandem engine pump (Fig. 15-24) is shown at G and N in Fig. 15-26. The engine valve is at the top of the pump and is shifted hydraulically. The upper rod is the pilot valve for the main valve. The engine-end of this pump is similar to Fig. 15-26F except that the underside of the bottom engine piston is exposed to produced fluid rather than power fluid. This pump is double acting. It is obvious that the pump-ends at H , I , and J in Fig. 15-26 are double acting because they pump on the upstroke and on the downstroke. The pump-end at K is obviously single acting because it pumps on the upstroke only. The pump-ends at L , M , and N are double acting, but this fact is not apparent because they pump only on the downstroke; they are sometimes referred to as “differential double-acting’’ pumps. A simple way to determine if a pump is single or double acting is to divide the pump-end displacement per SPM (from the specification table ) by the engine-end displacement per SPM. If the result is equal to the value in the P / E column of the specification table, then the pump is double acting (allow 2-3% variation for operation of the engine valve of Kobe pumps). The results will be 20-30% less than P / E for single-acting pumps. This test compares the pressure ratio ( P / E ) to the volume ratio and is a measure of the unproductive work performed by the engine-end. Example problem 15-1
The pump-end displacement per SPM for the first pump in Fig. 1 5 - l l ’ i s 1.15 and the engine-end displacement per SPM is 2.15; therefore: volume ratio
1.15 2.15
= -= 0.535
651
SPECIFICATIONSOILMASTER V-21 PUMP-TANDEM
ENGINE, SINGLE PUMP END DISPLACEMENT BPD per SPM
PUMP SIZE
Rated
Rated DESCRIPTION V-25-21-075
V-25-21-041
................... ................... (Courtesy ARMCO-Nation81
1072 550 550
10 00 8 38 10 00
6.31 3.93 3 93
170 140 140
Production Syslernri
Fig. 15-24. Specifications and cross-section of an Oilmaster V-21 pump; tandem-engine, single-pump-end.
658
IUPSTROKE
-
- DOWNSTROKE
Pl=HIGH PRESSURE POWER F L U I D Pz=EXHAUST PRESSURE POWER FLUID P3=PRODUCTION DISCHARGE PRESSURE P4=PRODUCTION INTAKE PRESSURE
Fig. 15-25. Simplified schematic diagram showing the operation of the Kobe type A pump.
a
c
h
/ I
1 1 - 1
1-1
I
1
1 Y
P
L
7
==T
'
L
659
k Fig. 15-26. Simplified schematic diagrams showing the opcrations of various engine-ends and pump-ends
660
This compares with the P / E value of 0.545. The 2%’ difference is the unproductive power used to shift the engine valve. The pump-end and engine-end displacement values for the first pump in Fig. 15-19 are 6.45 and 15.08, respectively; therefore: 6.45 volume ratio = -- 0.428 15.08 This compares with the P / E value of 0.52. The 21% difference is the unproductive power used to make the downstroke of the pump.
PUMP SELECTION
Selecting a pump for a given well is a simple and straightforward process using the specification tables in Figs. 15-11 through 15-24. One must select a pump (1) that fits the given tubing size, (2) that has sufficient capacity for the well. and (3) that will not operate at excessive pressure. Selection also involves price, brand preference. mechanical feature preference, and delivery and service availability. The “pump size or description” columns of the specification tables indicate what size tubing the pump will fit in. For instance, the Kobe and Guiberson pumps start with the nominal sizes: (in )
Nominal size
Tubing size (In )
2
2: OD
2:
2 i OD
3
3; OD
4
4; OD
The Oilmaster pumps use the following code in the first two digits: Code
Tubing size (in.)
20
2; OD
25
2; OD
30
3 f OD
The above information allows one to choose a pump to fit the given tubing size, whereas the following equation allows one to choose a pump with sufficient capacity for the well: Qp
= Minimum
pump displacement =
(Q,
X B ) + Qs 0.85 x 0.85
(15-1)
661
Example Pump Intake Pressure = 400 psia Oil Temperature = 100°F Oil Gravity = 3 5 O API Gas Specific Gravity = 0.7 Solution GOR = 65 scf/B
SOLUTION GAS-OIL RATIO A T PUMP INTAKE PRESSURE ( F r o m AIME Jour. Pet. Tech. T.P. 2931, 1950)
Fig. 15-27. Bubble point pressure correlation. (Modified after Borden and Rzasa, 1950, fig. 1, p. 346; courtesy of the SOC. Pet. Eng. of AIME.)
where: Q, = required rate of oil production, Q6 = required rate of water production, and B = formation volume factor. When choosing a pump, it is good practice to choose a pump that will operate at less than 85% of its maximum rated speed. In addition, a pump volumetric efficiency, usually 85%, should be assumed to allow for fluid slippage due to the pump wear over a period of time. These are the factors (0.85 X 0.85) in the denominator of eq. 15-1. The formation volume factor, B , should be obtained from actual PVT tests on the particular well fluid to be pumped; however, in the absence of this data, Figs. 15-27, 15-28, 15-29, and 15-30 can be used. Hydraulic pumping installations can be designed to require the pump to handle (compress) the free gas or they can be designed to allow the free gas to be vented (through the casing annulus or through a separate string of tubing) to the surface separately from the pumped fluid.
PBOPERTIES OF NATURAL HYDROCARBON MIXTURES OF GAS AND LIQUlO
m
Formation Volurrrc of Bubble Poiot Irtprid
m
N
EXAMPLE
GOR = 6 5 scf/B Gas Sp. Gr. = 0.7 Oil Gravity = 3 5 0 API Temperature = lOOOF B = 1.037
Copyright 1947 Chevron Rerearch CornReprinted by Psrrnmnmon
B, FORMATION VOLUME of BUBBLE POINT LIQUID -
Fig. 15-28. Oil formation volume factors of California natural hydrocarbon mixtures of gas and liquid at bubble point. (After Standing, 1952; courtesy of Reinhold Publishing Corp.)
P R W f R T I E S OF NATURAL HYDROCAKEON MIXTURES OF GAS A N D LIQUIU
Form;rtion Volrimc nf Gas Plus Liquid Phases
EXAMPLE GOR = 200 scf/B Gas Sp. Gr. = 0.7 Oil Gravity = 350 API Temperature = lOOOF Pressure = 4M) psia E = 1.1
m w m
Fig. 15-29. Total formation volume factors of California natural hydrocarbon mixtures of gas and liquid. (Alter Standing, 1952; courtesy of Reinhold Publishing C'orp.)
664
THEORETICAL VOLUMETRIC EFFICIENCIES OF CASING PUMPS Gas Gravity=0.8, Oil Gravity=40° API, Bottom Hole Temperature=150°F 3000
.
2500
2000
1500
8
1000
900 800
ui
I 3 VI
700
P r
600
W
500
Y
6
+
z
400
P
I
2
300 250
200
150
I FORMATION SHRINKAGE FACTOR
I
'I
01 .10
I
.20
.40
.30
.50
.20
.30
.40
.50
3
c a B
,
;
40
.20
Lu
.30
.40
.50 .60
4
.60
.80
.70
.90
.60
j
1.00
4
.70
.80
.90
,
1.oo
I
--d=
.90
1.00
.80
:
60
4
.30 80
S
I
20
c K
-
.
.40
+
;
.40
.50
.50
.60
.70
.a0
.90
1 .oo
+ .60
.70
.80
.90
1 .oo
Fig. 15-30. Determination of theoretical volumetric efficiencies of casing pumps.
If gas is to be vented, one can usually assume B = 1.0. To be precise, however, one should use Fig. 15-27 to arrive at the solution gas,/oil ratio at pump intake pressure. With this value, Fig. 15-28 can be used to determine the formation volume factor, B , for the produced oil. (The formation volume factor for water is taken as 1.O.)
665
Example: If gas is being vented, pump intake pressure is 400 psi, oil has 35”API gravity, the pump intake temperature is 100”F, and the gas gravity is 0.7. Fig. 15-27 indicates a solution gas/oil ratio of 65 scf/bbl. From Fig. 15-28, the formation volume factor is equal to 1.037. If 100 bbl/D of oil and 100 bbl/D of water are to be produced, the minimum required pump capacity from eq. 15-1 becomes:
Qpmin =
100 X 1.037 + 100 = 282 B/D 0.85 X 0.85
If free gas is to be compressed by the pump, Fig. 15-29 can be used to obtain the formation volume factor of the oil plus free gas. For example, if the gas/oil ratio in the above example is 200 scf/B and the free gas is not being vented, Fig. 15-29 gives a formation volume factor of 1.7. Again, using eq. 15-1, minimum pump displacement is equal to: Qp min =
+
100 x 1.7 100 0.85 X 0.85
= 374
B/D
Figure 15-30 has been obtained from Fig. 15-29 at the specific conditions shown. This chart takes into account the water produced but is made for “formation shrinkage factor”, which is the reciprocal of formation volume factor. If symbol, S , is used for this shrinkage factor, eq. 15-1 becomes: .
Q
=
pmn
Qs’Q6
(15-1.a)
(0.85 X 0.85) X S
In the above example, Fig. 15-30 gives a value of S of approximately 0.71. The minimum required pump capacity thus becomes: Qp
=
200 (0.85 X 0.85) X 0.71
= 389 bbl/D
When S is less than 0.5 or when [ ( Q , X B ) + Q 6 ]is more than [2 X (Q, + Q 6 ) ] ,it indicates that the pump will be required to handle more gas than liquid. Under these circumstances one should consider either (1)changing the tubing arrangement to vent the gas or (2) allowing a greater pump intake pressure (flowing bottomhole pressure). Vogel’s reference IPR curve for solution-gas-drive reservoirs (Fig. 15-31) illustrates that as the flowing bottomhole pressure is reduced, the incremental increase in production rate is decreased. For example, a deep well with a static reservoir pressure of 3000 psi (20.7 Pa) is to be produced at 1000 psi (6.89 Pa) flowing bottomhole pressure. Figure 15-31 shows that at this pressure (pwf/p, = $$$ = 0.3) one will be producing at 85-90% of the capacity of the well. To obtain the 10-15% additional production rate may require considerable additional
666 1 .oo
VOGEL'S C U R V E FOR INFLOW PERFORMANCE RELATIONSHIP FOR SOLUTION-GAS D R I V E RESERVOIR
0.80
0.60
0.40
0.2c
I
C 3
0 40
0.60
0 80
'0
PRODUCING RATE AS A FRACTION OF MAXIMUM PRODUCING RATE q,/iq,iMAX.
Fig. 15-31. Relationship between the bottomhole pressure (as a fraction of reservoir pressure) and producing rate (as a fraction of maximum producing rate) for solution-gas drive reservoir.
equipment costs, because as the flowing bottomhole pressure decreases the value of B increases rapidly. It was discussed above how to choose a pump: (1) to fit the tubing and (2) with sufficient capacity (displacement). Now one must (3) choose a pump that will not require excessive surface pressure. The following equation, which is derived in a later section, enables one to determine the surface power fluid pressure, ps: (15-2) where G4= fluid gradient of produced fluid, A p / h ; h , = pump setting depth; h4 = height of fluid above pump; P / E = pressure ratio of pump; and Fp = friction in pump.
667
Solving this equation for pressure ratio gives: P/E
=
P, - F p
(15-2.a)
G4(h1
Good practice is to select a pump that will not require over 3500 psi (24.1 Pa). Pressures up to 4000 psi (72.6 Pa) are occasionally found in practice, but 4500 psi (31.0 Pa) and above are rare. The quantity in the specification tables that relates to surface operating pressure is found in the P / E column. This quantity is called the “pressure ratio” or simply “ P over E”. It is the ratio of the effective pump area to the effective engine area. Substituting ps = 3500 psi and Fp = 400 psi (2.76 Pa), an average value, into eq. 15-2.a, one obtains: P/E(max.)
=
3500 - 400 G4(h1
- h4)
3100
(15-3)
G4hl - G 4 h 4
In SI units: P/E (max.)
=
24.1 - 2.76 G4(h1
-
21.3 G4h1
-
(15-3 .a)
G4h4
The quantity ( h , - h,) is the net lift or working fluid level in the well. When the pump is pumping from beneath a packer, there is no “working fluid level” in the well. It is, therefore, helpful to expand the denominator to (G,h, - G4h,) as shown in eq. 15-3. Flowing bottomhole pressure ( p4 = G4h4) can then be used instead of working fluid level. Thus, the pump selection procedure can be summarized as follows: (1) Select a pump to fit the given tubing size. (2) Select a pump to fit the given production rate (eq. 15-1). (3) Select a pump to operate at less than 3500 psi (24.1 Pa) surface pressure (eq. 15-3). Example problem 15-2
Given : Oil production = 100 bbl/D (15.9 m3/D) Water production = 100 bbl/D (15.9 m3/D) GOR = 500 scf/bbl (90 m3/m3) h , = 8000 ft (2440 m) h , = 2000 ft (610 m) G4 = 0.4 psi/ft (0.00904 Pa/m) Tubing size = 29 in. O D (7.3 cm)
Find: (A) The suitable pumps in Fig. 15-11 if gas is vented. (B) The suitable pumps in Fig. 15-11 if gas is not vented.
668
Solution: (A) Assume B
= 1.0
Minimum p u q p displacement, Qp min P/E(max.)
=
3100 0.4( 8000-2000)
=
2oo 0.85 X 0.85
= 277
bbl/D
= 1.29
According to Fig. 15-11, from 2i-in. pumps the second, third, fifth, and sixth meet the above criteria. The fifth pump has the lowest P / E and, consequently, the lowest surface operating pressure. Thus, it would probably be the ideal choice. (B) Flowing bottomhole pressure = 0.4 (2000) = 800 psi From Fig. 15-30, S = 0.62; thus: Minimum pump displacement, Qp min = P / E (max.)
=
3100 0.4(8000) - 800
200 0.85 X 0.85 X 0.62
= 446
bbl/D
= 1.29
According to Fig. 15-11, three 2i-in. pumps meet these criteria. If the well is 12,000 f t (3600 m) deep and the pump is set on bottom, the pump submergence, h , , will become 6000 ft (1830 m). This does not affect the pump selection if gas is to be vented. But if gas is not vented, a new value for p , and S (or B ) must be found: p4 = 0.4(6000) = 2400 psi From Fig. 15-30, S = 0.95 (approximately); thus: Minimum pump displacement, Qp min P / E (max .) =
3100 0.4(12000) - 2400
=
200 0.85 x 0.85 x 0.95
= 291
bbl/D
= 1.29
In this case, one can choose the second, third, fifth, or sixth pumps from Fig. 15-11. The maximum surface operating pressure and the maximum P / E value are not altered by setting the pump deeper. This example shows why it is almost always desirable to set hydraulic pumps as deep as possible. For a gas vented system, if the working fluid level is say 6000 ft, the surface power fluid flow rate and pressure will be the same whether the pump is set at 6000 ft, 12,000 ft or any other depth (neglecting friction of course). When the pump is required to compress gas, deeper setting depths provide higher pump intake pressures, which in turn provide higher values for S (lower values for B ) , and, therefore, improved pump loading conditions. To select a surface pump, it is necessary to know the power fluid flow rate and
669
TABLE 15-1 Pump selection procedure and the first approximation of surface power requirements
To find
Use
Pump to fit in given tubing (size)
Pump specification tables, Figs. 15-11 through 15-24 ( Q 5 X B)+Q6 Q5+Q6 (eq. 15-1) or 0.85 x0.85 0.85 X0.85 X S (eq. 15-3) P / E (max.) = 3100/(G4h, - G 4 h 4 ) P , = G ~ ( ~ I - ~ ~ ) PFD /E+ (eq. 15-2) ( Q 5 B + Q 6 ) 91 ( Q 5 +Q6)qi (eq. 15-4) 'I= 0.85x0.90xq4 0.85x0.90x9,xS
Pump with sufficient capacity Pump to operate below 3500 psi Surface power fluid pressure Surface power fluid rate
pressure required by the subsurface pump, or pumps, which have been selected. Equation 15-2 will give a suitable first approximation for the surface pressure required. (Equation 15-28 in a later section, with substitutions from Fig. 15-59, will give the precise value of p,). The following equation (derived in a later section) will give the value for power fluid flow rate, Q,: (15-4) where: q, = engine-end displacement factor; and q4 = pump-end displacement factor. Other quantities have been previously defined. The downhole pump specification tables list the values for q, and q4 in bbl/D per SPM (strokes per minute). In the denominator, the 0.85 is the pump-end efficiency and 0.90 is the engine-end efficiency. Table 15-1 summarizes pump selection procedure and the first approximation of surface power requirements. If these calculations are 75% below surface and subsurface pump maximum specification ratings, then it will usually be safe to omit the precise calculations covered in a later section. Example problem 15-3 Given: Same data as in Example problem 15-2 (B) Assume: Fp = 400 psi Find: (A) Surface power fluid rate. (B) Surface power fluid pressure. ( C ) Surface power required. Solution: In the solution of problem 15-2 (B) it was found that only one pump, the 2i-in. x l&-in. - l&-in., is suitable. This pump has a P / E = 1.0, q4 = 7.03, and q1 = 7.13. S was found to be equal to 0.5.
670
(A) From eq. 15-4:
=
200 x 7.13 0.85 X 0.90 X 7.03 X 0.5
= 530 bbl/D
(B) From eq. 15-2: p s = 0.4(6000)(1 .O) - 400 = 2000 psi
( C ) A useful hydraulic equation is: Q (bbl/D)
Xp
(psi) X 0.000017 = Horsepower
530 X 2000 X 0.000017 = 18 H P Class problems (1) Given: Production rate = 700 bbl/D (no water) Working fluid level = 7000 ft from surface G4 = 0.36 psi/ft Tubing size = 2i-in. O D GOR = 500 scf/D Pump setting depth = 12,000 ft Maximum surface pressure p s = 3500 psi Pump must handle all the gas. Find: List the pumps capable of producing t h s well.
(2) Giuen: Same data as in Example problem 15-2 except pump setting depth = 10,000 ft. Find: List the pumps capable of producing this well. TUBING ARRANGEMENTS
The most common type of tubing arrangement is the casing “free type” shown in Fig. 15-32. In this arrangement all the gas must be handled by the pump. For this reason some of the following well conditions usually exist: (1) Low gas/liquid ratio. (2) Near-zero pump intake pressure is not required because: (a) Excessive production rate may bring in too much sand. (b) Excessive production rate may increase water/oil ratio. (c) Incremental increase in oil production is not economically justified by increased operating costs. (d) Maximum well capacity (1)is not desired, (2) is not allowed, or (3) is beyond equipment cap ability.
671
The “parallel free” type system is shown in Fig. 15-33. This arrangement allows the gas to be vented to the surface through the casing annulus. The size of tubing string for returning the production and power fluid is usually smaller than the size of power fluid tubing string because of casing size limitations. If the fluid flow rates involved present an undue fluid friction problem, an arrangement such as the one shown in Fig. 15-34 (reverse-circulation system) can be used. In t h s system, the power fluid goes down the small string and production plus power fluid returns up the large string. T h s requires a hold-down to keep the pump on seat while pumping and requires that a releasing tool be dropped before circulating the pump out. Another alternative, which is a “casing free” type system with gas vent, is shown in Fig. 15-35. In this arrangement the small string is used to vent the gas to the surface. Figures 15-36 and 15-37 can be used for calculating the amount of gas that can be vented in t h s type of system. Up to this point, the writer covered only the “open power fluid” (OPF) systems. In Fig. 15-38, the “parallel free closed power fluid” (CPF) arrangement is shown. In this system the power fluid is not commingled with the production-it is returned to the surface through a third tubing string. This system and its slight variation as shown in Fig. 15-39, is used extensively in California, U.S.A., for the following reasons: (1) Gas must be vented because near-zero pump intake pressures are required. (2) Closed power fluid systems are used because: (a) Fine sand is produced, which is difficult to remove by settling tanks. (High pump repair costs result from solids in the power fluid.) (b) Because many oil fields in California are in ecologically sensitive areas, water is preferred as the power fluid. With hydraulic piston pumps (as opposed to hydraulic jet pumps), usually a lubricant must be added to power water. The CPF system conserves the lubricant, whereas in the OPF system the lubricant must be continuously replenished. In many cases, fresh water is used as the power fluid. Figure 15-40 shows a “casing free” CPF system, in which the produced fluid is conducted to the surface through the casing annulus and power fluid is returned through the side string of tubing. In this type of system, it is also possible to return power fluid through the casing annulus and produce through the side string. This latter arrangement would allow corrosion inhibitor in the power fluid to protect the inside walls of casing. Many combinations of the above arrangements are possible in dual-completed wells. Figure 15-41 shows an arrangement where the two pumps are run in tandem in one tubing string to produce a dual-zone well, whereas in Fig. 15-42 two pumps are run in tandem to produce twice the capacity from a single-zone well. This is a “casing free” OPF type system; however, all of the previously discussed arrangements can be made to accommodate single-zone tandem pumps. Figure 15-43 shows how a conventional subsurface safety valve can be used with hydraulic pumps. This valve is the kind typically installed below the mud level in offshore flowing wells. In this case, the valve is installed just above the packer. The actuating line is connected to the power fluid tubing just above the pump (10-20 ft).
Fig. 15-33.Parallel free type system.
r rn r n
8-n Fig. 15-32.Casing free type system.
rn
rn
I13
I-
%D
PARALLEL F R E E V
CASING F R E E
PROD. + PWR. FLD.
PWR. FLD.
Fig. 15-34. Parallel reverse-circulation type system.
Fig. 15-35,Casing free type system with gas vent,
674
PRESSURE-DROP IN P.S.I. Pressure drop through
1 ” gas v e n t line I C O u r t e w A R M C O - N a t i o n a l
Production SvrternP)
Fig. 15-36. Determination of pressure drop through 1-in. gas vent lines. (Courtesy of ARMCO-National Production Systems.) Gas gravity = 0.60; average temperature = 150°F.
615
PRESSURE-DROP IN P.S I Pressure drop through 1-114" gas vent lines ICourtesy A R M C O - N a t i o n a l
Production S y i f e m d
Fig. 15-37. Determination of pressure drop through 1:-in. gas vent lines. (Courtesy of ARMCO-National Production Systems.) Gas gravity = 0.60; average temperature = 150'F.
676
CPF
CPF
CPF
'c -=
'y K
;i
i
i
1
\
I
Pump Unseated By Produced Fluid
Pump unseated by power return fluid
PARALLEL FREE
PARALLEL FREE
Gas Pumped
CASING F R E E
Fig. 15-38. Parallel free closed power fluid (CPF) system. Pump is unseated by produced fluid. Fig. 15-39. Parallel free closed power fluid (CPF) system. Pump is unseated by power return fluid. Fig. 15-40. Casing free CPF system.
r
?
N
y
B
v
5'
5'
2
0
m
L
R
I
$
0
5'
e3 -
0 $
!-
4 0
Lu
3
N
0
E
a
?
Ei6
$ rn
2.
%
:: fZ I Ez
3
% !
Ei Ei
2
2
Ei
3 =0
4
4
2
?s EzZ
TJ
>
Y - I
b-
y
7
m
PROD. + PWR. FLD.BOTH PUMPS
PWR. FLD.BOTH PUMPS
PROD. UPPER PUMP PROD. LOWER PUMP + PWR. FLD. BOTH PUMPS
PWR. FLD. UPP. PUMP
PWR. FLD. LWR. PUMP
678
S A F E T Y VALVE-, requires high pressure to open and to keep open. Spring closes valve when pressures are balanced.
Fig. 15-43. Use of conventional subsurface safety valve with hydraulic pumps
When the pump is operating, the actuating pressure is the surface operating pressure (1000-3500 psi). When the surface pump is shut down for any reason, the hydrostatic pressure in the tubing and casing equalize, and the valve closes, shutting-in the tubing and casing at the packer. Some valves, sensitive to the pressure beneath
619
B
L
OPF
OPF
W
Z
$9
Gas Pumped !-Liquid F I X E D INSERT
FIXED CASING
Fig. 15-44. Fixed insert type system. Fig. 15-45. Fixed casing type system. Fig. 15-46. Fixed casing type system with separate gas vent tubing string.
FIXED CASING
680
them, will remain open if the hydrostatic pressure in the actuating line and tubing is above the pressure below them (bottomhole pressure). This does not present any problem because the only possibility of the well flowing is when the bottomhole pressure is greater than hydrostatic pressure, in which case these valves will close. If the safety valve is for 2i-in. tubing and the pump and tubing above it is 2; in. in diameter, the safety valve can be the wireline retrievable type. Figures 15-44, 15-45, and 15-46 show “fixed” type pumps. These arrangements require a tubing string to be screwed into the top of the pump and hence are not “free pumps”. Figure 15-44 shows the so called “fixed insert”, i.e., the arrangement used with the first hydraulic pump ever installed (Kobe pump, Inglewood, California, 1932; now in the Smithsonian Institute). This arrangement is occasionally used for one or both zones of a dual well. Figure 15-45 shows a “fixed casing” type system, which is used where the required pump capacity cannot be obtained with a pump that will fit inside the existing tubing. In Fig. 15-46, a “fixed casing” type system with separate gas vent tubing string is demonstrated.
WELLHEADS
Wellheads for hydraulic pumps can be simple and compact with a low profile. They can be closely spaced and can be below ground level. Wellheads are topped by a 4-way valve that provides the following functions: (1) Directs power fluid down the power tubing and production to the flowline for the “pump-in and operate” position of the valve. (2) Directs power fluid down the proper tubing (or casing) and returns fluid to the flowline for the “pump-out’’ position of the valve. (3) Provides a safety device to prevent high pressure from accidentally being applied to the casing. (4) Shuts surface line and provides a means to bleed pressure from the tubing for the “shut-off and bleed” position of the valve. (5) Provides a means to catch, hold, and release the pump.
COXTROL MANIFOLDS
A central control manifold is usually used to regulate the flow of power fluid to the individual wells of a multiple-well installation. These manifolds are usually located at a central location near the surface pumps, but can be located at satellite locations. In some cases, the control valves are attached to each wellhead. Control manifolds are available from the hydraulic pump manufacturers and consist of modular header sections that can be readily added to or subtracted from the manifold. Each header section has a “constant flow” control valve that automatically compensates for pressure variations to keep the rate of flow of power fluid to each well constant at the initial manual setting. In addition, the manifold
681
contains high-pressure meters, gauges, and a back pressure regulator to keep a constant pressure on the surface pumps.
SURFACE PUMPS
The surface pumps commonly used are designed specifically for power fluid service and are supplied by the downhole hydraulic pump manufacturers. For high-pressure clean oil service, these pumps usually use metal-to-metal plungers and liners and ball-type valve components which require little or no maintenance. For water service, plungers and liners with packing are usually used. Most pumps are skid mounted with electric motors or gas engines and include relief valves, back pressure regulators, gauges, safety switches and pulsation dampeners. Sometimes suction charging pumps are used. Suction lines are the most critical components of a surface pump installation. These lines should be one size larger than the pump inlet connection. A flow velocity of one foot per second is a good rule of thumb, and all valves should be full opening. Bypass and relief lines from the high-pressure side of the pump should not tie into the suction line, because t h s can cause pulsation (hammer) in the fluid end of the pump due to gas flashmg out of solution. A pulsation stabilizer installed on the suction side of the pump as close to the cylinders as possible is good insurance. Many pump discharge pulses and vibrations are cured by a pulsation stabilizer on the suction side of the pump. Charging or booster pumps are also an aid to proper loading of surface pumps. Elbows and tees should be kept to a minimum on the discharge side of the pump also. These offer a reflecting surface for pressure pulse waves and are the source of many vibrations. Sometimes just changing the geometry of discharge lines will eliminate troublesome vibrations. This happens when the geometry of the lines create a conduit tuned to a harmonic of the pulse wave length. Although most system vibrations are hydraulic in nature, sometimes they are caused by misalignment of the prime mover and pump or by loose foundation bolts.
POWER FLUID CLEANING SYSTEM
Power fluid quality, especially solids content, is an important factor contributing to pump life and repair costs. Power fluid leakage through a pump’s fits and clearances is a function of wear caused by abrasive solids and of the viscosity of the power fluid. The permissible solids content varies somewhat depending upon the definition of “acceptable pump life” and also on the viscosity; however, 10-15 ppm is usually acceptable for 30-40” API gravity oils. For heavier oils, more wear and, consequently, more solids may be tolerated, whereas for water, less wear and less solids are usually the rule. The maximum particle size should not exceed 15 pm, whereas the maximum salt content of oil should not exceed 12 lb/1000 bbl.
682
There are two basic types of power fluid systems: (1) The closed power fluid (CPF) system, where the surface and subsurface power fluid stays in a closed circuit and does not mix with the produced fluid. (2) The open power fluid (OPF) system, where the power fluid mixes with the production downhole and returns to the surface as commingled power fluid and production. The choice of oil or water for power fluid can be based on a number of factors. Following is a list of most of the factors involved in this choice: (1) Water is preferred for safety and environmental reasons. ( 2 ) For CPF installations, the addition of chemicals to power water for lubrication and corrosion control is not a large cost factor. (Fresh water is frequently used in CPF installations.) (3) For OPF installations, the addition of chemicals to power water can be a significant cost factor because the power water is commingled with the produced fluids. This requires continuous injection of chemicals which will add to the operating costs. (4) Treating power oil is seldom a large cost factor mainly because it seldom needs chemical additives for lubricity. One exception is when high-gravity oils are used at very h g h bottomhole temperatures. When these two factors produce a viscosity below 1.0 centistoke (see Fig. 15-62) a lubricant may be necessary for long pump life. ( 5 ) Maintenance on surface pumps is less when using oil, because metal-to-metal plungers and liners are usually used instead of packing. Also valves last longer and are usually ball-and-seat type rather than disc or poppet type, which are usually used for water. Additionally, the low bulk modulus of water causes much larger pressure pulses than oil, and these pulses are detrimental to pipe connections as well as contributing to fatigue failures of pump components. (6) Subsurface pumps are sensitive to viscosity and lubricating qualities of the power fluid. Because water has practically no lubricating qualities at bottomhole temperatures, it can, if not adequately treated, contribute to short pump life. Leakage of power fluid past the various sliding fits in the pump is a function of the viscosity and is greater with water than with most crude oils. (7) Testing a well for oil production is subject to an added source of error when oil is used for power fluid. (This statement is not true when using the “wellsite power plant”.) In an OPF system, the power oil must be metered in and subtracted from the total returned oil, and small errors in metering can be significant when the ratio of power oil to produced oil is large, e.g., when the well is producing a large percentage of water. For instance, if the ratio of power oil to produced oil is 10 : 1, an error of 2% in the power oil meter translates into a 20% error in measuring the produced oil. (8) Usually the surface pressure required will be less when using power water as compared to using power oil. (9) Although hydraulic pumps handle viscous crudes (7-20”API) very well, it is advantageous to use a higher gravity oil for power fluid and an OPF system. This
683
-
S U R F A C E PUMP
WELLS
Fig. 15-47. Surface facilities of CPF system.
commingles the two crudes at the discharge of the pump, thus diluting the heavy oil for ease in transporting it at the surface. CPF SYSTEM
In the CPF system, an extra downhole conduit must be provided for returning the spent power fluid to the surface. This causes the system to be more expensive than the OPF system and consequently its use is not widespread. The bold lines on Fig. 15-47 show the surface facilities for a CPF system. Because the power fluid tank is relatively small, this system is popular for urban locations and offshore platforms where surface space is at a premium. For example, it is quite popular in California, U.S.A.,due to the many townsite and offshore well locations. Frequently CPF systems use water for the power fluid because it is less hazardous and presents fewer ecological problems than high-pressure oil. In the case of water, however, corrosion inhibitors and lubricants must be added, and all oxygen must be removed -considerations that add to the operating costs. Figure 15-47 shows two wells on this system, but there is no reason that 30 or even 100 wells cannot be placed on a system of this type. Power jluid tank (CPF)
In most downhole pumps, the pump-end is lubricated with the power fluid and consequently part (typically 2-10%) of the power fluid is purposely “leaked” to the
684
production. This loss of power fluid must be replaced with clean fluid. The purpose of the power fluid tank in Fig. 15-47 is to remove abrasive particles from the make-up fluid and part of the recirculated fluid. One misconception concerning the CPF system is that the power fluid will remain clean because it has no source of contamination. In actual practice, however, this is not true due to the following reasons: (1) The power fluid tank does not completely remove all of the solid particles from the make-up fluid-cleanliness is relative, not absolute. ( 2 ) The power fluid is not completely non-corrosive. Again, this factor is relative, not absolute. The products of corrosion are generally abrasive solids. (3) When fluid containing solids, even a very small percentage of solids, is leaked through a long closely-fitted clearance space as in a downhole pump, the solids tend to be held back. This means that the fluid emerging from the fit is cleaner than the fluid trying to enter the fit. The tendency, then, is for the power fluid circuit to lose clean fluid and to retain the solid particles. Over a period of time, the above three factors allow the power fluid in the closed circuit to become “dirtier” than the fluid emerging from or the fluid entering the closed circuit, unless a part (10% is reasonable) of the recirculated power fluid is continuously cleaned, as in Fig. 15-47 by the power fluid settling tank. This “continuous cleaning of part of the recirculated power fluid” is an important feature in the design of the CPF system. When water is used for the power fluid, filters may be used instead of settling tanks for the cleaning process. These filters should remove particles down to 10 pm. When oil is used, experience has shown that the settling tank should be large enough to keep the upward velocity of the oil below 1 ft/hr for oils below 30”API gravity and below 2 ft/hr for oils above 30”API gravity.
OPF SYSTEM
In the OPF system, only two downhole paths are needed, one for conducting power fluid to the engine and one for conducting spent power fluid plus production to the surface. These conduits can be two strings of tubing or one tubing string and the tubing-casing annulus. Simplicity and economy are the important features of the OPF system. When water is used for power fluid in the OPF system, the chemicals added (for lubrication, corrosion inhibition, and oxygen scavenging) are generally lost when mixed with the production and, consequently, must be added continuously. The bold lines on Fig. 15-48 show the surface facilities for an OPF system with two wells. Central plants of this type can be used for any number of wells. Usually, the triplex pump and control manifold are located at the central tank battery, whereas control manifolds can be located at satellite locations. Even triplex pumps can be located at satellite locations if a small pump is used at the battery to get the fluid to the suction of the triplex pump. Satellite manifolds and satellite pumps reduce the length of high-pressure surface power fluid lines.
Fig. 15-48. Surface facilities of OPF system.
Power fluid tank (OPF)
The OPF power fluid tank shown in Fig. 15-49 has been proven over the years to represent an excellent design. This design and one with a slight modification are almost universally used. Oil generally enters the gas boot in surges and contains gas not removed in the treater, i.e., gas that was in solution at the 30-psi treater pressure. The purpose of the gas boot is to remove the last remnants of gas which would otherwise keep the tank stirred up. The top section of the boot should be 36 in. in diameter to be effective; however, even with this diameter, surges frequently occur that cause the oil to be carried over the top through the gas line. To prevent this oil carry-over from going into the top of the tank, and thereby bypassing the settling process, a loop with riser is used to tie the boot gas line to the tank gas line. Dead oil (gas-free) then enters the bottom of the tank which should have a level spreader. The oil entering here is power oil plus production. At the vertical midpoint, production is drawn off through the outside riser that keeps the tank full. From the midpoint up, the power oil settling process takes place. The settling out light solids are carried with the production to stock, whereas the heavier particles which settle to the bottom must be removed periodically. The one modification sometimes used is to allow the production to be taken off the boot or between the boot and the tank, but still with a riser to keep the tank full. This modification allows full use of the tank for power oil only, but imbalances in the column weight of the riser and of the tank can occasionally cause seemingly strange fluid level variations in the tank. With this modification, the solids should
GAS BOOT 4 MINIMUM DIAMETER CONNECTED TO T A N K 12 FROM BOTTOM RISES TO 18 IBOTTOM OF CONNECTION) FROM TOP AND THEN TO STOCK AS
3 BOLTED SECTIONS 8 x 2 0 SWAGED TO TOP 8' x 36' SFCTION TOP SECTION SHOULD HAVE SIDE DPFNlNG IY M I N D I A I LOCATED I N CENTER.
CLOSED EXCEPT
. ,
'//
O I L FROM TREATING PLANT
FROM TRIPLEX RELIEF
PRODUCTION I
WITHDRAWAL LINE
LINE TO TANK
r
MINIMUM DIAMETER 1' ABOVE BOTTOM OF TANK R I M V A L V E MUST BE F U L L OPENING END OF
TRIPLEX SUPPLY LINES
3" DIAMETER MOUNTED 3" OFF BOTTOM INSIDE TANK, LINE SHOULD GO TO CENTER OF TANK AND HAVE DROP PIPE DOWN I N SUMP TO 6 OFF BOTTOM OUTSIDE LINE SHOULD BE CONNECTED BOTH WITH STOCK TANK IAS SHOWN UNLESS STOCK T A N K IS HIGHER THAN 1 6 ) A N D PIT I F F L A T BOTTOM TANK I S USED. INSIDE LINE SHOULD HAVE B U L L HEADED TEE MOUNTED HORIZONTALLY I N CENTER OF TANK
T MINIMUM DIAMETER TWO LINES ARE FLANGED
X TO 7 FROM TOP OF TANK, R U N 1 ' APART, JOINED ABOUT 5' FROM BOTTOM OF R I M GATE VALVES ARE F U L L OPENING FLANGES SHOULD BE AS FAR FROM SlOCK DRAW OFF AS POSSIBLE NOT LESS T H A N 90 DEGREES AROUND TANK
Fig. 15-49. Power fluid settling tank in OPF system.
SPREADER ROUND WITH SERRATED OR PERFORATED SKIRT 8 MINIMUM DIAMETER MOUNTED LEVEL 2' ABOVE BOTTOM R I M OF TANK
POWER OIL TANK
687
Fig. 15-50.Wellsite power plant unit. (Courtesy of Kobe, Inc., subsidiary of Baker International Corp.)
be drained from the bottom more frequently. The boot can be placed inside the tank; however, experience has shown cases of pitting and corrosion at the fluid level. The outside location, therefore, is preferred. In order to assure adequate particle settling, the power oil tank should be sized to allow the upward velocity in the top half to be less than 2 ft/hr. This velocity is equivalent to 1500 bbl/D in a 750-bbl, 24-ft high tank. The velocity should be lower for oils heavier than 30"API gravity and for operations in extremely cold climates. Treating the oil for sale to the pipeline must be done ahead of the power oil tank.
INDIVIDUAL WELLSITE POWER PLANT
The use of individual wellsite power plants is becoming more and more popular, and these units are competitive with other types of artificial lift. Typical wellsite units are shown in Figs. 15-50 and 15-51. A wellsite power plant is a package of components, installed at or near a wellsite, that accomplishes the functions normally performed at a central plant. The basic components consist of a liquid-gas separator, centrifugal separators for removing solids from the power fluid (oil or water), and a surface pump. These units are portable, require a minimum of installation labor, and eliminate the need for advance long-range planning of a central power plant. Although they are always used with an OPF tubing arrangement, they have one feature similar to a CPF
688
Fig. 15-51. Wellsite power plant unit. (Courtesy of Guiberson Division, Dresser Industries, Inc.)
system: the net production from the well goes into the flowline while the power fluid is recirculated at the wellsite. This feature simplifies well testing and does not increase the load on the treating system at the tank battery. Wellsite power plant is simple, flexible, compact, and portable, features that are of great interest to the design engineer, the production foreman, and the lease operator. Generally, there is a choice of having either a central system or an individual wellsite system. Some choices are obvious, as in the case of a central system for an offshore platform or any cluster of wells (such as in a downtown area or islands constructed for that purpose). For those wells that are isolated or widely spaced, the individual wellsite system will probably be preferred. Flow schematics in Figs. 15-52, 15-53, and 15-54 show models of Kobe (Solo Unit), Oilmaster (Unidraulic), and Guiberson (Econodraulic) wellsite power plants. The important functions these units must perform are: (1) Provide gas-free fluid to the surface pump. (2) Provide means to choose oil or water for the power fluid. (3) Remove the solids from the power fluid. (4) Provide surge and reserve capacity for circulating a subsurface pump to the surface after a pump failure. (Vessel sizes range from 40 in. X 10 ft to 5 ft X 20 ft.) If a well produces 600 bbl/D of water, 200 bbl/D of oil, and uses 2000 bbl/D of power water, these flow rates are shown on the simplified schematics of Figs. 15-55, 15-56, and 15-57.
689 RELIEF VALVE
Q FROM
WELL
0
' I
\
RELIEF VALVE
I
TO WELL
FLOW LINE-
-P O -METER a -
P D METER
CLOSED
CLOSED
t OPEN FOR RESERVE POWER O I L SUPPLY OTHER V A L V E S SHOWN I A I FLOAT.OPERATED DUMP V A L V E (61 VESSEL BACK-PRESSURE V A L V E ICI M E T E R BY-PASS V A L V E
Fig. 15-52. Kobe Solo wellsite power plant unit.
ACCUMULATOR
RESERVOIR
ANK ERY
Fig. 15-53, Oilmaster Unidraulic wellsite power plant unit.
690
CHECK VALVE
L
-.
29
- 1-
MOTOR
DAMPENER
POWER FLUID PUMP ::NTROL
CHECK
PANEL
Fig. 15-54. Guiberson Econodraulic wellsite power plant unit.
r
I
'
I
\I
I 2000
Fig. 15-55. Flow rates (flow diagram) in a Kobe Solo unit, assuming water production of 600 bbl/D, oil production of 200 bbl/D, and use of 2000 bbl/D of power water.
691
I -1
I
I
I
I
-9
2000
-
-
2000
Fig. 15-56. Flow diagram in an Oilmaster Unidraulic unit, assuming water production of 600 bbl/D, oil production of 200 bbl/D, and utilization of 2000 bbl/D of power water.
Removing solids from the power fluid is usually accomplished by cyclone centrifugal separators (Fig. 15-58). These cyclones require 30-60 psi pressure drop from inlet (feed) to top outlet (overflow). The ratio of overflow to underflow out of the apex of the cone is controlled by the relationslup of overflow pressure to underflow pressure. Usually, the overflow must be 5-10 psi greater than the underflow pressure to insure a positive rather than a negative underflow rate. The cyclone internals, feed nozzle, vortex finder, and apex can be sized to accommodate various rates of flow.
I
\
2200
1
0
N 0
--
800
L -
2000
600
0 --
2000
wELb
Fig. 15-57. Flow diagram in the Guiberson Econodraulic unit, assuming water production of 600 bbl/D, oil production of 200 bbl/D, and utilization of 2000 bbl/D of power water.
692 OVERFLOW (CLEAN LIQUID)
VORTEX
FINDER
-L
(2) LIQUID ROTATION DEVELOPS H I G H C E N T R I F U G A L FORCES THROUGHOUT CYCLONE
FEED NOZZLE
/
(1) FEED INLET PRESSURIZED L lQ U l D ENTERS T A N G E N T I A L L Y CONE A N G L E
(4) LIQUID MOVES INWARD A N D UPWARD AS S P I R A L L I N G VORTEX
SUSPENDED SOLIDS DRIVEN TOWARD W A L L A N D DOWNWARD I N ACCELERATING SPIRAL
UNDERFLOW (SOLIDS A N D LIQUID)
Fig. 15-58. Cyclone centrifugal separator
In Fig. 15-55, the underflow rate is controlled by valve number ( I ) . Valve number ( 2 ) , a pump internal relief valve, is used to adjust the pressure drop across the cyclones to 40-50 psi, which produces optimum rates through the cyclones. This unit is designed to continually recycle approximately of the cleaned power fluid back through the cyclones for additional solids removal. In Figs. 15-56 and 15-57 the underflow is controlled by valve number ( I ) . Cyclone internals are changed to obtain pressure drop for wells with different flow rates. Disposing of the underflow (solids) can be a problem on low-volume wells. In the Kobe Solo Unit system, if the underflow is set at a rate that is constantly, or even occasionally, greater than the production rate, the excess will go back to the vessel and has to be separated by the cyclones a second time. In the Oilmaster Unidraulic system, these conditions can cause the underflow to be shut-off and thereby cause the cyclone to be ineffective. In the Guiberson Econodraulic system, these conditions can cause the level in the vessel to drop.
4
693 POWER FLUID RATE CALCULATIONS
The pump specification tables list the engine- and pump-end displacement factors, q1 and q4, in bbl/D per SPM. At 100% efficiency, the power fluid rate would be ( q1 X S P M ) . Usual practice is to assume 90% engine-end efficiency so that power fluid rate'Q, is equal to: =
q1 x S P M 0.90
(15-5)
The pump-end displacement is ( q4 X S P M ) . As discussed in a previous section, one must take into account the formation volume factor, B, and usually assume 85% pump-end efficiency. Pump-end displacement, therefore, becomes: q4 x SPM
=
QsB + Qi 0.85
(15-6)
(In eq. 15-1, an additional factor, 0.85, was used in order to choose a pump that would operate at less than 85% of its maximum rated speed.) By combining eqs. 15-3 and 15-6, one obtains eq. 15-4:
POWER FLUID PRESSURE CALCULATIONS
The various pressures, friction losses, and fluid densities involved in CPF and OPF systems are shown in Fig. 15-59. Figure 15-60 illustrates those cross-section areas of the Kobe Type A pump which are involved with the various pressures. (Other pumps have different configurations.) On adding the upstroke forces on this pump and assigning the plus sign to upward-acting forces, one obtains: -PIA,
-P2
-
'Pi
-AR)
-P3
(P1 - P 2 ) ( A E - A R ) - ( P 3 - P 4 ) ( A P - A R )
(
- A R ) 'P4
(
-AR) + p l A R =
=
Pump friction, Fp,is not shown in Fig. 15-60, because it does not operate against an area. It is a function of pump speed and fluid passageways in the pump. Inasmuch as it opposes motion it will have a negative sign and the above equation becomes:
PRESSURES & FRICTION LOSSES AFFECTING HYDRAULIC PUMPS
694
m W P
8
m
c
I
8
0
u
P,
c
-
ln 2
'L-
=
Surface Power Return Back Pressure. psi
PFL
=
Surface Flow Line Back Pressure, psi
m
Y
m
"
e
v) 3
I
E
PPR
c
Surface Operating Pressure, PSI
a:
=
8
C
F1, Fp, F g = Fluid Friction in Tubing, psi Fp = Friction in Pump, psi
GI,
G2, G3. G4
~
Fluid Gradient rn
E
C
5
E
=-
u
Pump Submergence, ft.
P
=
P
h4
ln
Pump Setting Depth, ft. P
=
5
hl
L
P
P
f r
'1 ;1
i
1
'1
,-
h4G4
Fig. 15-59. Pressures and friction losses affecting hydraulic pumps.
PFL
1
=
+
U
P4
P4 = h4G4
m
h l G 3 + F3
+
=
5
P3
c
+
U
P3
5-
I
=
1
5 U
U
c?
c U
a" m
+
U
P
+
'7,
m
I,
Closed Power Fluid System
P2
tI
Open Power Fluid System
695
PRESSURES ACTING ON A KOBE TYPE A PUMP
Net Cross-section Area
AR
AE - AR
AR
Fig. 15-60,Pressures acting on a Kobe type A pump.
Pressure Upstroke
Pressure Downstroke
696
The quantity [ ( A , - A R ) / ( A , - A R ) ] is the ratio of net pump area to net engine area and, for this pump, is the same for the downstroke and the upstroke. This quantity is sometimes referred to as the pressure ratio, but more commonly it is called the P over E ratio. The numerical values for P / E ratio are listed in the specification tables. Inasmuch as the algebraic equation relating pump areas to P / E ratio is different for different pumps, one must substitute P / E in the above equation to arrive at the following general CPF equation for all hydraulic pumps: (15-7) Some pumps have an additional term in their CPF equation that is a function of ( p z - p 3 ) , but this term is always ignored because it is very small. For Fig. 15-13,
the term is:
Inasmuch as the value of ( p 2 - p 3 ) is usually less than 500 psi, this term is usually less than 35 psi. (For OPF system p 2 = p 3 , so this term drops out.) As shown in Fig. 15-59: PI
=hF1-
Fl + Ps
P2 = hlG,
+ F2 + Ppr
P3 = hlG4
+ ‘3 +Pwh
P4
= h4G4
Substituting these relationships into eq. 15-7 and assuming fluid friction, F,, F2, and F3, and wellhead back pressure, ppr and pwh to be zero, eq. 15-7 becomes eq. 15-2:
p , - ( hlG4 - h 4 G 4 )P / E - Fp = 0 T h s is the equation which was used in a previous section to get a first approximation of surface pressure and to arrive at a maximum value for P/E (eq. 15-3). Inasmuch as in the OPF system p 2 = p 3 , the general OPF equation for all hydraulic pumps becomes:
(PI - P 3 )
- (P3 -P4)P/E-
Fp=0
where (as shown in Fig. 15-59):
PI
= hlG1 -
P3 = h l G 3
P4
= h4G4
4 + Ps
+ F3 +Pwh
(15-8)
691
PRESSURE INCREASE DUE TO MECHANICAL AND HYDRAULIC FRICTION I N PUMP AND BOTTOM HOLE ASSEMBLY vs PERCENT OF RATED SPEED 200c
1500
1000 9oc
800 700
600
500 400
300
EXAMPLE
200
PUMP - Flg 15 - 13
SPEED - 50% VISCOSITY - 1 0 C S SP GR. - 0 8 A P = 270 x 8 = 216 100
PERCENT OF R A T E D SPEED
FIG
13- 1 7 , 1 5 - 1 9
FIG 1 5 - 1 6 t l 5 - 2 4 FIG 1 5 - 1 5 , 1 5 - 2 3
FIG
is-
12,15-22
FIG 15- 1
7
15-18,15-21
Fig. 15-61. Relationship between the percent of rated speed and pressure increase due to mechanical and hydraulic friction in pump and bottomhole assembly.
PSI
0
20
40
60
80
100
120 140
160
B3 200 220 240 260 280 300
Temperature - degrees Fahrenheit Gtavity
API 9.5
7
10.7 14.0 15.0 19.8 25.6 26.8
-
Field Boscan, Venezuela Boscan, Venezuela Maricopa, Calif. Wilrnington, Calif. Sansinena, Calif. Scholern Alechern, Okla. Seal Beach, Calif.
Line
Gravity ’API
8 9 10 11 12 13
306 31 1 364 346 440 50 7
Fig. 15-62. Relationship between viscosity of crude oils and temperature
Field Ventura, Calif. Kettleman Hills, Calif. Oklahoma City, Okla. Kettlernan Hills, Calif. Denton, New Mexico Kettlernan Hills, Calif.
699
100
0
200
Temperature -degrees Fahrenheit
Fig. 15-63, Relationship between viscosity of water and temperature at 0 psi
The unknowns in eq. 15-8 are pump friction, Fp;fluid friction, Fl and F,; and the fluid gradient in the return column, G,. Pump friction, Fp, is obtained from Fig. 15-61 after obtaining SPM from eq. 15-6. The fluid gradient in the return column is calculated after obtaining the ratio of power fluid flow rate, Q,, and produced fluid flow rate, (Q, Q 6 ) ,in this column. Thus, it is necessary first to calculate Q,, by using eq. 15-5 or 15-4. Fluid friction in the tubing is then found by use of Figs. 15-65 through 15-98. The value of p 3 may be obtained from multiphase flow corrections; however, in this chapter the equation p , = h,G3 + F3 +pwh is used. This equation does not account for gas in the return column, but it offers a conservative procedure that will provide a safe design. Pump friction is obtained from Fig. 15-61, which represents the mechanical and hydr'aulic friction in the pump. From the curves in Figs. 15-62 and 15-63, the power fluid viscosity at the bottomhole temperature can be obtained to use with the pump friction chart. Conversions from API gravity to specific gravity can be made from Fig. 15-64. The values obtained from Fig. 15-61 show maximum values based on the largest pump piston operating at 100% pump-end efficiency. When the fluid flow rate through the pump-end is reduced by smaller pistons, or by gas, the total friction will be somewhat lower than that predicted by the chart. This discrepancy is due to
+
700 G r a v i t i e s a n d U n i t Pressure of Oil Columns
Specific
Note-First line opposlte each API gravity
IS
sp gr at 60'F. Second line is column pressure in psi/ft
-
~~~
Degrees A P I
0
I
.2
.3
4
.5
6
7
.8
.9
10000 4331
9993 43'8
9986 4325
9979 4322
9972 4319
9965 4316
9958 4313
9951 4310
9944 4307
,9937 4304
11
9930 4301
9923 4295
9916 4295
9909 4212
9902 4289
9895 4286
9888 4282
9881 4279
9674 4276
,9888 ,4274
12
9861 4271
9854 4268
9847 4265
9840 4262
9833 4259
9826 4256
9820 4253
9813 4250
9808 4247
,9799 4244
9792
10
13
424'
9786 "238
9779 $235
9772 4232
9765 4229
9759 4226
9752 4224
9745 4221
9738 4218
,9732 4215
14
9725 $212
9718 4209
9712 4.06
9705 4203
9696 4200
9692 4198
9685 4195
9679 4192
9672 4189
,9685 4186
15
9659 ~183
9652 41.80
9646 4178
9639 4175
9632 4172
9626 4169
9619 4166
9613 4163
9606 4160
,9600 4158
16
9593 4155
9587 4152
9580 4139
9574 *'46
9567 4143
9561 4141
9554 4138
9548 4135
9541 4132
9535 4130
17
,9529 4127
9522 4124
6121
9509 4118
9503 4116
9497 4113
9490 4110
9484 4108
9478 4105
9471 4102
,9465
,9459 4097
9452 4094
9446 4091
9440 4088
9433 4085
9427 4083
9421 4080
9415 4078
9408 4075
9383 4361
9377 4061
9371 4059
9365 4056
9358 4053
9352 4050
9346 4048
9321 4037
9315 4034
9309 4032
9303 4029
9297 4027
9291 4024
9285 4021
9260 501 I
9254 4038
9248 4005
9242 4003
9236 4000
9230 3998
9224 3995
9200 3985
9194 3982
9188 3979
9182 3977
9176 3974
9170 3972
9165 3969
18
4099
9516
19
,9402 4072
,9398
9390
4069
C067
20
9340 4045
,9334 4043
GO40
-.
~...~. ~
~
9328 ~~~~
__
~
21
9279 4015
,9273
9267
4016
4014
22
9216 3932
9212 3933
3987
9159 3967
9153 396.:
??6?
9141 3959
9135 3956
9129 3954
9123 3951
9117 3949
9111 3946
9108 3944
24
,9100 394;
,9094 3933
9088 3936
9082 3333
9076 3931
9071 3929
9065 3926
9059 3923
9053 3921
9047 3918
25
,9042 39:6
9036 3913
,9030
I
9024 3x8
9016 3906
9013 3934
9007 3901
9001 3898
8996 3896
8990 3894
26
8984 339!
8978 3858
8973 3886
8967 3883
8961 3881
6956 3579
8950 3876
8944 3874
8939 3871
8933 3869
27
8927 3566
,8922 3861
,8916 3832
891 1 3859
8905 3857
8899 3854
8894 3852
8888 3849
8883 3847
8877 3845
28
8871 3847
,8668 38;0
8860
3837
8855 3835
8849 3833
6844 3830
8838 3828
8833 3826
8827 3823
8822 3821
29
,8816 3818
3815
8605 3813
8800 381 I
8794 3809
8769 3807
8783 3804
8779 3802
8772 3799
8767 3797
6756 3792
,8751 3790
8745 3787
8740 3785
8735 3783
8729 3781
8724 3778
8718 3776
8713 3774
8686
23
__
30
9762
3795 8708
981 1
9206 9147
33!
I
31
3'71
8702 3/60
6697 3767
8692 3'65
3762
8681 3760
8676 3758
8670 3755
8685 3753
8660 3751
32
8654 3748
8649 3746
8644 3744
8639 3742
8833 3739
9628 3737
8623 3735
8618 3732
8612 3730
8607 3728
33
6602 3726
8597 3723
8591 3771
8586 3719
8581 3716
8576 3714
8571 3712
8585 3710
8560 3707
8555 3705
34
35
8550
8545
8540
8534
3703
370'
3b99
3696
8529 3694
8524 3692
8519 3690
8514 3687
8509 3685
8504 3683
8498 3680
6493
3678
9488 3676
8483 3674
8478 3672
8473 3670
8468 3667
8463 3665
8458 3663
3661
8453
Fig. 15-64a Conversion table-'API gravity (10-35) to specific gravity and pressure gradient in psi/ft.
701 S p e c i f i c G r a v i t i e s and U n i t Pressure of O i l C o l u m n s Note-First line opposite each API gravity is s p gr at 60°F. Second line is column pressure in Psl/tt
.4
.5
.6
.7
.I
.9
,8423
,8413
,8408 ,3642
,8403 ,3639 ,8353 ,3618
___ 38
,6448 ,3659
,3657
,8438 ,3654
,4433 ,3652
.M28 ,3650
,3648
.a418 ,3646
37
.098 ,3637
,8393 ,3635
,8388 ,3633
,8383
,8373 3626
,8368 ,3624
,063 3622
,8358
,3631
,8378 3629
.ma
,8343 ,3613
,8338 ,3611
,033 3609
,83211 ,3607
,8324 ,3605
,8119 3603
,8314 3601
,8309 ,3599
,8304
,3616
39
,8299 ,3594
,8294 ,3592
,8289 ,3590
,8285 ,3588
.I280 3586
,8275 ,3584
,8270 ,3582
,8265 ,3580
,1260 ,3577
,8256 ,3576
40
,8251 ,3574
,8248 ,3571
,8241 ,3569
,8236 ,3567
,8232 ,3565
,8227 ,3563
,8222 ,3561
,8217 3559
,8212 3557
,8208 ,3555
,8203 ,3553
.Il9B ,3551
,8193 3548
,8189 3547
,8184 3544
,8179 ,3542
,8174 3540
,8170 3538
,8165 ,3536
,8160 3534
,8155 ,3532
,8151 ,3530
,8146 3528
,8142 3526
,8137 3524
,8132 3522
,8128 ,3520
,8123 3518
,8118 3516
,8114 3514
,8109 3512
,8104 ,3510
,8100 3508
8095 3506
,8090 3504
,8086 3502
,8081 3500
,3016 ,3498
,8072 3496
,8067 ,3494
,3063 ,3492
,8058 ,3490
.SO54 3488
,8049 3486
,8044 3484
,8040 3482
,8035 3480
.5031 ,3478
,8026 ,3476
,8022 ,3474
.a017 ,3472
,8012 3470
,8008 3468
,8003 3466
,7999 3454
,7994 ,3462
,7990 3460
,7985
,3458
,7981 ,3457
,7976 ,3554
46
,7972 ,3453
,7987 ,3451
,7963 ,3449
,7958 ,3447
,7954 ,3445
,7949 3443
,7945 3441
,7941 3439
,7936 3437
,7932 ,3435
47
,7927 ,3433
,7923 ,3431
,7919 ,3429
,7914 3428
,7909 ,3425
,7905 ,3424
,7901 ,3422
,7896 3420
,7892 3918
,1887 3416
48
.7883 ,3414
,7879 3412
,7874 ,3410
,7870 ,3408
,7865 ,3406
,7861 3405
,7857 ,3403
,7852 3401
7848 3399
.1a44 3397
49
,7839 ,3395
,7835
,7831 ,3392
,7826 ,3389
,7822 ,3388
,7818
,3393
,3386
,7813 3384
,7809 3382
,7805 3380
,7800 3378
50
,7796 ,3376
,7792 ,3375
,7788 ,3373
,7783 ,3371
,7779 ,3369
,7775 3367
,7770 ,3365
,7766 3363
7762 3362
38
,6443
,3644
,3620
,3596
,7758
,3360 -
51
,7753 3358
,7749 ,3356
,7745 ,3354
,7741 3353
,7736 3350
,7732 3349
7128 3347
,7724 3345
,7720 3344
,7715 ,3341
52
,7711 3340
,7707
,7703 3336
,7699 3334
,7684 3332
,7690 3331
,7686 3329
,7582 3327
7618 3325
,7674 ,3324
53
,7669 3321
,7665
,3320
,7861 ,3318
,7657 3316
,7653 ,3315
7649 3313
,7645 3311
,7640 3309
,7636 3307
,7632 3305
,7628
,7824
,7820
3304
,3302
,3300
,7816 3298
,7612 ,3297
,7608 3295
,7603 3293
,7599 3291
,7595 3289
,7591 328P
.7W 3286
,7583 ,3284
,7579 ,3282
,7575 3281
,7571 ,3279
,7567 3277
,7563 3276
7559 3274
,7555 3272
,7551 ,3270
,7527 3260
,7523 ,3258
,7519 ,3256
,7515 3255
,7511 3253
54
55
,3338
~
56
,7547 ,3269
,7543 3267
,7539 ,3265
,7535 ,3263
,7531
57
,7507 ,3251
,7503 ,3250
,7499 ,3248
,7495 ,3246
,7491
,3244
,7487 3243
,7483 ,3241
,7479 ,3239
,7475 3237
,3236
58
,7467 ,3234
,7463 ,3232
,7459
,7455 ,3229
,7451 ,3227
,7447 3225
,7443
,3230
,3224
,7440 ,3222
,7436 ,3221
,7431 ,3219
59
,7428 ,3217
,7424 ,3215
,7420 ,3214
,7416 ,3212
,7412 ,3210
,7408 3208
,7405 ,3207
,7401 ,3205
,7397 ,3204
,7393 ,3202
.7389
,7385 ,3198
,7381 ,3197
,7377 ,3195
,7374 ,3194
,7170 3192
,7366 ,3190
,7362 ,3188
,7358 ,3187
,7354 ,3185
80
,3200
,3262
,7471
Fig. 15-64b. Conversion table-"API gravity (36-60) to specific gravity and pressure gradient in psi/ft.
702
PRESSURE DROP IN PIPE ’?” STANDARD
PIPE (0.62” I.D.)
Fig. 15-65. Relationship between the fluid flow and pressure drop in a +-in. standard pipe (0.62 in. I.D.).
the fact that approximately 25% of the total friction is in the pump-end of the pump. This value is not well documented for all pumps, but can be used to estimate the reduction in pump friction due to actual pump-end liquid flow rate. In equation form, the A p from Fig. 15-61 is equal to:
where Fee= engine-end friction = 0.75 A p , and Fpe= pump-end friction = 0.25 A p .
703
PRESSURE DROP IN PIPE 3h” STANDARD PIPE (0.82’’ I.D.)
05
01
10
50
PRESSURE DROP-PS1/1000’-MULTIPLY
10
50
0
BY SPEC GRAV
Fig. 15-66. Relationship between the fluid flow and pressure drop in a :-in. standard pipe (0.82 in. I.D.).
In the example shown in Fig. 15-61, the A p is 216 psi; therefore:
Fee = 0.75 X 216 = 162 psi Fpe= 0.25
X
216 = 54 psi
If it is a 2t-in. pump from Fig. 15-13 (29 in.X l a h-1: in. pump-the largest 2t-in. pump), which is operating at 100% pump-end efficiency, the Fpe= 54 psi is correct. If, however, it is a 2 t in. x l a in-1; in. pump operating at 80% pump-end
704
PRESSURE DROP IN PIPE 1” STANDARD PIPE (1.05” 1.0.) 1000
100
-
il
u N
P
n a m
I
%
s P
3 10
10
PRESSURE DROP-pS1/1000 -MULTIPLY BY SPEC G R A V
Fig. 15-67. Relationship between the fluid flow and pressure drop in a I-in. standard pipe (1.05 in. I.D.).
efficiency, Q4 will be less than that used to construct the chart. Because the correction to Fpe is a small quantity, it is customary to ignore it. The error thus introduced is on the safe side. On assuming: Fp= A p (from Fig. 15-61)
the procedure for calculating ps in the OPF system is:
705
PRESSURE DROP IN TUBING 1-%”EU API TUBING 1000
100
-
Y
0 N
z
n
m 1
3
s 0 LL J
10
10
PRESSURE DROP-PSI/lOOO
-MULTIPLY
BY SPEC G R A V
Fig 15-68 Relatlonahp between the fluid flow and pressure drop in a l i - i n EU API tublng
Find Step 1: Formation volume factor, B or Step 1: Formation shrinkage factor, S Step 2 : Pump speed, SPM Step 3: Pump friction, Fp Step 4: Power fluid rate, Q1 Step 5 : Power fluid tubing friction, Fl
From Figs. 15-27, 15-28 or 15-29 Fig. 15-30 eq. 15-6 Figs. 15-61 and 15-62 or 15-63 eq. 15-4 or 15-5 Figs. 15-65 through 15-98
706
PRESSURE DROP IN TUBING 1-kz” EU API TUBING
0.1
05
50
10
PRESSURE DROP-PSl/lOOO’-MULTIPLY
10
50
BY SPEC G R A V
Fig. 15-69. Relationshp between the fluid flow and pressure drop in a 1:-in. EU API tubing.
Step 6: Return fluid gradient, G, Step 7: Return fluid tubing friction, F3
GiQi + G5Q5 + G6Q6 G, = Qi + Q, + 426 Figs. 15-65 through 15-98
One can now calculate p , after rearranging eq. 15-8 to: Step 8: ps = p 3 ( l + P / E ) - p 4 ( P / E )
+ Fp- hlGl + F,
too
707
PRESSURE DROP IN TUBING 2-1/16" API TUBING 10 000
1000
Y U N
f
n
m n I
B
s D
3
100
10
05
01
50
10
PRESSURE DROP-PSI/lOOO'-MULTIPLY
10
50
100
BY SPEC GRAV
Fig. 15-70. Relationship between the fluid flow and pressure drop in a 2&-in. API tubing.
+
Inasmuch as Q4 = Q , Q 6 , where Q, is the flow rate of produced oil and Q6 is the flow rate of produced water: G, = G5Q5 G 6 Q 6 ,where Q, Q 5 + Q6 +
+ Q6
= Q4
(15-9)
708
PRESSURE DROP IN TUBING 2-910" EU API TUBING
10 000
1000
Y c7 N
9 0 a m I
3
s 0
3 100
10 01
05
10
50
PRESSURE DROP-PS1/1000'-MULTIPLY
10
50
BY SPEC. GRAV
Fig. 15-71. Relationship between the fluid flow and pressure drop in a 22-in. EU API tubing.
Example problem 15-4 Given: Q, = 200 bbl/D of 40"API gravity oil Q6= 100 bbl/D of 1.03 sp. gr. water p 4 = 600 psi h , = 10,000 ft Power fluid: 40"API gravity oil pwh= 1 5 psi Bottomhole temperature = 180°F
100
709
PRESSURE DROP IN TUBING 2-78'' EU API TUBING
' -
---
I
0.1
0.5
1.o
5.0
PRESSURE DROP-PSl/lOOO'-MULTIPLY
10
50
0
BY SPEC. GRAV
Fig. 15-72. Relationship between the fluid flow and pressure drop in a 2 i - h . EU API tubing.
Assume: 252016 pump from Fig. 15-20 OPF parallel free type installation using 24-in. OD power tubing and 2i-in. OD return tubing, with gas to be vented to the surface through the casing annulus. B=l.
Find: (A) Surface horsepower using first approximation method. (B) Surface horsepower using method outlined in this section. From Fig. 15-20: P / E = 0.64, q 1 = 16.5 bbl/D per SPM, q4 = 10.6 bbl/D per SPM.
710 PRESSURE DROP IN TUBING 3-%” EU API TUBING 10 000
1000 I
Y u N
I 0
a
m I 3
s D LL _I
100
10
01
05
10
50
PRESSURE DROP-PSI/lOOO’-MULTIPLY
10
50
BY SPEC GRAV
Fig. 15-73. Relationshp between the fluid flow and pressure drop in a 3:-in. EU API tubing.
Solution: (A) From eq. 15-9: G,
=
(0.8251) (0.433)(200) + (1.03) (0.433)(100) 300
From eq. 15-2:
p , = (0.386 X 10,000 - 600)0.64 + 400 = 2490 psi
= 0.386
psi/ft
100
71 1
PRESSURE DROP IN TUBING 4-'P EU API TUBING
0.1
05
10
50
PRESSURE DROP-PS1/1000'-MULTIPLY
10
50
BY SPEC GRAV
Fig. 15-74. Relationship between the fluid flow and pressure drop in a 4:-in. EU API tubing.
From eq. 15-4: 300 16*5 0.85 x 0.90 x 10.6
= 610
bbl/D
Hydraulic horsepower = 2450 X 610 x 0.000017 = 25.4 HP (B) Step 1: was given; B = 1.0 Step 2: From eq. 15-6: SPM =
300 10.6 X 0.85
= 33.3
SPM
0
712
PRESSURE DROP IN TUBING FLOW BETWEEN P a ' ' EU TUBING & %" (1.050" O.D.) PIPE 1000
100
-
Y a
N
If_ 0
m
I
z
s 2
3 10
10 01
05
10
50
PRESSURE DROP-PSI/lOOO'-MULTIPLY
10
50
100
BY SPEC GRAV
Fig. 15-75.Relationship between the fluid flow and pressure drop when flow occurs between 2i-in. EU tubing and :-in. pipe (1.050 in. O.D.).
Step 3: From Fig. 15-62, viscosity of 40"API gravity oil at 180°F is between lines 11 and 12 at approximately 1.5 centistokes. From Fig. 15-64, specific gravity is equal to 0.825, whereas from Fig. 15-20, rated speed is 51 SPM. Percent rated speed = 33.3/51 X 100 = 65% From Fig. 15-61, A p is equal to approximately 410 psi, whch if multiplied by specific gravity gives Fp:
Fp= 410 X 0.825 = 338 psi
713
PRESSURE DROP IN TUBING FLOW BETWEEN 2-'a" EU TUBING L 1" PIPE ( 1 . 3 l S ' O . D . )
0.5
0.1
1.o
5.0
PRESSURE DROP-PSl/lOOO'-MULTIPLY
10
50
100
BY SPEC. GRAV.
Fig. 15-76. Relationship between the fluid flow and pressure drop when flow occurs between 2i-in. EU tubing and 1-in. pipe (1.315 in. O.D.).
Step 4: From eq. 15-5:
33.3 X 16.5 0.9
= 610
bbl/D
Step 5 : In step 3, viscosity of the power fluid was found at bottomhole temperature; however, to be precise for tubing friction calculations, one should estimate the average temperature of the fluid from the bottom to the surface. If the fluid reaches
714
PRESSURE DROP IN TUBING FLOW BETWEEN 2 - 7 w 3 II
01
05
10
i
-
1
EU ~ API TUBING
50
PRESSURE DROP-PS111000'-MULTIPLY
10
50
100
BY SPEC GRAV
Fig. 15-77. Relationship between the fluid flow and pressure drop when flow occurs between 2 i - h and 1f -in. EU API tubing.
the surface at loo", which will give an average temperature of 140"F, from Fig. 15-62 the average oil viscosity is found to be approximately 2.1 centistokes. From Fig. 15-72, at Q, = 610 bbl/D and oil viscosity of 2.1 centistokes, the pressure drop is approximately equal to 1.6 psi per 1000 ft. Upon multiplying it by specific gravity, Fl is found to be: F,
= 1.6 X
10 X 0.825 = 13 psi
715 PRESSURE DROP IN TUBING FLOW BETWEEN 3-”2”&1-”r”EU APl TUBING
05
10
50
PRESSURE DROP-PS1/1000’-MULTIPLY
10
50
100
BY SPEC GRAV
Fig. 15-78. Relationship between the fluid flow and pressure drop when flow occurs between 3:-in. and 1:-in. EU API tubing.
Step 6:
G, =
(610) (0 3 2 5 ) (0.433)
= 0.367
+ (200) (0.8250) (0.433) + (100)(1.03) (0.433) 610 + 300
psi/ft
Step 7: The viscosity of water at 140°F from Fig. 15-63 is equal to 0.46 centistokes.
716
PRESSURE DROP IN TUBING FLOW BETWEEN 3-'z"6 1 - ' 2 " EU API TUBING
I
"
I
01
05
10
50
PRESSURE DROP-PSI/lOOO'-MULTIPLY
10
50
1 I0
BY SPEC, G R A V
Fig. 15-79. Relationship between the fluid flow and pressure drop when flow occurs between 3i-in. and 1:-in. EU API tubing.
The average viscosity in the return string, v3, is equal to:
v3 =
(610)(2.1) + (200)(2.1) + (100)(0.46) 910
= 1.92 centistokes
From Fig. 15-71, at 910 bbl/D and viscosity of 1.92 centistokes, the tubing friction in the return string of 2i-in. tubing is 11 psi per 1000 ft. Upon multiplying it by
PRESSURE DROP IN TUBING FLOW BETWEEN 3-w a z-J/s~* EU API TUBING 1000
100
-
v)
a
l3 N
0 0
a m I
B
s P
3 10
10
Fig. 15-80. Relationship between the fluid flow and pressure drop when flow occurs between 3;-in. and 2i-in. EU API tubing.
specific gravity, F3 is found to be:
F3 = 11 X 10 X
0.367 0.433
= 93 psi
Step 8:
p 3 = (10,000 X 0.367)
+ 93 + 75 = 3838 psi
718
PRESSURE DROP IN TUBING FLOW BETWEEN 4 - 1 ~ 3 . 32-31~33 EU API TUBING 10.000
10
Fig. 15-81. Relationship between the fluid flow and pressure drop when flow occurs between 4:-in. and Zi-in. EU API tubing.
From eqs. 15-8 and 15-9:
ps = (3838)(1.64) - (600)(0.64) + 338 - (10,000 X 0.825 X 0.433)
+ 13 = 2690 psi
Hydraulic horsepower = 2690 X 610 X 0.000017 = 27.9 HP By using the first approximation method (Solution A), one arrives at a surface hydraulic horsepower requirement that is about 10% low. [One could easily add pWh(l P / E ) = 75(1.64) = 123 psi and improve the accuracy.] If the first approxi-
+
119
PRESSURE DROP IN TUBING FLOW BETWEEN 4-‘2” 6 2 . ’ ~ ” EU API TUBING 10 000
1000
s
Y
5 n P m
I
3
E: P
2
100
10
0.5
1.o
5.0
PRESSURE DROP-PS111000’-MULTIPLY
1’0
50
100
BY SPEC. GRAV.
Fig. 15-82. Relationship between the fluid flow and pressure drop when flow occurs between 4:-in. and 2;-in. EU API tubing.
mation method yields a power requirement 20% less than a standard size surface pump, it is usually safe enough to ignore the more precise calculations outlined in this section. Class problems
(1) Given: Same problem as example problem 15-4, except power fluid is water having specific gravity of 1.03. Find: Surface hydraulic horsepower and compare to example problem 15-4A and 15-4B.
720
PRESSURE DROP IN TUBING FLOW BETWEEN S”,l a # CSG. 6 2-’/a” EU API TUBING 10 ooc
1 OO(
-
Y 0
N
p1
n
a I
m
3
s 0 Y _I
1O(
11
Fig. 15-83. Relationship between the fluid flow and pressure drop when flow occurs between Sin., 18-lb casing and 2 i - h EU API tubing.
( 2 ) Giuen: Same problem as class problem 1 above, except: return tubing string is 1$ in. in diameter. Find: Surface hydraulic horsepower. BOTTOMHOLE PRESSURE CALCULATIONS
Equations 15-7 and 15-8 are frequently used to calculate the flowing bottomhole pressure, p 4 , in existing hydraulic pumping installations. This calculation is similar
721
PRESSURE DROP IN TUBING FLOW BETWEEN 5-”2”, 20# CSG. 6 2-%” O.D. EU TUBING
0.1
05
10
50
PRESSURE DROP-PSI/lOOO’-MULTIPLY
10
50
100
BY SPEC G R A V
Fig. 15-84. Relationship between the fluid flow rate and pressure drop when flow occurs between Sf-in., 20-lb casing and 2;-in. O.D. EU tubings.
to finding p,, except steps 1, 2, 3 and 4 are eliminated because SPM and Q, are given. Often the last-stroke method is used. The procedure for this method is to close the valve on the power fluid supply line to the well and record the pressure when the pump stops stroking. This last-stroke pressure is the operating pressure at zero pump speed and zero fluid flow. It is the pressure where Fp, Fl and F3 are zero. Inasmuch as it takes less than a minute for the pump to stop stroking, p4 and h,G, do not change appreciably from their producing values. If p , , is the last-stroke
122
PRESSURE DROP IN TUBING FLOW BETWEEN 5-’2’’,
-
14# CSG. (L 2-’a”
O.D.EU TUBING
I
0.1
1 .o
0.5
5.0
PRESSURE DROP-PSl/lOOO’-MULTIPLY
10
50
3
BY SPEC. G R A V .
Fig. 15-85. Relationshp between the fluid flow rate and pressure drop when flow occurs between 5+-in., 14-lb casing and 2;-in. O.D. EU tubing.
pressure, eqs. 15-7 and 15-8 become:
For CPF system: p4 =
P I S
- Ppr
~
P/E
- hlG4 -pwh
(15-10)
(15-11)
123
PRESSURE DROP I N TUBING FLOW BETWEEN S - l r " . 22# CSG. & 2-'s" EU TUBING
0
Fig. 15-86. Relationshp between the fluid flow rate and pressure drop when flow occurs between the 5;-in., 22-lb casing and 2 i - h . EU tubing.
Use of these equations eliminates the need for calculating the various friction losses in the system. This last-stroke method must not be used with a severely worn pump, however, because fluid slippage in the pump-end of the pump can cause the pump to continue strohng past its balance pressure. If precise bottomhole pressure data is required, a pressure bomb can be attached to most hydraulic pumps to record actual producing bottomhole pressures while the pump is operating. This method is preferred to calculations based on eqs. 15-7, 15-8,
124
PRESSURE DROP IN TUBING FLOW BETWEEN 5 - 3 4 ' , , 22# CSG. & 3-12'' EU TUBING 10 000
1000
Y u N
z
n
n
1
3
s il LL J
100
10 01
05
10
50
10
50
100
PRESSURE DROP-PS1/1000 -MULTIPLY BY SPEC GRAV
Fig. 15-87. Relationship between the fluid flow rate and pressure drop when flow occurs between the 5:-in., 22-lb casing and 3;-in. EU tubing.
15-10, or 15-11. The principal sources of error in calculations using these equations are: (1) Pump friction values are obtained using eqs. 15-7 and 15-8. Values for Fp using Fig. 15-61 are sometimes 200-400 psi high. This is due to the fact that pump friction varies with clearance, concentricity, and surface finish of the sliding parts, which vary due to manufacturing tolerances and to pump wear. The values shown on the chart are purposely on the high side of the range to allow a safety factor for
125
PRESSURE DROP IN TUBING FLOW BETWEEN 8 - 5 8 " , 28# CSG. 6 3-'2'' EU TUBING 10 000
01
05
10
50
10
50
0
PRESSURE DROP--PS1/1000 -MULTIPLY BY SPEC G R A V
Fig, 15-88. Relationship between the fluid flow rate and pressure drop when flow occurs between the 6:-in., 28-lb casing and 3i-in. EU tubing.
design purposes. Using eqs. 15-10 and 15-11, solves this problem, but G, and G, become sources of error as explained below. (2) When gas is present in the production column, multiphase flow correlations must be used with eqs. 15-7 and 15-8, because even a 300 scf/bbl, gas/liquid ratio will reduce p 3 by 600-1200 psi at depths below 5000 ft. Although this error provides a safety factor when designing an installation, it is intolerable for bottomhole pressure calculations. Equations 15-10 and 15-11 eliminate the error in Fp, but they also eliminate tubing friction, F3. Inasmuch as multiphase flow correlations include
7 26
PRESSURE DROP IN TUBING FLOW BETWEEN 6-50", 28# CSG. & 4-'r" E U TUBING 10 000
01
05
10
50
PRESSURE DROP-PSI/lOOO'-MULTIPLY
10
50
0
BY SPEC GRAV
Fig. 15-89. Relationship between the fluid flow rate and pressure drop when flow occurs between the 6:-in., 28-lb casing and 4:-in. EU tubing.
tubing friction. a separate calculation must be made to determine tubing friction, which must then be subtracted out of the multiphase flow value. SAMPLE PROBLEM AND QUESTIONS
(1) The following well is to be equipped with a subsurface hydraulic bottomhole piston pump, whch will operate off of the existing lease open-power oil system. A The help extended by Charles E. Arnold is indeed greatly appreciated by the authors
727
PRESSURE DROP IN TUBING FLOW BETWEEN 7”, 26# CSG. h 2%” O.D. EU TUBING 100,000
10,000
-
u5
2a N
P n 0
m
I
e
s P
3 LL
1ooc
lot 0.1
0.5
1 .o
5.0
PRESSURE DROP-PSI/IOOO’-MULTIPLY
10
50
100
BY SPECIFIC GRAVITY
Fig. 15-90. Relationship between the fluid flow rate and pressure drop when flow occurs between the 7-in., 26-lb casing and 2;-in. O.D. EU tubing.
casting type free pump installation is desired using available 2i-in. O.D. API tubing. Design the whole system. Given : (a) Well depth (both vertical and measured) = 6000 ft (b) Pump setting depth = 5900 f t (c) Pumping fluid level (from surface) = 5000 ft (d) Gross production = 400 bbl/day (e) Water cut = 25%
728
PRESSURE DROP IN TUBING FLOW BETWEEN 7”, 2 w CSG. a 2-7s” O.D. EU TUBING 10 000
1000
I_J n
a N
9 Q 0
m
I
3
2 0 Y2
1oc
1(
PRESSURE DROP-PSI/lOOO -MULTIPLY
BY SPEC GRAV
Fig 15-91 Relationship between the fluid flow rate and pressure drop uhen flou occurs between the 7-1n. 26-lb and 2 i - i n 0 D EU tubing
( f ) Oil gravity = 33” API (8) Oil viscosity at 100 OF
= 46 SSU (h) Water specific gravity = 1.05 (i) Gas/oil ratio = 0 6) Casing size and weight = 7-in., 32-lb (k) Flowline back pressure = 60 psi (1) Static fluid level (from surface) = 3700 ft
129
PRESSURE DROP IN TUBING FLOW BETWEEN 7”, 26# CSG. (I 3 - ’ P O.D. EU TUBING
3 PRESSURE DROP-PSl/lOOO’-MULTIPLY
BY SPEC. GRAV.
Fig. 15-92. Relationshp between the fluid flow rate and pressure drop when flow occurs between the 7-in., 26-lb casing and 3i-in. O.D. EU tubing.
Also solve this problem using gas/oil ratio of 200 cu ft/bbl. (2) What are the four basic components in every type of artificial lift system? (3) What is the primary advantage and disadvantage in a free-casing OPF hydraulic pump system? (4) What are the two basic power fluid systems? ( 5 ) Name the basic components of a hydraulic pumping system. (6) What are the two main disadvantages of a Jet Pump? (Consult literature on jet pumps, e.g., Trico Industries, Inc.)
730
PRESSURE DROP IN TUBING FLOW BETWEEN 7”, 30# CSG. 6 4-’h” EU TUBING 10.000
1000
-
r: ff N
P
0
I 3
s 0
3
100
10
05
10
50
PRESSURE DROP-PSI/lOOO-MULTIPLY
10
50
100
BY SPEC GRAV
Fig. 15-93. Relationship between the fluid flow rate and pressure drop when flow occurs between the 7-in., 30-lb casing and Of-in. EU tubing.
(7) Engine valve is shifted (check correct answer): (a) Mechanically by piston. (b) Hydraulically by valve rod. (c) Automatically by packer nose. (8) Normal power fluid loss expected in a hydraulic pump system (check correct answer): (a) None, (b) 20%,and (c) 5%. (9) Running the pump too fast for the well’s production often results in (check
731
PRESSURE DROP IN TUBING FLOW BETWEEN S’h”, 43%# CSG. 6 2%” O.D. EU TUBING 100,000
1
100 0.1
’ ’
I
PRESSURE DROP-PSI/lOOO-MULTIPLY
BY SPECIFIC GRAVITY
Fig. 15-94. Relationship between the fluid flow rate and pressure drop when flow occurs between the 9;-in., 43;-lb casing and 2;-in. O.D. EU tubing.
all that apply): (a) Pounded production valving. (b) “Governed” engine valve assembly. (c) More production than expected. (d) Excessive wear and damage to pump. (e) Short pump runs. (f) Pump unseating while running. (10) What is the hydraulic horsepower (HHP) if the pump requires 1250 bbl/D of power fluid at 3000 psi operating pressure?
732
PRESSURE DROP IN TUBING FLOW BETWEEN 8-5s”, 43-’r# CSG. 6 3 - l ~ ”O.D. E U TUBING
01
0.5
1 .o
5.0
PRESSURE DROP-PS1/100O-MULTIPLY
10
50
100
BY SPEC. GRAV.
Fig. 15-95. Relationship between the fluid flow rate and pressure drop when flow occurs between the 9;-in., 43i-lb casing and 3 i - h . O.D. EU tubing.
(11) When considering jet pump sizing: (a) What ratio gives high lift capacity? (1) “A” ratio, (2) “C” ratio, (3) “E” ratio, (4) “H” ratio. (b) What amount of submergence is probably necessary? (1) 5 % , (2) 20%, (3) 50%, (4) None. (12) If minimum surface horsepower is important, which type of pump should be chosen?
133
PRESSURE DROP IN TUBING FLOW BETWEEN 9-%", 43-'z# CSG. & 4-'z" O.D. EU TUBING
01
0.5
10
PRESSURE DROP-PSI/lOOO
50 -MULTIPLY
10
50
100
BY SPEC G R A V
Fig. 15-96. Relationship between the fluid flow rate and pressure drop when flow occurs between the 9;-in., 43i-lb casing and 4;-in. O.D. EU tubing.
(13) Which of the following are advantages of hydraulic pumping? (a) Wide production range (flexible). (b) No large equipment over wellhead. (c) Requires no attention. (d) Operating pressure allows pump intake pressure to be accurately estimated. (e) Directional wells do not present problems. (14) (a) For piston pumps, is the P/E ratio used to determine maximum pump setting depth or not? (b) They should be sized so that maximum operating speed is: (1) Used only by foreman. ( 2 ) Is no more than 85% of maximum rated speed. (3)
134
PRESSURE DROP IN TUBING FLOW BETWEEN lo%’’, 40.5# CSG. & 3%”O.D. EU TUBING
PRESSURE DROP-PSl/lOOO’-MULTIPLY
BY SPECIFIC GRAVITY
Fig. 15-97. Relationship between the fluid flow rate and pressure drop when flow occurs between the loi-in., 40.5-lb casing and 3i-in. O.D. EU tubing.
Within local speed limit. (4) Within OSHA specifications. (15) (a) A nozzle to throat ratio of 1OC in jet pumps means throat is a No. throat. (b) Power fluid cleanliness is less or more important in the case of jet pumps than with a piston pump? (c) Describe several throat materials available according to the type of service.
735 PRESSURE DROP IN TUBING FLOW BETWEEN lo%”,40.5# CSG. 6 4%” O.D. EU TUBING
01
0.5
1.o
PRESSURE DROP-PSI/lGOO’-MULTIPLY
I
5.0
10
50
1 3
BY SPECIFIC GRAVITY
Fig. 15-98. Relationship between the fluid flow rate and pressure drop when flow occurs between the lOi-in., 40.5-lb casing and 4i-in. O.D. EU tubing.
736
Fig. 15-99. Grid for Figs. 15-65 through 15-98.
REFERENCES Coberly, C.J., 1961. Theoiy and Application of Hydraulic Oil Well Pumps. Kobe Inc., Huntington Park, Calif., 127 pp. Borden, Jr., G. and Rzasa, M.J., 1950. Correlation of bottom hole sample data. Trans. A.I.M.E., 189: 345-348. Neely, B., Gipson, F., Clegg, J., Capps, B. and Wilson, P., 1981. Selection of artificial lift method. 56th Annu. Fall Tech. Conf. and Exhibition of the SOC. Pet. Eng. of A.I.M.E., San Antonio, Tex., Oct. 5-7, SPE 10337,6 pp. Standing, M.B., 1952. Volumetric and Phase Behavior of Oil Field Hydrocarbon Systems. Reinhold, New York, N.Y., 122 pp. Wilson, P.M., 1976. Introduction to Hydraulic Pumping. Kobe Inc., Huntington Park, Calif., 103 pp.
131
Chapter 16
ELECTRIC SUBMERGIBLE PUMPS
W.J. POWERS, CLARENCE DUNBAR, and GEORGE V. CHILINGARIAN
INTRODUCTION
The electric submergible pump (ESP) is perhaps the most versatile of the major oil production artificial lift methods. The purpose of this chapter is to provide the reader with a broad-based understanding of the key factors in selecting, installing, and operating electric submergible pumps. The ESP topics covered here are: electric submergible pump system, applications, electric submergible system components, selection data and methods, handling, installation and operation, and troubleshooting.
ELECTRIC SUBMERGIBLE PUMP SYSTEM
The Electric Submergible Pump System (ESP) is usually comprised of a downhole pump, electric power cable and surface controls. In a typical application, the downhole pump is suspended on a tubing string hung on the wellhead and is submerged in the well fluid (see Fig. 16-1). The pump is close-coupled to a submergible electric motor, which receives power through the power cable and surface controls. The ESP has the broadest producing range of any artificial lift method. The standard 60 Hz producing range of the ESP is from a low of 100 bbl of total fluid per day up to 90,000 bbl/day. Variable-speed drives can extend the producing range beyond these rates. Although most operators tend to associate ESPs with “highvolume” lift rates, the average ESP produces less than 1000 bbl of total fluid per day. The ESPs are used to produce a variety of fluids as well as the gas, chemicals, and contaminants commonly found in these fluids. They are currently economically operated in virtually every known environment found in the oilfields of the world. The oil/water ratio is, in general, not significant in assessing an application. Relatively high gas/fluid ratios can be handled using “tapered” design pumps and a special gas separator pump intake. Aggressive fluids (those containing H,S, CO,, or similar corrosives) can be produced utilizing special materials and coatings. Sand The help extended by Marshall Behrens and W. Duane Anderson was indeed invaluable.
738
and similar abrasive contaminants can be produced with acceptable pump life using specially modified pumps and operation procedures. The ESPs usually do not require storage enclosures, foundation pads, or guard fences. An ESP can be operated in a deviated or directionally drilled well, although the recommended operating position is in a straight section of the well. Because the ESP can be up to 200 ft long, operation in a bend or “dog leg” could seriously affect the unit’s run-life and performance by causing hot spots where the motor rests against the casing. The ESP can operate in a horizontal position. In this case, run-life will be determined by the protectors ability to isolate well fluid from the motor. These pumps are currently operated in wells with bottomhole temperatures up to 350’F. Operation at elevated ambient temperatures requires special components in the motor and power cable that are capable of sustained operation at a high ambient temperature. The ESPs can efficiently lift fluids in wells in excess of 12,000 ft deep, and can be operated in casing as small as 4: in. O.D. Many studies indicate that ESPs represent the most efficient lift method and the most economic on a cost per lifted barrel basis. Overall system efficiency can be as high as 68%, depending upon fluid volume, net lift. and pump type. The major disadvantage of the ESP is that it has a narrow producing rate range compared to other artificial lift forms. It does handle free gas well but the impact of large volumes of gas can be destructive to the pump. The run-life can be adversely affected by poor quality electric power, but this is not limited to the ESP. APPLICATIONS
The ESP has historically been applied in lifting water or in low oil cut wells that perform similar to water wells. Within t h s apparently narrow segment, however, there are numerous types of installations and equipment configurations. The following equipment applications and configurations are covered in this section: (1) Typical Installation ( 2 ) Booster/Injection (3) Bottom Intake/Discharge (4) Cavern Storage/Shrouded (5) Offshore Platforms Typical installation
A typical ESP installation is shown in Fig. 16.1. The ESP system’s major surface and downhole equipment is presented here. In this installation, the available surface power is transformed to the downhole power requirements by three single-phase transformers. The transformed power is supplied by a power cable to a switchboard and then through a junction box and wellhead-tubing support. The power cable is run in with the production tubing string and is banded to the tubing to prevent
139
U
0 0 "
0
--
Fig. 6-1. Typical installation of an electric submergible pump system (ESP)
740
mechanical damage during installation and removal. The power cable is spliced to a motor flat cable, whch is banded to the exterior of the pump-protector-motor unit. The centrifugal pump is located at the tr>pof the downhole unit, and is hung on the tubing string by the discharge head. Below the pump is a standard intake which provides for fluid entry to the pump. The center component is the protector. The protector both equalizes external and internal pressures and isolates the motor oil from the well fluid. The lowest component is the motor which drives the centrifugal pump. The downhole unit is landed above the perforations. This is necessary so that fluid entering the well flows past the motor. This flow cools the motor which would otherwise be likely to overheat and fail. ELECTRIC SUBMERGIBLE SYSTEM COMPONENTS
Motor
The ESP system’s prime mover is the submergible motor (see Fig. 16-2). The motor is a two-pole, three-phase, squirrel cage induction type. Motors run at a nominal speed of 3500 rpm in 60 Hz operation, and are filled with a highly refined mineral oil that provides dielectric strength, bearing lubrication, and thermal conductivity. A standard motor is capable of operating in wells at temperatures up to 250’F. Higher temperature capability motors for hotter wells are also manufactured. Heat generated by motor operation is transferred to the well fluid as it flows past the motor housing. A minimum fluid velocity of one foot per second is recommended to provide adequate cooling. Because the motor relies on the flow of well fluid for cooling, a standard ESP should never be set at or below the well perforations or producing zone unless the motor is shrouded. Reda motors, for example, are manufactured in four different diameters (series): 3.75 in., 4.56 in., 5.43 in., and 7.38 in.. Thus motors can be utilized in casing as small as 44 in. in diameter. Horsepower capabilities range from a low of 74 hp in 3.75-in. series, to a high of 1000 hp in the 7.38-in. series. Motor construction may be a single section or several sections in tandem bolted together to reach a specific horsepower. Motors are selected on the basis of the maximum outside diameter that can be easily run in a given casing size. The standard motor housing material is a heavy-wall, seamless, low-carbon steel tubing. The motor shaft material is carbon steel. The rotor(s) are supported by sleeve bearings made of Nitralloy and bronze. The squirrel cage rotor is made of one or more sections depending on motor horsepower and length. The motor stator is wound as a single unit in a fixed housing length. The standard motor thrust bearing is a fixed pad Kingsbury type. Its purpose is to support the thrust load of the motor rotors, and to vertically align the rotor-shaft assembly within the stators’ magnetic field while the motor is running. Axial alignment to maintain the air gap between the stator and the rotors is maintained by special rotor bearings.
741
Center Tandem Motor
Motor Head Showing Power Cable Connection
Fig. 16-2. Motor of ESP system. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
742
d Pump with Standard Intake
Fig. 16-3.Example of a multistage centrifugal type electric submergible pump. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
143
Pump
The electric submergible pump is a multistage centrifugal type (see Fig. 16-3). The type of stage used determines the design volume rate of fluid production. The number of stages determines the total design head generated and the motor horsepower required. The materials used in manufacturing impellers include Ni-Resist, Ryton, and bronze. Diffusers are universally manufactured of Ni-Resist. The standard shaft material is K-monel. Optional, high-strength shaft materials are used in deep setting applications where conventional shaft material horsepower limits are exceeded. “Bolt-on” design makes it possible to vary the capacity and total head of a pump by using more than one pump section. Large-capacity pumps, however, typically have integral heads and bases. The housing of the pump contains the impellers and diffusers and supports the weight of the complete unit. Housings are seamless, heavy-walled, low-carbon steel tubing. The housing is capable of containing the normal operating pressure of the pump with an adequate safety factor. The nominal outside pump diameter ranges from 3.38 in. to 11.25 in. Protector
The protector’s primary purpose is to isolate the motor oil from the well fluid while balancing bottomhole pressure and the motor’s internal pressure. There are two types of protector design: the positive seal (Fig. 16-4) and the labyrinth path (Fig. 16-5). The “positive seal” design relies on an elastomer fluid barrier bag to allow for the thermal expansion of motor fluid in operation, and yet still isolate the well fluid from the motor oil. The “labyrinth path” design utilizes differential specific gravity of the well fluid and motor oil to prevent the well fluid from entering the motor. T h s is accomplished by allowing the well fluid and motor oil to communicate through tube paths connecting segregated chambers. The protector performs four basic functions: (1)connects the pump to the motor by connecting both the housing and drive shafts, (2) houses a thrust bearing to absorb pump shaft axial thrust, (3) isolates motor oil from well fluid while allowing wellbore-motor pressure equalization, and (4) allows thermal expansion of motor oil due to operating heat rise and thermal contraction of the motor oil after shutdown. The standard protector housing material is the same carbon steel material as is used in motors and pumps. The drive shaft is also K-monel. Pump intake
Two types of intakes are used to allow fluid to enter the pump. These are the standard intake shown in Fig. 16-3 and the gas separator intake shown in Figs. 16-6 and 16-7. A gas separator intake is used when the vapor/liquid ratio is greater than that which can be handled by the pump. If the gas remains in solution, the pump
Positive Seal Protector
Fig. 16-4. Positive-seal protector. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
745
Labyrinth Path Protector
Fig. 16-5. Labyrinth-path protector. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
746
Static Type Gas Separator
Fig. 16-6. Static type gas separator intake. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
141
Fig. 16-7. Rotary gas separator intake. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
748
will perform normally. Once the gas/liquid ratio exceeds a value of about 0.1, however, the pump may produce a head lower than normal. As the vapor/liquid ratio increases above 0.1 and free gas increases, the pump will eventually “gas lock”, which usually drastically reduces fluid production and in extreme cases can damage the pump. The pump, however, can tolerate progressively higher vapor/liquid ratio at higher operating pressure. There are two types of gas separator intakes: (1) the static type (Fig. 16-6) and (2) the KGS rotary type (Fig. 16-7). The static type induces gas separation by reversing the fluid flow direction. At the fluid entry ports, the reversal of fluid flow direction creates lower pressure which allows the gas to separate. The separated gas moves up the annulus and vents at the wellhead. The fluid, which still contains some gas, enters the separator and moves downward into the standtube. The fluid is picked up by the rotating pick-up impeller. The impeller creates a vortex which forces dense, relatively gas-free fluid to the outside and allows additional gas breakout along the shaft. This provides the first stage of the pump with a higher density of fluid than if the gas broke out in the pump. The Reda KGS rotary gas separator (Fig. 16-7) utilizes a rotary inducer-centrifuge to centrifugally separate the gas and produced liquids. The well fluid enters the intake ports and moves into the inducer. The inducer increases the fluid pressure discharging into the centrifuge. The centrifuge forces the denser fluid to the outside. Gas rises from the center of the centrifuge through the flow divider into the crossover section where gas vents to the annulus and fluid is directed into the first stage of the pump.
Power cable Electric power is supplied to the downhole motor by a special submergible cable. There are two different cable configurations: (1)flat (or parallel) and ( 2 ) round (see Fig. 16-8). Round construction is used except where casing-tubing clearance requires flat constructions’ lower profile. The standard range of conductor sizes is 1/0-6 AWG (American Wire Gauge). This range meets virtually all motor amperage requirements. Almost all cable types use copper conductors. Although aluminum conductors are sometimes used in high H,S concentration wells, mechanical protection is provided by armor made from galvanized steel or, in extremely corrosive environments, from monel. Cable is constructed with three individual conductors-one for each power phase. Each conductor is enclosed by insulation and sheathing material. The thckness and composition of the insulation and sheathing determines the conductors resistance to current leakage, maximum operating temperature capability, and resistance to withstand permeation by well fluid and gas. For example, Reda electric power cable is rated to operate up to 400°F at 1500 psi (see Table 16-1). Round cable is also manufactured with an “I” wire. The “I” wire serves as an electrical link between a downhole instrument and surface reading-processing equipment.
749
Round a n d
Parallel
P o w e r Cable
Fig. 16-8.Round and parallel power cable. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
Motor frat cable The motor flat cable is in the lowest section of the power cable string. The motor flat cable has a lower profile than standard flat power cable so that it can run the
TABLE 16-1 Reda cable types and applications Cable type
Configuration flat or round
Maximum bottomhole temperature ( O F at 1500 psi)
KVA rotary (000)
Armor
Redalene
F, R
205"
3-5
Redahot
F,R
275'
3
Redalead
F
350'
3-5
Redared
R
400"
3-5
Polyethylene
R
180"
3
Galvanized steel and monel Galvanized steel and monel Galvanized steel and monel Galvanized steel and monel None
750
Fig. 16-9. Switchboard with recording ammeter and RedalertTM motor controller. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
length of the pump and protector in limited clearance situations (see Fig. 16-1). The motor flat cable is manufactured with a special terminal called a “pothead”. The function of the pothead is to allow entry of electric power into the motor while sealing the connection from well fluid entry. Switchboard The switchboard is basically a motor control device (see Fig. 16-9). Voltage capability ranges from 600 V to 4900 V on a standard switchboard. All enclosures are NEMA-3R (National Electrical Manufacturers Association), which is suitable for virtually all outdoor applications. The switchboards range in complexity from a
751
simple motor starter-disconnect switch to an extremely sophisticated monitoring-control device. There are two major construction types: electromechanical and solid state. Electromechanical construction switchboards provide overcurrent-overload protection through three magnetic inverse time delay contact relays with pushbutton, manual reset. Undercurrent protection is provided by Westinghouse SCR (Silicone Controlled Rectifier) relays. These features provide protection against downhole equipment damage caused by conditions such as pump-off, gas lock, tubing leaks, and shutoff operation. The solid state switchboards incorporate the highly sophisticated motor controller. The purpose of the motor controller is to protect the downhole unit by sensing abnormal power service and shutting down the power supply if current exceeds or drops below preset limits. This is accomplished by monitoring each phase of the input power cable to the downhole motor. The monitoring function applies to both overload and underload conditions. When a fault condition occurs, the controller shuts down the unit. It can be programmed to automatically restart the downhole motor following a user selected time delay if the fault condition is caused by an underload. The programmed time delay can be from one minute up to twenty hours. Overload condition shutdown must be manually restarted, but t h s should be done only after the fault condition has been identified and corrected. A valuable switchboard option is the recording ammeter. Its function is to record, on a circular strip chart, the input amperage to the downhole motor. The
*
HAND HOLE ACCESSTOTAPCHANGER
12 500 UNITS 95 BIL HV BUSHINGS 2 h 5 0 UNITS I50 BIL HV B U S H I N G S 3
12GAUGETANK
[
EXTRA CREEP BUSHINGS
PAINT SYSTEM MEETS OR EXCEEDS ASTM 5% SALT FOR 1519 HRS. TAP CHANGER CORE a COIL HOLD DOWN SUPPORT
2"O.D. (14 GA) COOLING TUBES BUSHING LEADS COIL BLOCK
COIL, ALL COPPER WINDING 65°C INSULATING SYSTEM.
CORE a COIL POSITION GUIDES
CORE, LOW LOSS MATERIAL GRAIN ORIENTED STEEL HEAVY DUTY FRAME
Fig. 16-10. Single-phase transformer. (Courtesy of Reda Pump Division, TRW Energy Products Group )
152
Hercules HHS Wellhead
Fig. 16-11. Hercules HHS wellhead, (Courtesy of Reda Pump Division, TRW Energy Products Group.)
ammeter chart record can show if the downhole unit is performing as designed or if abnormal operating conditions exist. Abnormal conditions can result when a well’s inflow performance is not correctly matched with pump capability or when electric power is of poor quality. Abnormal conditions that are indicated on the ammeter chart record are primary line voltage fluctuations, low current (A), high current, and erratic current. Transformer The ESP system utilizes three different transformer configurations: (1) three single-phase transformers (Fig. 16-10), (2) one three-phase standard transformer, or (3) one three-phase autotransformer. Transformers are generally required because primary line voltage generally does not meet the downhole motor voltage requirement. Oil-immersed, self-cooled (OISC) transformers are used in land-based applications. Dry-type transformers are sometimes used in offshore applications, which exclude oil-filled transformers. Wellhead The wellhead or tubing support may be used as a limited pressure seal (see Fig. 16-11). The wellhead provides a pressure-light pack-off around the tubing and power cable. High-pressure wellheads-up to 3000 psi-use an electrical power feed-through mandrel to prevent gas migration through the cable. Wellheads are manufactured to fit standard casing sizes from 4 i to 134 in.
153
-5OMlNlMUM-
SWITCHBOA
15 FEET MINIMUM JUNCTION B O X
Fig. 16-12. Junction box. (Courtesy of Reda Pump Division, TRW Energy Products Group )
Junction box A junction box connects the power cable from the switchboard to the well power cable (see Fig. 16-12). It is necessary to vent, to the atmosphere, any gas that may migrate up the power cable from the well. This prevents accumulation of gas in the switchboard, which can result in an explosive and unsafe operating condition. A junction box is required on all ESP installations.
Accessory options
The following covers two major accessory options: (a) the pressure-sensing instrument and (b) a variable-speed drive. Pressure-sensing instrument ( P S I ) The pressure-sensing instrument (PSI) provides the operator with precise down-
154
Fig. 16-13. Pressure and temperature sensing instrument. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
hole pressure and temperature data. The PSI has two components: (a) a downhole transducer-sending unit and (b) a surface readout unit (see Fig. 16-13). The downhole transducer-sending unit connects electrically and bolts to the base of the motor. Both pressure and temperature data are transmitted from the transducersending unit to the surface read-out through the motor windings and the power cable on a DC carrier signal. The transducer receives operating power from the motor’s neutral winding. This allows the operation of the PSI even when the motor is not running. The major use of the PSI unit is determination of the producing potential of a well. This is accomplished by determining both static and dynamic reservoir pressures. By correlating the change in pressure with a given producing rate, a well’s inflow performance can be accurately quantified. This, in turn, will allow equipment selection which optimizes well production. Variable speed drive The variable speed drive (VSD) is a highly sophisticated switchboard-motor controller (see Fig. 16-14). A VSD performs three distinct functions: (a) varies the capacity of the ESP by varying the motor speed, (b) protects downhole components
755
Fig. 16-14. Variable-speed drive switchboard motor controller. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
from power transients, and (c) provides “soft start” capability. Each of these functions is discussed in succeeding paragraphs. A VSD changes the capacity of the ESP by varying the motor speed. By changing the voltage frequency supplied to the motor and thus motor rpm, the capacity of the pump is also changed in a linear relationship. Thus, well production can be optimized by balancing inflow performance with pump performance. This applies to
756
both long-range reservoir changes as well as short-term transients, such as those associated with high GOR wells. This may eliminate the need to change the capacity of a pump to match changing well conditions or it may mean longer run-life by preventing cycling problems. This capability is also useful in determining the productivity of new wells by documenting pressure and production values over a range of drawdown rates. The change in voltage frequency can be made manually or automatically. The VSD can operate automatically in a “closed loop” mode with a programmable controller and PSI instrument. The VSD also protects the downhole motor from poor quality electric power. The VSD is relatively insensitive to incoming power balance and regulation, while providing closely regulated and balanced output. The VSD will not output electric power transients to the downhole motor, but it can be shutdown or damaged by such transients. Given the choice, most operators would prefer to repair surface installation equipment rather than pull and run downhole equipment. Within limits, the VSD upgrades poor quality electric power. The VSD takes a given frequency and voltage AC input, converts the AC to DC, and then rebuilds the D C to an AC waveform. The shape of the waveform is a six-step square wave. The “soft start” capability of a VSD provides two major benefits. The first is that it reduces the start-up drain on the power system. Secondly, the mechanical strain on the pump shafts is significantly reduced when compared to a standard start. This capability is valuable in gassy or sandy wells. In some cases, slowly ramping a pump up to operating speed may avoid pump damage. The VSDs are manufactured in sizes ranging from 125 to 1000 KVA. These units are designed for sustained operation at maximum ratings when properly applied to a given set of pump and well conditions.
SELECTION DATA AND METHODS
The data requirements and calculation procedures required for pump selection in a typical ESP application are presented in this section. The single most important factor in selection of an ESP is the input data. The data used in sizing an ESP must be accurate and reliable to insure that the unit is properly matched to the well’s inflow performance. The data required for selection of an ESP are: (1) mechanical data, (2) production data, (3) fluid data, and (4) power supply. Mechanical data The ESP well data form (Fig. 16-15) exemplifies a specific pump selection (Jones Oil Company; Smith No. 1 well). The mechanical data includes: (1) Casing size and weight: 5) in. O.D.; 17 lb/ft. (2) Tubing size, weight and thread: 2$ in. O.D.; EUE. (3) Well depth-both measured and true vertical: 7070 ft.
151
Request For Information Gt73
REDA P U M P COMPANY Bartlesville, Oklahoma GENTLEMEN:
Will you please send me complete information concerning a Reda unit for the following oil well: DESCRIPTION: Company Well Name &! No. ............ Field or Pool Name ..... ......................... County ............................................ State . ...................... Production Formation .... Sand or Lime ............................. Total Well Depth ..21.0.0 ...... Fr. Caring: . . . . In. O.D. . -. Lbs. per Ft. ................... Liner: .................. In. O.D. . . . . . . . Lbs per EL. T o p of Liner at .... ... ft. Bottom of Liner at ...................... ft. .. to ... ft. Open Hole: . . . . . . . . . . . . . . to fr. Perforations:
J . c l ~ . ~ ~ % o i \ . . b ~ ) / Sm'ltk..b.1 sand
1.7..
.5.&
..?025.
7add.
PRESENT P R O D U C T I O N A N D P R O D U C T I V I T Y DATA: Bbl. Water/Day. Average Production Rate: ............... .-............. Bbl. Oil/Day wtrh dQ.0 ............. Bbl./Day total fluid using .................. method of arriiical I l f r . . Cu. Ft./Day gar produced with with ...... ........... feet of . . . . . . . In. O . D . E U E I M . Tubing. ft P.S.I. at . . Static Fluid Level . . . . ft. from top. Static Bottom Hole Pressure ................ Producing Fluid Level ..... f t . from top while producing A..a.o.. ....... Bbl./Day total fluid. ....... Bbl./Day per P.S.I. of draw-down or t Bbl./Day per foot of draw-down Productivity index is
i!mo
~..L./s.
.385f
238/
7'763.
Fluid gradient 1 s P.S.I./Ft. From all observations i t is estimated this well is capable of producing . feet from the cop. producing fluid level of
500
.z?s/
GENERAL INFORMATION: Well has had a history of problems due to the following checked condmonr: 0 Paraffin, 0 Other such as . . . . . . . . . . P S.I. back pressure
IS
maintained on the caring annulus, which
1:
0 Sand
.....
Bbl./Day
total fluid from a
Production, L! Severe Corrosmn,
nol (cross out o n e ) vented
14
1.
Gravity of 011 . . "API Specific gravity of brine .../!I . Bottom Hole Temperature 0 'F Viscosity of Oil: ...... SSU a t ............... "F a n d ........... SSU at the bottom hole temp. of . 'F volts and primary Y O I C ~ R P Electric Power Supply: 3 .... phase, .6.O..... cycle with secondary voltage of of 0 0 Bbl./Day ratal fluid (oil plus Calculations indicate a Reda unit designed to pump a desired capacity of feet would ftr this well's particular characteristics. T h e calculated warer) against a total dynamic head of . . feet of frmron loss in feet plus ( 2 ) total dynamic head is based o n ( I ) a producing fluid level of PSI tubmg discharge pressure feet of .. In. O.D. E U E / o . b tubing plus ( 3 ) Other Information or Remarks: ................... ............. . . . .
fZ500
5
b.gO6.
.*.??../
a..%
7aoo
. .
.
.
.
35
.a00
.
.
.
.
.
.
.
.
Yours very truly.
Name Company Address Date
.
.
...
......
City
..........
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
(If above rnformation is submitted as completely as possible, it will greatly facilitate a prompt reply 1 R-144 - Revised 1/20/65.
Fig. 16-15. Request for information form. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
758
(4) Perforations depth-both measured and true vertical: 7025-7050 ft. (5) Unusual conditions such as tight spots, dog legs and deviation from true vertical at desired setting depth. The casing size and weight determine the maximum diameters of the motor, pump, and protector components that will physically fit in the well. In general, the most efficient installation is obtained when the pump having the largest possible diameter, in the target flow range, is selected. The depth of the well and the performance determine the maximum setting depth of the ESP. If the motor is to be set below the perforations, a motor shroud must be used to provide a flow of well fluid past the motor for cooling. In this example the pump setting depth is 7000 ft. Production data
The production data needed for correct pump sizing includes (Fig. 16-15): (1) Present and desired production rate: 200 bbl/day and 500 bbl/day. (2) Oil production rate: 0 bbl/day. (3) Water production rate: 200 bbl/day. (4) Gas/oil ratio (free gas and solution gas or gas bubble point): 0 cf/bbl. (5) Static bottomhole pressure and fluid level: 1500 psig at 7000 ft; fluid level = 3851 ft. (6) Producing bottomhole pressure and stabilized fluid level: 1100 psig at 7000 ft; fluid level = 4691 ft.
w a
I
2000 I000 -
3000
lift requirements
I I I I I I
I I I I
I I
Fig. 16-16. Capacity curve for the example well
-
-
159
(7) Bottomhole temperature: 160°F. (8) System back pressure from flowlines, separator, and wellhead choke: 200 psig at the wellhead. The inflow performance of a well establishes the maximum economic and efficient rate at which it can be produced. Liquid level data may be used as a substitute for producing pressures and rates in water or low oil cut wells with no gas. In this case, a straight-line productivity index ( J ) curve may be used as a reasonable approximation of well capacity (see Fig. 16-16). Figure 16-16 graphcally illustrates the J of the above example well. The static fluid level is at 3851 ft. At a production rate of 200 bbl/day, the producing level is at 4691 ft. To obtain the pumping fluid level at the 500 bbl/day production, the curve is extended to 500 bbl/day point at which the pumping fluid level is approximately 5951 ft. Most oil wells do not exhibit a straight-line J curve because of interference caused by gas. The Vogel (1968) technique yields a downward sloping curve that
INFLOW PERFORMANCE RELATIONSHIPS FOR SOLUTION - G A S DRIVE WELLS BY: J. V. VOGEL From SPE Paper 1476 January 1968 JPT Qo [Z) MAX = 1 - 0.20
PR
- 0.80
(2;
P, = Static Reservoir Pressure at Producina Zone PwF= Well Bore Pressure at Producing Zone Oo = Producing Rate (BPDI at PwF Q0 Max = Producing Rate at 100%Drawdown PwF= 0
Fig. 16-17. Inflow performance relationships for solution-gas-drive wells. (After Vogel, 1968; courtesy of the Society of Petroleum Engineers of AIME.)
760
corrects for gas interference. The IPR curve (Fig. 16-17) applies when wellbore pressure in the producing zone drops below the bubble point resulting in two-phase flow as the gas breaks out of the fluid. It is extremely important to remember that the data obtained for t h s approach in sizing an ESP must be both accurate and reliable to insure proper equipment selection. Fluid data The fluid data includes: (1) Oil API gravity, viscosity, pour point, paraffin content, sand content, and emulsion tendency; currently, 100% water. (2) Water specific gravity, chemical composition, corrosion potential, and scaleforming tendency; specific gravity = 1.1. (3) Gas specific gravity, composition, and corrosive potential; no gas present. (4) Reservoir formation volume factor, bubble point pressure, and viscosity-temperature curve; no gas present. The specific gravity of the produced fluid is directly related to the horsepower required to turn a given size pump. Although relatively few applications encounter fluid viscosities high enough to affect pump performance, it is important to be aware that capacity, head and horsepower correction factors may be required. In wells with a water cut of 65% or higher, the fluid generally will not require viscosity correction factors (except for emulsions). The PVT data are required when gas is present. Various computer programs ( e g , COMPSEL, a Reda computer program) for pump selection (discussed later in this section) contain a subroutine which utilizes Standing’s (1952, 1975) correlations in approximating the PVT values when actual data are not available. In high GOR applications, PVT data are very desirable because the three standard correlation estimates of Standing (1952, 1975), Lasater (1958) and Vasquez and Beggs (1980) yield large differences in calculated downhole gas volume. Electric power supply The electric power supply includes: (1)Voltage available and frequency: 12,500 V; 60 Hz. (2) Capacity of the service: adequate. (3) Quality of service (spikes, sags, etc.): good quality. The power data is important because it partially determines transformer and switchboard requirements. Frequency affects pump rotation speed, capacity, and head. Selection process Once the required data have been gathered and analyzed, the first ESP selection step is to determine the well’s production capacity at a given pump setting depth.
761
This involves analysis of the inflow performance data as well as desired production rate. Two key factors that must be considered are the minimum pump intake pressure (net positive suction head), which the well will permit without pump-off or gas lock, and the producing rate, which draws the fluid level down to an optimum level. The next selection step is to determine the total dynamic head ( T D H ) . The TDH is the sum of three variables: (a) the true vertical lift distance from the producing fluid level to the surface, (b) friction loss in the tubing string, and (c) discharge pressure head at the wellhead. The design T D H determines the number of stages required in a pump. In the above-described example, the J is calculated by taking the present production rate of 200 bbl/day divided by 4691 ft pumping fluid level minus the static fluid level of 3851 ft. Thus, J is equal to 0.2381 bbl/day/ft of drawdown. Next, the pumping fluid level is calculated at the 500 bbl/day rate. Dividing 500 bbl/day by 0.2381, gives 2099.96 ft drawdown. On adding t h s 2099.96 f t of drawdown to the 3851 ft of static fluid level, a pumping fluid level of 5950.96 f t is obtained. The total head that must be developed by the pump will consist of the lift, the tubing friction loss, and the wellhead pressure. The 200 psi wellhead pressure is converted to feet of head by first calculating a pressure gradient for the 1.1 specific gravity brine. The latter is multiplied by 0.433 psi/ft gradient for fresh water to obtain 0.4763 psi/ft fluid gradient for the 1.1 specific gravity brine. On dividing the 200 psi wellhead pressure by the salt water gradient of 0.4763 psi/ft, the head of 419.9 ft is obtained. The tubing friction loss can be calculated from Fig. 16-18. The figures for the old tubing are used in order to be on the conservative side. In this particular example, the tubing friction loss at 500 bbl/day is 5 ft per thousand feet and with 7000 ft of tubing a total tubing loss is 35 ft: (7000/1000) X 5 = 35. To obtain the total dynamic head requirements, the 5950.96 f t is added to the 35 ft of tubing friction loss and 419.9 ft wellhead pressure (which in this example represents 200 psi pressure at the surface). Thus, total head requirement is equal to 6405.86 ft. Selection of a specific pump involves identifying a pump of the largest possible diameter whch can be run in the well. The pump should have the target capacity within its recommended operating range and close to its peak efficiency. The initial pump capacity selection can be made from Table 16-11, for example, which lists all Reda pumps grouped by series for various casing sizes. The first group is the “A” series pumps which will go inside 4i-in. O.D. casing. The next group is the “400” series pumps that will go into 5;-in. O.D. casing. In the recommended range column, it is necessary to identify the pump that is in the 500 bbl/day range in 5:-in. casing. It is the D-550 pump that will fit this particular set of conditions. The individual pump curve should then be reviewed to determine the optimum producing range and the proximity of the design producing rate to the pump’s peak efficiency (see Fig. 16-19 for a typical pump performance curve). It is very important to choose a producing rate which is in the recommended capacity range
LOSS OF HEAD DUE TO FRICTION OF WATER I N P I P E AS LISTED
FRICTION LOSS PER I uuu 1
BASED ON WILLIAM & HAZEN TABLES NEW - SCHEDULE 40 NEW P I P E
2
1
3
5
3 .
4
BPD X 100 10
5 678910 15 20
30
2
3 4 5 678910 BPD X 1000 50 100 150 200 300 GALLONS PER MINUTE
2 500
3 4 5 67a910 BPD X 10,000 1000 1500 2000 3000
Fig. 16-18. Determination of friction loss in tubing. (Courtesy of Reda Pump Division, TRW Energy Products croup.)
763 TABLE 16-11 Engineering tables for the initial pump capacity selection (pumps 60 Hz/3500 rpm) (Courtesy of Reda Pump Division, TRW Energy Products Group) Capacity range recommended limits
Series
Outside diameter (inches)
Pump type
Max BHP a rating for pump shaft
(BPD)
(m3/D)
338
3.38
A400 AN550 AN900 A1200 A1500
94 94 94 94 125
280-500 425-700 660-1 100 875-1575 1100- 1900
45 - 80 68-111 105-175 139-250 175-302
400
4.00
DN280 D400 DN450 D550 D700 D950 DNlOOO DN1300 D1350 DN1750 D2000 DN3000 DN4000
44 94 94 94 94 125 125 125 125 125 125 256 256
100-450 280-550 320-575 375-650 500-900 600-1150 760-1250 975-1650 950-1800 1200-2050 1400-2450 2 100- 3700 3400-5000
16-72 45-81 51-91 60-103 80- 143 95-183 121-199 155-262 151-286 191-326 223-390 334-588 540-795
450
4.62
EN1250 El450 EN3600
160 160 256
950-1600 1050-1800 2800-4500
151-254 167-286 445-715
540
5.13
G2000 GN2500 G2700 G3100 DN4000 GN5200 G5600 GN7000 GNlOOOO
256 256 256 256 375 375 375 375 637
1500-2500 2000-3 100 2100-3400 2200-3700
238-397 318-493 334-541 350-588
4200-6600 4500-7250 5500-8500 8000-12000
668- 1049 715-1153 874-1351 1272-1908
562
5.62
HN13000
375
9200-16400
1463-2607
650
6.62
IN7500 IN10000
637 637
6000-9500 8000-12250
954-1510 1272-1948
675
6.75
JN16000 JN21000
637 637
12800- 19500 16000-25000
2035 3100 2544,3795
862
8.62
M520 M675
631 637
12000-24000 19000-32500
1908-3816 3021-5167
950
9.50
N1050
1000
24000-47500
3816-7552
1000
10.00
N1500
1000
35000-59000
5564-9380
1125
11.25
P2o00
lo00
53600-95800
8521- 15240
-
For additional shaft horsepower capability contact a TRW Reda company. Note: Larger capacity pumps of larger horsepower are available upon request.
a
164
m
0
8
s
N 0
m
-a
s
P
m
i-"
.-a
Fig. 16-19. Typical pump performance curves.
765
of the specific pump. When a pump is operated outside this range, premature failure can result. The recommended operating range for 60 Hz is defined by vertical lines. The D-550 pump has an operating range of 380-650 bbl/day, and the 500 bbl/day rate falls near the peak efficiency range of the pump. By intersecting the head capacity curve at the 500 bbl/day rate, one can determine that the D-550 pump develops approximately 22.40 ft of head per stage (2240/100 stages = 22.40 ft per stage). The horsepower motor load curve is used similarly. Each stage will require approximately 0.175 horsepower per stage. Once a pump is chosen, the number of stages required can be calculated using the lift (ft) per stage data from the performance curve: Number of stages =
TDH (ft) Lift/stage( ft)
(16-1)
The horsepower required by the pump design can then be calculated. To accomplish this, the horsepower required per stage is read from the specific pump performance curve. The required motor horsepower is determined by multiplying the horsepower required per stage by the number of design stages. The performance curve horsepower data applies only to liquids with a specific gravity of 1.0. For liquids having other specific gravities, the water horsepower must also be multiplied by the specific gravity of the fluid pumped. Thus, the following equation can be used for the horsepower calculation: HP = (HP/stage) x (Number of design stages) X (sp. gr.)
(16-2)
In the present example, the total head required is 6405.86 ft and the rate is 500 bbl/day of 1.1 specific gravity water. To determine the number of stages required, the total head of 6405.86 ft must be divided by 22.40 ft/stage. T h s equals 285.98, which rounds off to 286 stages. To determine the horsepower required to drive the 286-stage pump, the horsepower per stage of 0.175 is multiplied by the 286 stages and then by the 1.1 sp. gr. to get a horsepower requirement of 55.06, whch may be rounded off to 55 horsepower. Once the design motor horsepower is determined, specific motor selection is based on setting depth, casing size and motor voltage. Although the cost of the motor is generally unrelated to voltage, overall ESP system cost may be lowered by utilizing higher voltage motors in deep applications. This lower cost can sometimes occur because a higher voltage can lower the cable conductor size required. A smaller conductor size (lower-cost cable), can more than offset the increased cost of a higher-voltage switchboard. Setting depth is a major variable in motor selection because of starting and voltage drop losses, which are a function of the motor amperage and cable conductor size. Table 16-111 lists the most common Reda motors and voltages. The 375-series motors go in 44-in. O.D. casing. The 456-series motors will go in 54-in. O.D. casing,
766
TABLE 16-111 List of the most common Reda motors and voltages
Moms 60 Hz 375 Series 13 75" 001
HP
Volrs
75 10 5
415
13 5
400
20
15
330 415 415 650
34 27 35 22 5 38 5 22 5 29 5 74 5
13 5 2; i
440
22 E
750 650 7Rn
AMP
Tandem Mo'ors
30 39
630
575
HP 10 15 20 25 30 35 40
50 35 5 51
60 45 51
58 5 67 5 76 5
90 102 1125 1275
660 740 1000 1250 860 990
1110 1320 1480 1650 1850
51 5 51 37 31
51 51 5 51 51 5 51 51 5 51
10
80
90
100 738 S e w 110
220 240 260 340
1325 2275 2280 2235
110 70 76 101
'"
14@ 160
Tandem Motors
400 440 480 520 600 680 720
2300 2050 2230 2420 3450 3170 3345
115 142 143 143 115 142 143
180
200
220 240
456 Serles 14 56" 001 Volts 435 655
450 750 410 690 425 1260 750 385 675 785 430 110 880 1340 675 815 955 1390 640 745 810 970 1330 540 750 945 1135 635 860 1085 1310 710 960 1135 1220 1460 1960
920 1075 1355 2205 1190 945 1125 1295 2245 Tandem Motors 1 80
1:oo 1890 2270 1270 1720 2',70 2620 1420 1920 2270 2440 2920 1580 1840 2150 2710 2380 ,890 2250 2590
540 Series 1543" 001 AMP 15 23 16 28 5 17 39 22 44 5 25 5 150 57
HP
Vnlt?
AMP
33 28 59 33 29 19 47 39 33 23 59 52 41 39 29 82 5 60 47
39 80 60 46
39
81 59 50 46 39 29 80 70 59 46 28 5 60 81 70 59 35 a2 5 60 47 39 80 60 46 39 81 59 50 46 39
ao
70 59 46 60 a1 70
59
Tandem Motors
240 260 300 320
360 400 450 480 540 600
1710 2060 2590 1850 2250 2150 1650 2230 1890 2250 2200 2270 2415 3345 2835 3300
88 73 59 88 67 87 122 88 5 120 89 115 127 122 88 5 120 115
161
540-series motors will go in 64-in. O.D. casing, and the 738-series motors will go into 84-h. O.D. or larger casing. In the 456-series motors, a 60-HP motor is the closest size to the requirement of 55 HP. There are several voltages listed for this motor horsepower. The reason is that with a very shallow setting, one could use a lower-voltage motor resulting in a less expensive switchboard. In this particular case, with a 7000-ft setting, it is desirable to maintain a reasonably high motor voltage to keep the amperage down to a low level whch will allow the use of a smaller size cable. Thus, one can select the 60-HP, 970-V, 39-A motor. Cable selection variables are amperage, voltage drop, annulus clearance, ambient well temperature, and corrosion conditions. The recommended maximum voltage drop should be limited to 30 V per thousand feet. If voltage drop exceeds this value, a larger conductor size should be used. Round cable is normally used unless tubing collar-annulus clearance dictates flat or parallel construction. The maximum operating temperature of a cable in relation to the specific well's ambient temperature determines the specific type of cable. Armor and lead sheathing may be recommended in a well with mechanical or clearance problems or in the presence of corrosive gas such as H,S. For the present example, Fig. 16-20, which shows cable voltage drop losses, can be used to determine the size of a cable that is appropriate for a 35.75-A motor [ ( 5 5 HP/60 HP) X 39 A]. In this particular installation, size No. 4 cable can be selected. Referring to Fig. 16-20 at 35.75 A, in the case of the size No. 4 cable there is less than 3% loss per thousand feet. With 7000 f t of cable, voltage drop loss is 147 V from the surface to the motor: [(7000/1000) x 211. In order to determine the required surface voltage, the 970 V requirement of the motor is added to the 147 V loss in the cable. Thus, a total of 1117 V is required. The surface electrical equipment (switchboard and transformer(s)) selection is based on the required motor horsepower, voltage, amperage, voltage loss, and cable size. The surface voltage is the sum of the downhole motor no-load voltage plus the voltage losses due to cable size and other component losses. Voltage losses associated with transformers range from 2.5 to 6% depending on the manufacturer. Additional impedance is associated with VSD transformer sizing. Transformers must also be selected based on the primary voltage available and the required hookup method: Delta Delta, Y Delta, or Delta Y. To obtain the KVA requirement for the transformer bank, the total voltage required is multiplied by the current (A) and by the square root of 3 for three-phase power and' then divided by 1000: (1117 V X 35.75 A X 1.73): 1000 = 69.084 KVA. The transformer banks are normally made in sizes of 25 KVA, 37f KVA, 50 KVA, 75 KVA, etc. A bank of three 25-KVA transformers gives a total KVA of 75, which will be more than adequate for the 69.084 KVA needed in this installation. The next step is the switchboard selection for this particular installation. There are several types of switchboards available starting with the DFH-2 switchboard for a 440-V installation. The MFH switchboard is for voltages up to 1000 V and the
I-
U
a r
768
06
mu
mc
0 1
m
-I0 W e 4
U u
Fig. 16-20. Voltage drop chart. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
769
MDFH switchboard is for voltages up to 1500 V. Larger switchboards are for 2400and 5000-V installations. In this example, a size 3 MDFH switchboard is selected. Size 3 would give a maximum current of 100 A and a voltage capability of up to 1500 V. The protector selection variables are motor and pump series, motor horsepower, and well temperature. Normally the protector is of the same series as the pump and motor. Large horsepower motors (150 HP and larger) may require a larger oil capacity. For large horsepower motors, one can use, for example, a Positive Seal Double Bag model or a tandem “Labyrinth Path” model (Reda motors). An ambient well temperature of 250°F or higher generally requires the use of the “Labyrinth Path” protector. Figure 16-21 lists the flat cable extensions for various motor sizes. In the present example, inasmuch as the bottomhole temperature is 160”F, one can select the standard flat cable material, which is described as 3KV KEOTB. If the bottomhole temperature was, for example, 180”F, one could use a lead-covered flat cable extension, which is described as 3KV KELB. The length of the flat cable should be at least 5 f t greater than the length of the protector and the pump, so that the flat to power cable splice is well above the pump discharge for clearance. Thus, the major components of the installation have been selected: motor, pump, protector, flat cable extension, the round cable going to the surface and the switchboard. The remaining items to complete the installation are the check valve, the bleeder valve, the tubing head, and the junction box at the surface. The purpose of the check valve is to prevent fluid from running back to the pump when the unit is shut down and causing the unit to spin in reverse rotation. It also eliminates the necessity of filling the tubing each time the pump starts up. The check valve is normally run 300-500 ft above the pump in a well where gas production is anticipated. By setting the check valve 300-500 ft above the pump, in the event of gas lock and pump shut-down, the gas has an opportunity to work out of the pump and be replaced by liquid fluid for the subsequent startup. The bleeder valve is normally run one joint above the check valve. It enables drainage of the fluid from the tubing so that the tubing will not have to be pulled “wet”. The wellhead is designed to support the tubing string and to enable the packoff around the cable. The ESP equipment selection in high water cut, low GOR wells are relatively straightforward. Equipment selection in high GOR or viscous crude wells, however, can be very complicated. Reda, for example, has developed a sophsticated computer program called COMPSEL to comprehensively analyze alternative equipment selections in such situations. COMPSEL can be used to select equipment or to evaluate previously selected equipment. This capability means that, over time, the engineer can evaluate the fit of a pump as oil well conditions change. If well inflow performance changes significantly, the downhole equipment may need resizing both to optimize production and prevent premature failure. Pump performance is directly affected by parameters such as the formation volume factor, GOR, viscosity, volume of free gas in the pump, and water cut.
770
Cable Flat Cable Extensions Size
6 6 6 Price Adder
6 6 6 Price Adder
6 6 6 Price Adder
6 6 6 Price Adder 6
6 6
Description 3 KV KELTB l t P l 3 KV KELTB l t P l 3 KV KELTB l t P l Per Foot 3 KV KEOTB 3 KV KEOTB 3 KV KEOTB Per Foot 3 KV KEOTB 3 KV KEOTB 3 KV KEOTB Per Foot 3 KV KELB 3 KV KELB 3 KV KELB Per Foot 3 KV KELB 3 KV KELB 3 KV KELB
Price Adder Per Foot 4 3 KV KEOTB 4 3 KV KEOTB 4 3 KV KEOTB Price Adder Per Foot 4 3 KV KEOTB 4 3 KV KEOTB 4 3 KV KEOTB Price Adder Per Faor 4 3 KV KELB 4 3 KV KELB 4 3 KV KELB Price Adder Per Foot 4 3 KV KELB 4 3 KV KELB 4 3 KV KELB Price Adder Per Foot 4 3 KV KEOTB 4 3 KV KEOTB 4 3 KV KEOTB Price Adder Per Foot
Series
375 375 375
Type Armor Monel Monel Monel
Dimensions Inches 0 420 x 1 050
59
length Ft. 40 0 55 0 70 0
0 484 x 1 142 0 484 x 1 142 0 484 x 1 142
29 39 49
40 0 55 0 70 0
27 37 47
40 0 55 0
0 420 x 1 050 0 420 x 1 050
Weights lbs.
34 47
456 456 456
Gab
456 456 456
Monel Monel
0 484 x 1 142 0 484 x 1 142 0 484 x 1 142
456 456 456
Galv Galv Galv
0 485 x 1 150 0 485 x 1 150 0 485 x 1 150
44 60 77
40 0 55 0 70 0
456 456 456
Monel Monel Monel
0 485 x 1 150 0 485 x 1 150 0 485 x 1 150
42 58 74
40 0 55 0 70 0
540 540 540
Galv Galv Galv
0 531 x 1 280 0 531 x 1 280 0 531 x 1 280
36
63
40 0 55 0 70 0
540 540 540
Monel Monel
0 531 x 1 280 0 531 x 1 280 0 531 x 1280
35 48 61
40 0 55 0 70 0
540 540 540
Gab Galv Galv
0 520 x 1 245 0 520 x 1 245 0 520 x 1 245
54 74 94
40 0
540 540 540
Monel
52 72
40 0
Monel
0 520 x 1 245 0 520 x 1 245 0 520 x 1 245
738 738 738
Galv Galv Galv
0 531 x 1 280 0 531 x 1 280 0 531 x 1 280
Galv Galv
Monel
Monel
Monel
50
91
36 50 63
70 0
55 0 70 0
55 0 70 0
40 0 55 0 70 0
Fig 16-21 Flat cable extensions for various motor sizes (Courtesy of Reda Pump Division, TRW Energy Products Group )
771
These factors result in greater intake volume than realized at stocktank conditions. Under routine conditions of high water cuts, no produced gas, low viscosities, etc., the previously described selection technique is sufficient. If, however, the free produced gas or fluid viscosity is of sufficient magnitude to influence “normal” pump performance, a COMPSEL study using the best available productivity and PVT data should be used to accommodate the complexities of these factors. The COMPSEL program contains analytic models for pump performance, reservoir response and fluid characteristics. It utilizes the Chew and Connally (1959) correlations for live fluid viscosity based on the quantity of gas in solution. Another option uses Orkiszewski’s (1967) two-phase vertical flow model to compute pump discharge pressure and horsepower required. Standing’s (1952) correlations are used to provide surrogate PVT data when actual values are not available. The fluid intake volume (oil, water and free gas) can be fixed by inputting the surface production and intake pressure, or the program can utilize a subroutine to calculate drawdown, flowing bottomhole pressure, etc., using the straightline J method. This calculation can be refined, when producing conditions are below the bubble point of the oil, by using the subroutine incorporating the Vogel (1968) technique (IPR method). If PVT data (oil formation volume factor and solution GOR versus pressure) are not available for use in the COMPSEL calculations, the subroutine incorporating Standing’s correlations is automatically utilized in the program. Basically, COMPSEL has two semi-independent logical functions: selection and simulation modes. In selection mode, the program takes a set of inpu‘t data, including well, fluid, and reservoir characteristics, and selects an optimal submersible pump configuration based on a non-overlapping set of optimal flow limits. Conceptually, the program does nothing which could not be done by hand; but the time required to accurately compute manually the pressure, temperature, and fluid conditions between every stage during a pump selection would make pump analysis tedious and difficult. In a selection case, all data of pump intake are calculated first, an initial stage is then selected, and the performance for that first stage is computed. A new pressure, at the top of that stage, is estimated by adding the pressure developed by the first stage to the suction pressure. At this new pressure, a new total fluid flow is calculated and a new stage selected. In this manner, as shown in Sample Problem 16-1, the pump is built up from the bottom. The key to the selection case lies in the program’s ability to calculate a new pressure and, consequently, a new flow and pressure gradient between each stage of the pump. The new flow rate is significant because the total flow determines the next stage selection. The flow changes from stage to stage because, as pressure and temperature increase, gas compressibility and solubility change as do the water and oil viscosities (and the dissolved gas effect on formation volume factor). The pressure gradient (density) change is significant because, in general, the pressure developed by a stage is the gradient times the head in feet, corrected for viscosity. Thus, the selection mode in COMPSEL monitors and updates each of these variables at each stage in the pump selection.
772
In simulation mode, COMPSEL allows the user to input a pump configuration and production environment and then computes for the user the expected flow rate under the input conditions. T h s technique determines the overall efficiency of the system and points out conditions that may dictate resizing the pump to avoid excessive pump wear or performance deterioration due to a stage acting as a choke.
HANDLING, INSTALLATION AND OPERATION
This section provides recommended practice on the handling, installation and operation of an ESP system. Because both safety and economic run-life are dependent on correct procedures, the importance of following the recommended practice cannot be overemphasized. Handling
The downhole components, i.e., motor, pump, protector, and intake, are shipped in a metal shipping box (see Fig. 16-22). The shipping boxes are painted red on one end which should be placed toward the wellhead when the equipment is delivered to
Fig. 16-22. Shipping boxes positioned at the wellhead.
113
0‘
Fig. 16-23. Cable handling procedures
the wellsite. The shpping boxes should be lifted with a spreader chain or bridled with a sling at each end. Severe equipment damage can result from dropping, dragging, or bouncing the boxes. The shipping boxes should never be lifted by the middle of the box only. The cable reel should be lifted using an axle and a spreader bar (see Fig. 16-23). If a fork lift is used, the forks should be long enough to support both reel rims when the reel is picked up from an end. The ends of the cable should be covered or sealed to protect against corrosion. Transformers and switchboards are provided with lifting hooks. To avoid damage, the recommended practice is to lift with a spreader bar in order to maintain a vertical position. Variable-speed drives are normally skid mounted with fork lift slots and lifting eyes. Some VSD models are manufactured with pull bars. Additional information on ESP handling and installation procedures is available in the American Petroleum Institute publication entitled “API Recommended Practice for Electric Submersible Pump Installations” (American Petroleum Institute, 1980). Installation
There are three steps to every ESP installation: (a) well preparation, (b) site layout, and (c) running equipment in the well and start-up. The well preparation procedure involves determining the downhole clearance conditions. Site layout prescribes equipment and rig locations as well as size and capacity. Running equipment in the well and start-up procedures include equipment handling, testing, and responsibility of the rig crew and servicemen. Prior to beginning installation of the ESP equipment, the well must be cleared of any tubing, rods, packers, etc., that would prevent the downhole equipment from
114
reaching target setting depth. The casing flange and wellhead should be examined for burrs or sharp edges. T l s is very important in small-diameter casing, because cable damage can be caused by burrs or edges catching cable bands. A gauge ring should be run in (particularly in 44-in. and smaller casing) to below the setting depth of the downhole equipment. If gauging indicates tight spots, a scraper or reamer should be used to remove the obstruction (scale, paraffin, burrs, or partially collapsed casing). This will ensure adequate clearance for the ESP downhole equipment as it is run into the well. The BOP (if used) should be checked for adequate clearance as well as burrs and sharp edges. Cut-out rams are available for most tubing and cable sizes. They should be installed in the BOP for well control in the event of a “kick” during equipment installation. The pulling unit should be centered over the well as closely as possible. A guide wheel-cable sheave should be safely secured to the rig mast no higher than 30-45 ft above the wellhead. The guide wheel should be at least 54 in. in diameter. The cable reel or spooling truck should be positioned about 100 ft from the wellhead in direct line of sight of the rig operator. One person should be responsible for the cable operation. The responsibilities of this person are to insure: (a) that there is a minimum tension on the cable-the cable should be run at the same speed as the tubing, (b) that the cable is kept clear of power tongs during tubing make-up or break, and (c) that no one stands in front of the cable reel-spooler. The cable junction box must be located at least 15 f t from the wellhead (see Fig. 16-16). The switchboard must be located a minimum of 50 ft from the wellhead and a minimum of 35 ft from the junction box. The junction box is normally located 2-4 ft above ground level to insure adequate air circulation and easy access. The junction box must never be located inside a building. The pump manufacturer’s field representative checks all equipment prior to installation. During installation, his responsibility is to supervise the pulling and/or running of the downhole equipment. All equipment delivered to the wellsite is checked to determine that all components necessary to complete the installation have arrived and are not damaged. Once the equipment, cable, and verification procedures are completed, the assembly and run-in of downhole equipment can begin. The manufacturer’s field representative directs the assembly and checks equipment as it is being run-in. Once the run-in procedures are completed and final electrical tests are made, the manufacturer’s representative will complete the electrical connections. The switchboard underload and overload adjustments are set according to the conditions expected for each well. The pump is then started. Fluid pump-up time and load and no-load voltage and amperage on each phase are recorded. Phase rotation should be carefully checked to insure that the pump is rotating in the correct direction. The quantity of production of oil, gas, and water should be monitored on start-up and then regularly for the time required to achieve stability. A careful study should be made on a pump installation that does not produce as designed. As much information as possible should be gathered to aid in specific identification of the
775
problem and appropriate remedial action. This will insure that subsequent installations will provide satisfactory run-life. Pulling equipment out of a well involves essentially the reverse process of run-in. The equipment and cable should be handled with the same care as when new because it is still valuable. Cable damage and missing bands should be recorded at the depth they occurred to aid in subsequent repair and evaluation. If the equipment failure is judged to be premature, the condition of cable, flat cable, pump rotation, and motor-protector fluid may be useful in determination of the cause of the failure. TROUBLESHOOTING
The purpose of this section is to outline a recommendation for identification and solution of typical ESP problems. The only way a failure can be analyzed, and a solution reached as to its cause, is by data collection. When a problem occurs, one should have as much information as possible. Data which should be routinely compiled on each ESP well is: (1) Production: (a) Water, oil, and gas. (b) Run-life. (c) Unit starts and stops. (d) Dynamic and static fluid levels. (e) Pump setting and perforation depths. (2) Ammeter charts. (3) Well conditions (presence and amounts of abrasives and corrosives, e.g., H,S). (4) Electric power quality (surges, sags, balance, negative sequence voltages, etc.). (5) Visual observations of equipment and cable condition on prior pulls. (6) Reasons for equipment pull (failure, workover, size change, etc.). (7) Bottomhole temperature (BHT). When an ESP well is first put on production, data should be collected daily for the first week, weekly for the first month, and a minimum of monthly after the first month. Production data during the first month is very important, because it will indicate if the pump is performing as designed. If a downhole pressure instrument is installed, operating bottorrihole pressure (BHP) is equally important. The major source of information when troubleshooting an ESP installation is the recording ammeter. The recording ammeter is a circular strip chart accessory mounted in the switchboard, which records the amperage drawn by the ESP motor (see Fig. 16-24). A number of changes in operating conditions can be diagnosed by interpreting ammeter records. Ammeter chart “reading” identifying typical problem situations is .described below. Normal operation A normal chart (Fig. 16-24) is smooth with amperage being at or near motor nameplate amperage draw. Actual operation may be either slightly above or below
116
Fig. 16-24. Normal ammeter chart. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
nameplate amperage. As long as the curve is symmetric and consistent over time, however, operation is considered normal. Normal start-up A normal start-up will produce a chart similar to that shown in Fig. 16-25. The start-up “spike” is due to the inrush surge as the pump comes up to the operating speed. The subsequent amperage draw is high but trends toward a normal level. This is principally due to the fluid level being drawn down to the design TDH, resulting in a high but declining amperage draw.
777
Fig. 16-25. Normal start-up ammeter chart. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
Power fluctuations
Operating ESP amperage will vary inversely with voltage. If system voltage fluctuates, the ESP amperage will inversely fluctuate to maintain a constant load (Fig. 16-26). The most common cause of this type of fluctuation is a periodic heavy load on the primary power system. This load usually occurs when starting-up another ESP or other large electric motor. Simultaneous start-up of several motors should be avoided to minimize the impact on the primary power system. Ammeter spikes can also occur during a thunderstorm accompanied by lightning strikes.
778
Fig. 16-26. Chart showing ammeter spikes (power fluctuations). (Courtesy of Reda Pump Division, TRW Energy Products Group.)
Gas locking
Gas locking occurs as fluid level drawdown approaches the pump intake and intake pressure is lower than the bubble point pressure. This situation is shown in Fig. 16-27. This ammeter chart shows a normal start-up and amperage decline as the fluid level is drawn down. The chart shows erratic fluctuations, however, as gas breaks out near the pump beginning at approximately 6:15 a.m. As the fluid level continues to draw down, cyclic loading of both free gas and fluid slugs leads to increasingly wider amperage fluctuations, ultimately resulting in shutdown at approximately 7:15 a.m. due to undercurrent loading.
119
Fig. 16-27. Ammeter chart showing gas locking. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
There are three possible remedies for gas locking: (a) installation of a gas separator intake and/or a motor shroud; (b) lowering of the setting depth of the pump (but not lower than the perforations unless the motor is shrouded); and (c) reducing the production rate of the pump by using a surface choke (but insuring that the production rate remains within the recommended range for that pump). If the above solutions are not satisfactory, the pump should be replaced with a pump that does not drawdown the fluid level, or the intake pressure must be reduced below the bubble point pressure. Another possible solution is to add a variable-speed drive (VSD) to the existing system. The VSD controls the speed of the pump which, in turn, controls the pump
780
Fig. 16-28. Ammeter chart illustrating fluid pump-off. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
capacity. Thus, the pump output can be fine-tuned to protect against pump-off and gas lock while contributing to improved pump life.
Fluid pump-off Fluid pump-off occurs typically when an ESP is too large in relation to the inflow capacity of the well. This condition is illustrated in Fig. 16-28. This chart shows a normal start-up at 7:OO a.m. and normal operation until approximately 1O:OO a.m. Then amperage draw begins to slowly fall until the underload setting is reached and
781
the pump is shut-down about 2:15 p.m. Subsequent automatic restarts at 4:15 p.m. and 8:15 p.m. produce similar results. The remedial actions are much the same as those listed for gas lock and, in addition, a well stimulation treatment may increase the well’s productivity closer to a match with the pump. In general, cycling an ESP is not conducive to optimum run-life. As a temporary measure, the amount of time delay before automatic restart can be increased if the switchboard is equipped with a Redalert motor controller, for example. This may allow the fluid volume to build up to prevent as high a frequency of shutdown
Fig. 16-29. Ammeter chart showing excessive cycling. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
782
occurrence. Nevertheless, the pump and well are not compatible and the pump should be checked as to size on the next changeout, or the well must be worked over to improve productivity. A form of frequent, short-duration cycling is shown in Fig. 16-29. This shows an extreme pump-off condition. Although the initial reaction is to suspect a badly oversized pump, there may be another cause. If a fluid level sounding, taken immediately after pump shutdown, indicates fluid over the pump, the problem may be a tubing leak or a restricted valve or discharge line. A tubing leak is typically accompanied by somewhat low discharge pressure and low surface production rate.
Fig. 16-30. Ammeter chart illustrating gassy conditions. (Courtesy of Reda Pump Division. TRW Energy Products Group.)
783
If shutdown is due to a plugged valve or discharge line, tubing pressure should be abnormally high. Gassy conditions-emulsion
An ammeter chart of gassy but normal producing well is shown in Fig. 16-30. The continuous amperage fluctuations result from alternating free gas and heavy fluid pumping. Generally, this condition results in a reduction of stocktank barrels in relation to pump design rate. This figure is also typical of an emulsion. The
Fig. 16-31. Ammeter chart illustrating presence of solids and debris. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
784
amperage fluctuations are caused by the frequent, temporary blockage of the pump intake. If it is an emulsion block, spikes are normally lower or below the normal amperage line. Solids-debris
When debris or solids are found in a well, the amperage will display fluctuations immediately after start-up. This condition is shown in Fig. 16-31. Typically, when solids such as sand and scale, or weighted mud are found in a well, special care must
Fig. 16-32. Ammeter chart showing overload condition. (Courtesy of Reda Pump Division, TRW Energy Products Group.)
185
be taken on start-up to avoid pump damage. It may be necessary to put back pressure on the well to prevent excess amperage until the kill fluid is removed and/or sand production begins to decline to a safe volume. Overload shutdown
A pump will also automatically shutdown in an overload condition. This condition is shown in Fig. 16-32. When an overload condition shutdown occurs, however, the unit must not be restarted until the cause of the overload has been identified and corrected. Some motor controller overload detection circuits contain a built-in time delay, ranging from 1-5 sec at 500% of the set point to 2-30 secs at 200% of the set point; they will not automatically restart the unit on an overload condition. A restart attempt in an overload condition can destroy the downhole equipment if the cause of the overload is not first identified and corrected. The most common causes of an overload condition are increased fluid specific gravity, presence of sand, emulsion, and of scale, electric power supply problems, worn equipment, and lightning damage. EXAMPLE PROBLEM 16-1-PUMP GOR
SELECTION ON A WELL WITH A HIGH PRODUCING
Given : Casing size and weight: St-in.; 17 lb/ft Tubing size, weight and thread: 2;-in.; 8rd EUE Well depth (measured and true vertical): 7200 ft Perforations depth: 7025-7050 ft Present and desired production rate: 500 bbl/day Oil production rate: 250 bbl/day Water production rate: 250 bbl/day Gas/oil ratio: 500 scf/bbl Static bottomhole pressure: unknown Producing bottomhole pressure and fluid level: 500 psi at 7000 ft Bottomhole temperature: 160'F Wellhead temperature: 120°F System back pressure (flowlines, separator, and wellhead): 200 psi at wellhead Oil gravity (API): 35" API; viscosity; pour point; paraffin content; sand and emulsion tendency. Water specific gravity: sp. gr. = 1.1; chemical content; corrosion potential; and scale tendency. Gas specific gravity: 0.65 with respect to air Reservoir formation volume factor: FVF = B = 1.077; bubble point; and viscosity- temperature curve. In t h s example, it is assumed, that the gas ingestion percentage ( G I P ) is 50%, i.e., that 50% of the free gas is produced through the pump and the remaining 50% is
786
2
,33%,:7f,D6,39?
k3tR75°/*G'P
3 500 STBPD 500 GOR 100%GIP 50% OIL, 50% W A T E ~ 4 500 STBPD, A L L WATER
3200 2800
200016001200-
2400a
g Y) 3
$ n.
i,
400 800
'0
I
200
400
600
800
2
3
I000 1200 1400
CAPACITY, bbl/day
Fig. 16-33. Capacity curves for example problem 16-1. BHT = 160 O F .
produced up the annulus. The first step is to determine the pump discharge pressure required in this well. This can be done using any one of several multiphase flow correlations. The most commonly used methods are those of Hagedorn and Brown (1965), Orkiszewski (1967), Duns and Ros (1963), Beggs and Brill (1973), and Poettmann and Carpenter (1952). Figure 16-33 shows four different oil, water, and gas combinations producing 500 stocktank barrels: (1) 100% water (2) 50% water; 50% oil; 50% GIP; GOR = 500 scf/bbl (3) 50% water; 50% oil; 75% GIP; GOR = 500 scf/bbl (4) 50% water; 50% oil; 100% GIP, GOR = 500 scf/bbl As a higher percent of free gas is produced through the pumps (GIP), discharge pressure decreases. The intake volume at the pump, however, increases dramatically. In t h s example, the Hagedorn and Brown (1965) correlations are utilized. The vertical flowing pressure gradients are presented in Fig. 16-34. In order to utilize tlus correlation, the gas/liquid ratio (GLR)must be calculated first. Then the pump discharge pressure is determined.
R,
=
( R - R , intake) GIP + R , = (500 - 80) X 0.5 + 80 = 290 scf/bbl
and
GLR through the pump = R , X GIP = 290 X 0.5 = 145 scf/bbl,
787 a
4
8
12
16
20
28
Fig. 16-34. Vertical flowing pressure gradients for various gas/liquid ratios (50% water and 50% oil). (Courtesy of Reda Pump Division, TRW Energy Products Group, and PennWell Publishing Company: Brown, 1980, p. 454.)
where R, = cumulative produced gas/oil ratio, scf/bbl, R = total producing gas oil ratio, scf/bbl; R , = solubility of gas in oil at 500 psi (gas in solution), scf/bbl; GIP = gas ingestion, percent; and GLR = gas/liquid ratio, scf/bbl. Using Fig. 16-34, the 200 psi wellhead pressure is equal to 1500 ft of head. Thus, the total head at 7000 f t is equal to: 7000 1500 = 8500 ft. Given 8500 f t head and
+
788
a GLR of 145 scf/bbl, from Fig. 16-34 the pressure required to flow is 2630 psig at the pump discharge. Next, the volume at the pump intake and successive stages is computed given the bottomhole temperature and pressure: Volume intake = (oil + dissolved gas)
+ (water) + (gas)
Y n = qo(1- 4)'0 + qO(W,)Bw + qo(1- K > ( R- Rs)(Bg)(GIP) where qo = total fluid; Wc= percent water cut; B, = formation volume factor (FVF) for oil; B, = formation volume factor of water; Bg= formation volume factor of gas; R,= solubility of gas in oil, scf/bbl; R = total gas/oil producing ratio, scf/bbl; and GIP = gas ingestion percent. Standing's correlations are used here to determine the solution gas/oil ratio at various pressures and the formation volume factor of oil. Refer to Figs. 16-35 and 16-36. (a) Determine how many scf of gas are dissolved in a barrel of oil at 500 psi intake pressure using Fig. 16-35: At a pressure of 500 psi, 80 cu ft of gas are dissolved in a barrel of oil at 160°F BHT. (b) From Fig. 16-36, at 80 scf/bbl, formation volume factor of oil ( B , ) is equal to 1.077. As fluid moves through the pump, pressure increases and volume decreases. These charts enable the determination of the gas in solution and the B, at each incremental step in the pump calculation. V,, = 500(1 - 0.5)1.077 + 500(1 - 0.5) X 500(1 - 0.5) X (500 - 80)0.5(0.00577) = 269.25(oil) + 250(water) 302.93(gas) = 822.18 bbl where 0.00577 = Bg = 0.00504ZT/p = 0.00504 x 0.95 x 620/514.7 (bottomhole conditions), 2 = 0.95 (compressibility factor obtained from Fig. 16-37; T = 620"R; and p = 514.7 psia (submergence pressure). The same method of calculating V,, is used to determine the volume at successively higher pressures until the discharge pressure of 2630 psi is reached. The COMPSEL program performs this calculation in one-stage increments. In t h s example, 50-psi increments are used to shorten the calculation process. On successive calculations, the value for GZP is 100% and, consequently, R becomes equal to R,, where: R , = ( R - Rs)GZP+ R,; ( R - R,) = gas going up the casing; and R , = (500-80)0.5 + 80 = 290 scf/bbl. Gas/liquid ratio through the pump = total fluid X oil cut = 290 X 0.5 = 145 scf/bbl. Thus the equation becomes: in' = q(1- wc>[Bo+ (Wc/' - Wc) + ('p - Rs)Bgl. For 550 psi: V,, = 500(0.5)[1.08 + 1 (290 - 88)0.00539] = 250[1.08 + 1 1.091 = 792.5. For 600 psi: V,, = 250(1.084 + 1 + [(290 - 100)0.00480]) = 749 bbl/day.
+
+
+
-1 W
Fig. 16-35. Determination of bubble point pressure for crude oils from California. (After Standing, 1952; courtesy of Chevron Research Company.)
W
790
EXAMPLE ffEQUIffED: Formation vo/ume at POO'F of a bubble point fiquid having a gar-oil ratio of 350 CFB, a g a s g r ~ v i t yof 0.75, a n d a h n k o,/ g r a v i t y o f 30 .AP/.
PRO c E o u a E : Starting at the / r f t side of the chart, proceed horizonta//y d u n 9 t h e 350 CFB /inr to a gas gravgty of 0. 7 5 . From this point d r o p vwtica//y fo the 30%P/ /in= Proceed horrzontaNy from fie tank oil gravity scde to the 200'F h e . The requirrd formation volume is rbund to be f.22 barrel ner bar-/ of t a n k od.
-~
~
FORMATION
VOLUME of BUBBLE POINT LIQUID
-
Fig. 16-36. Oil formation volume factor of California natural hydrocarbon mixtures of gas and liquid at bubble point. (After Standing. 1952; courtesy of Chevron Research Company.)
791
Fig. 16-37. Compressibility factors for a 0.65 gravity natural gas. (After Brown and Beggs, 1977, fig. 2.17, p. 86; courtesy of PennWell Publishing Company.)
For 650 psi: V, = 250(1.093 1 [(290 - 117)0.00441]) = 714 bbl/day. For 700 psi: = 250(1.095 + 1 [(290 - 119)0.00410])= 699 bbl/day. For 750 psi: = 250(1.097 1 [(290 - 130)0.00380])= 676 bbl/day. For 800 psi: F, = 250(1.102 1 + [(290 - 138)0.0035])= 658 bbl/day. For 850 psi: = 250(1.109 1 + [(290 - 152)0.0033])= 641 bbl/day.
F, F,
F,
+ + + + + + +
792
For 900 psi: V,, = 250(1.112 1 + [(290 - 160)0.0031])= 629 bbl/day. For 950 psi: V,, = 250(1.115 1 + [(290 - 171)0.0030])= 618 bbl/day. For 1000 psi: V,, = 250(1.119 + 1 + [(290 - 180)0.0028])= 607 bbl/day. For 1200 psi: V,, = 250(1.136 + 1 [(290 - 255)0.0023]) = 554 bbl/day. For 1400 psi: V,, = 250(1.164 + 1 [(290 - 275)0.0019]) = 548 bbl/day. For 1600 psi: V,, = 250(1.18 + 1 + [(290 - 310)0.0019]) = 536 bbl/day. At high pressures, R , increases to become equal to or to exceed the value of R , (specifically, above 1600 psi). When this occurs, there is no free gas. In Figs. 16-35 and 16-36, R , = R,, and at 290 scf the pressure is equal to 1535 psia; at a pressure of 1535 psia, B, = 1.175. At pressures of 1535 psi through 2630 psi, the volume is equal to: V,, = 250(1.175 + 1 + [290 - 29010.0019) = 544 bbl/day. The specific weight (y,) of gas moving through the pump can be calculated using the following equation:
+ +
+ +
Y, =
28.97 X SG, X p ZR T
where SG, = specific gravity of gas, with respect to air; p = pressure of 1 atmosphere; 2 = compressibility factor; R = gas constant; and T = temperature, OR. At standard conditions:
=
28.97 X SG, X p 1 X 10.73 X 520
= 0.0496
lb/cu ft.
The weight of the gas is equal to: x GLR = 0.0496 X 500 = 24.8 lb/bbl. The specific weight of oil, yo, is equal to: yo = SG, x 350 lb/bbl = 0.8489 X 350 = 297 lb/bbl. The specific weight of water, y,, is equal to: yw = SG, x 350 lb/bbl = 1.1X 350 lb/bbl. Thus, the weight of one stocktank barrel = W, W, + W, = (24.8 X 0.5) + (297 X 0.5) (385 x 0.5) = 353.4 lb/bbl. The net weight W, is equal to: W, = 353.4 - 5.21(gas up casing) = 348.19 lb/bbl. The weight rate of flow of 500 stocktank barrels per day = 500 X 348.19 = 174,095 lb/day. The weight of the pump fluid intake = 822 X 350 lb/bbl (at SG = 1.0) =
y,
+
+
793
287,700 lb/day. Thus, specific gravity is equal to: 174,095/287,700 = 0.61 and pressure gradient = 0.61 X 0.433 psi/ft = 0.264 psi/ft. The pressure gradients at various pressures are presented below: Pressure (Psi)
500 550 600 650 700 750 800 850 900 950
lo00 1200 1535
(bbl/day)
Specific gravity SG
822 792 749 714 699 676 658 641 629 618 607 554 544
0.610 0.628 0.664 0.697 0.712 0.736 0.753 0.776 0.791 0.805 0.819 0.871 0.914
Flow
Pressure gradient (PSi/ft) 0.264 0.272 0.288 0.302 0.308 0.319 0.326 0.336 0.342 0.349 0.355 0.377 0.396
The pump selection for the pump intake rate of 822 bbl/day through the pump discharge rate of 544 bbl/day would initially utilize the DN-750 pump (see Fig. 16-38). The initial rate of 822 bbl/day is very close to the peak efficiency of the DN-750 pump. At pressures of 500 and 550 psi, the average volumetric rate of flow is equal to: [(822 + 792)/2] = 807 bbl/day; the average specific gravity is 0.62; and the pressure gradient is 0.268 psi/ft. The DN-750 pump develops about 22.7 ft/stage at the 807 bbl/day rate; thus: pressure in psi per stage = pressure gradient X ft/stage
= 0.268 X
22.7 = 6.08.
The number of stages is equal to: total pressure/(pressure/stage) = 50/6.08 = 8.22. Horsepower per stage at 807 bbl/day is 0.235. Thus, after correction for specific gravity, the horsepower required for 8.22 stages is equal to: 8.22 x 0.235 x 0.62 = 1.19 HP. For additional ranges: Pressure range (Psi) 500-550 550-600 600-650
Average flow rate (bbl/day) 807 770 731
Number of stages
Horsepower
8.22 7.60 6.92 22.74
1.19 1.14 1.06 3.39
794
795 main S t o p -. P r o g r a m t e r m i n a t e d . DATE 3/21/1984 NEW C A S E
-
TIME
C O R R E C T I N G F'LIMP FERFORMANCE F O R V I S C O S I T Y G I F OVER ONE. WILL. D l V l D E B Y 100. COOPER 1.0 3.0 WRFIDE 1. 1 1 l:ll:l~:ll:~ .1 5 0 0 . I;I~:I~>~:I PRESSLIRE 2 BARRELS F R E E .9886 30. 0445 1 2 7021:) 7. 4507 -, 9'7 44 3.7'1.57 .9465 2.3677 4 1.6527 .95e7 5 95 10 1 .:'I I0 6 .Y4.16 Y 11.14 '7 c,921 8 :'j 2 4 Y 4ASb 1 1:)
.
. . .
.
. . I
11 12 i3 14 15
7'78.2
GR
.2 h 10 .2095 . t9r.19 .31R7
I
3650
.:.ti44
Fig. 16-39. Computer printout of Sample problem 16-1.
-,
-
12:36:54
pm
796
.
F i 4 I hh 541.M
TOTAL DYNAMIC HEAD MINIMCIM FUMF DISCHARGE PRESS!JRE PRESSURE DHUP ACROSS F'IJMF
****
5Z9CI. 971)C) 2 5 2 8 . .3:)l:)l:l 2 128. 3l:~l:ll:l
RESERVOIR DATA AND PUMP SEL.ECTION R E S U L T S
*++*
WELL NAME E X A M F L F F'R0BLk.M # 1
FLUID C H A R A C T E K I S T I C S
WELLBOFE: CONDI T I O N 9
5ol:l. I:Kl SCF/BBL. GOR & ;:,a. ClO DEti RP I O I L GRAVITY GAS G R A V I T Y hS c. WATER CUT . ~ l , >00 . PERCENT VISCOSITY 8.00 CP A T 11:)o.oO F: VISCOSITY 2.415 CP A T 200.00 F MEAN TEMPERATLIRE 140.10 DEE F F I 1 . l-)l:) BF'D/I;'ST
..
Fuw
C r m F I GUHAT I or4
60 CYCLE POWER
.
ME: AS1.IRE.I) DEP T t i
SIBHP INTAYE' F'RESSURE WELL..HEAD PRESSURE PERFORATIONS ROTTOM HOLE TEMP FREE G A S A T I N T A K E 'TLIBINO S I Z E VERT PUMP DEPTH WELLIiEAD TEMF FUMF' FERFURMANCE
I NTAI.: E VOLUME
D I SCHARGE VOLCIME DISCHARGE PKESSUFiE DR A w Do w IN
FBHF' STAGES
82 264 DATE .-
TYPE DN-.7St:)
D E S I R E D FLOW HOF(sjEp:mOWER
822.38 541.66 2628.30 501:i.oo
BFD HPD PSI FSI 5":l 1:10 PS I 500.00 BPD cc ad. 85
.
D--5513 ;/2i/i9a4
TIMF
-
1:2:47:25
pm
Fig. 16-39 continued.
Using the same approach, the DN-750 stage is utilized followed by the D550,until reaching the target rate of 544 bbl/day and 2630 psi discharge pressure. The COMPSEL selection for this problem was 82 stages, DN-750,and 264 stages, D550. The horsepower requirements were 55.85 HP after the well is stabilized at a design rate of 541.7 bbl/day. The design discharge pressure is 2628.3 psi. An important point to remeber is that the COMPSEL designed pump is at a stable producing rate. If the well was killed with a heavy fluid, additional stages and
797
horsepower may be required to unload the well. Figure 16-39 represents the computer printout of this problem.
SAMPLE PROBLEMS A N D QUESTIONS
(1) Determine the equipment series required, given the following information: (a) Well total depth ( T D ) = 3900 ft (b) 8;-in., 36-lb casing to a depth of 3000 ft (c) 6;-in., 24-lb liner to TD; top at 2950 ft (d) Depth of perforations = 3000-3450 ft (e) Pump setting depth = 3250 ft ( f ) Flow rate: 2100 bbl/day STB-100 bbl/day of oil and 2000 bbl/day of water (g) Tubing size: 2$-in., 8rd EUE (h) Maximum fluid entry is below 3300 ft. (2) Given the following information, design the pump, motor, cable string, and transformer for an optimum ESP installation: (a) Casing: 6i-in., 24-lb, I D = 5.921 in., drift = 5.796 in. (b) Setting depth: 4000 ft vertical depth (VD), 4000 ft measured depth ( M D ) (c) Bottomhole temperature = 150 O F (d) Tubing: 2$-in. EUE (e) Surface pressure = 100 psi ( f ) Water cut = 90%; sp. gr. water = 1.0 (g) Voltage supply: 480 V (h) Desired gross production rate: 1500 bbl/day (i) PFL (lift) = 3800 f t TDH = Net lift + Friction loss + System back pressure. (3) Determine the expected production rate, given the following information: (a) Gas absent (b) Specific gravity = 1.1 (c) Lift: 5000 ft vertical; 7000 ft measured. (d) Downhole pressure unit: 100 psi (e) Tubing pressure = 150 psi ( f ) Installed pump: 230-stage, DN 750 (g) Motor length = 17.5 ft (h) Protector length = 5.3 ft. (4) Find (a) pump size (OD) from the manufacturer’s selection chart, and (b) stage type (i.e., flow rate recommended range) using manufacturer’s selection chart, given the following information: (a) Casing: 7-in., 23-lb (b) Tubing: 3-in., 8rd EUE (c) Bottomhole temperature = 175O F (d) Wellhead temperature = 130 O F (e) Tubing pressure = 280 psi
798
( f ) GOR = 1290 cu ft/bbl
(g) Specific gravity of gas = 0.674 (air = 1.0) (h) Oil gravity = 47 OAPI (i) Water cut = 28.6% 0) Pump intake pressure = 1140 psi (k) Gross production rate = 1900 bbl/day (1) Gas separator efficiency = 50% (m) PVT data:
Pressure (Psi)
0
97
253
269
500
750
1000
1250
Rs (cuft/bbl)
BO ZatBHT
599 375 491 0 239 300 1.47 1.54 1.34 1.40 1.08 1.29 0.90 0.95 0.92 1 0.98 0.97
704 1.59 0.89
809 1.65 0.87
(5) Given the following data: (a) 6850 No. 2 cable; maximum amp. = 70 A (b) SBHP = 2500 psi at 7000 ft datum; 8241-1. casing (c) FBHP = 1800 psi (d) Tubing: 2H-in. (e) Present total production = 525 bbl/day (75 bbl/day of oil) ( f ) ftube: 17 ft/lOOO f t (g) GOR = 80 cu ft/bbl (h) B H T = 160°F (i) Perforations depth: 6900-7450 ft 0) Surface pressure = 100 psi (k) Pump 17 = 60% (1) Motor 17 = 90% (m) Bo = 1.2 (n) PF = 0.9 (0)Fluid gravity data: sp. gr. oil = 0.904 (25OAPI), sp. gr. water = 1.02, sp. gr. gas (air = 1)= 0.7. (p) Pump setting depth = 6800 ft (q) Pump intake desired pressure = 200 psi Determine: (a) Shut-in BHP at the pump (b) Production capacity: (i) using IPR curve, (ii) using J ( P I ) (c) Determine volumetric rate of flow (reservoir bbl) at intake (d) Determine friction loss (e) Determine TDH (total dynamic head) = Lift + Tubing friction loss + Wellhead pressure - Intake pressure
799
( f ) Determine hydraulic I-€'(H,) (g) Determine motor Ep(H,,,) (h) Calculate the minimum motor voltage (6) What factors determine the maximum setting depth of the ESP? Why? (7) Describe different types of pump intakes. (8) What factors determine the maximum diameter of the pump, motor, and protector? Why?
REFERENCES American Petroleum Institute, 1980. API Recommended Practice for Electric Submersible Pump Installation. RP 11R. API, Dallas, Tex. American Petroleum Institute, 1982. Recommended Practice for the Operation, Maintenance and Trouble Shooting of Electric Submersible Pump Installations. API, Dallas, Tex. Beggs, H.D. and Brill, J.P., 1973. A study of two-phase flow in inclined pipes. J. Pet. Tech., 25(5): 607-617. Brown, K.E., 1980. The Technology of Artificial Lift Methob, Vol. 2a. PenWell, Tulsa, Okla., 720 pp. Brown, K.E. and Beggs, H.D., 1977. The Technology of Artificial Lift Methods. Vol. 1. PennWell, Tulsa, Okla., 487 pp. Chew, J. and Connally Jr., C.A., 1959. A viscosity correlation for gas-saturated crude oils. Trans. A I M E , 216: 23-25. Duns, Jr., H. and Ros, N.C.J., 1963. Verticalflow of gas and liquid mixtures in wells. Proc. 6th World Pet. Congr., Frankfurt, p. 451. Hagedorn, A.R. and Brown, K.E., 1965. Experimental study of pressure gradients occumng during continuous two-phase flow in small-diameter vertical conduits. J. Pet. Tech ., 17(4): 475-484. Lasater, J.A., 1958. Bubble point pressure correlation. J. Pet. Tech., lO(5): 65-67; Trans. A I M E , 213: 379-381. Martin, J.W. and Vatalaro, F.J., 1979. Testing of Oil Well Power Cables Under Simulated Downhole Conditions. TRW Reda Pump Div. Publ. Mead, H., 1975. Oasis submersible lift operations. SPE 5287. Paper, presented at the Soc. Pet. Eng. European Spring Meeting, London, April 14-15, 5 pp. O'Neil, R.K., 1976. Application and selection of electrical submergible pumps. SPE 5907. Paper, presented at the Annu. Rocky Mountain Regional Meet., Casper, Wyo., May 10-11, 12 pp. Orkiszewski, J., 1967. Predicting two-phase pressure drops in vertical pipes. J. Pet. Tech., 6: 829-838. Poettmann, F.H. and Carpenter, P.G., 1952. The multiphase flow of gas, oil and water through vertical flow strings with application to the design of gas life installations. Drill. Prod. Prac., A P I : 257. Standing, M.B., 1975. A pressure-volume-temperature correlation for mixtures of California oils and gases. Drill. Prod. Prac., A P I : 275. Standing, M.B., 1952. Volumetric and Phase Behavior of Oil Field Hydrocarbon Systems. Reinhold, New York, N.Y. Schultz, H.F., 1973. Extraordinary application of electrical submergible centrifugal pump equipment. S P E 4723. Paper, presented at the South Plaine Prod. Tech. Symp., Lubbock, Tex., Nov. 1-2. Swetnan, J.C. and Sackash, M.L., 1977. Performance review oftapered submergible pumps in the Three Bar (Devonian) Field. SPE 6854. Paper, presented at the 52nd Annu. Fall Tech. Conf. and Exhibition, Dallas, Tex., Oct. 9-12, 5 pp. Vasquez, M. and Beggs, H.D., 1980. Correlations for fluid physical property predictions. J. Pet. Tech., 32(6): 968-970. Vogel, J.V., 1968. Inflow performance relationships for solution-gas drive wells. J. Pet. Tech,, 20(1): 83-92; Trans. A I M E , 243: 83-92.
r~
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
801
REFERENCES INDEX * Numbers in italics refer to references lists Abraham, H., 147 Agnew, B.G., 413 Alden, R.C., 254, 256, 277 Al-Hussainy, R., 357, 370 Allen, T.O., 373, 374, 413 Amin, M.B., 51, 58 Anderson, R.L., 529 Andrews, D.E., 291, 326 Babson, E.C., 529 Baker, O., 293, 326 Bancroft, T.A., 529 Bansbach, P.L., 147 Barcenas, G.H., 414 Barnickel, W.S., 114, 132, 147 Barry, E.G., 253, 276 Beal, C., 244-246, 276 Beasly, W.L., 620, 633 Beauregard, E., 467, 469. 529, 530 Beeson, C.M., 108, 147, 204, 210, 429, 465, 467, 528, 529 Beggs, H.D., 222, 244, 246, 258, 276, 277, 279, 281. 282, 286, 287, 290, 292, 296, 326, 363, 371, 464, 760, 786, 799 Bessler, D.U., 147 Bicher, L.B., 49, 58 Binder. R.C., 23-25, 27, 41 Bird, R.B.. 259 Blair, C.M.. 124 Bogart, A.J., 391-400, 413 Bommer, P.M., 620, 633 Bonnet, J.L., 414 Borden, ti., Jr., 661, 736 Brear, A,, 427 Brigham, W.E., 368, 371 Brill, J.P., 222, 258, 276, 286. 287, 290, 292, 326, 786, 791, 799 Brons, F., 346. 355. 371 Brown, G.G., 67, 108, 254, 256, 260, 276, 277 Brown, K.E., 277. 279, 281, 282, 286, 291, 296, 300, 326, 363, 371, 416, 419, 428, 429, 432, 439, 441, 442, 445, 449, 452-457, 459, 460.
464, 465, 539, 582, 609, 633, 786, 787, 791, 799 BUITOWS, D.B., 257-259, 277 Byrd, J.P., 537, 539, 569, 579, 580, 582, 593, 608, 609, 619, 620, 633 Campbell, A.T., Jr., 347, 371 Campbell, J.M., 73-76, 85, 108, 277 Cantrell, R.C., 414 Capps, B., 449, 465, 531, 634, 736 Carpenter, P.G., 286, 326, 786, 799 Carr, N.L., 49, 58, 257-259, 277 Carter, R.D., 371 Champion, C.A., 413 Chernikin, V.I., 239, 250, 277 Chew, J., 245, 246, 277, 771, 799 Chilingar, G.V., 41, 108, 147, 159, 170, 204, 210, 413, 429, 465, 467 Chilton, C.H., 69. 108, 219, 261, 277 Clavier, C., 413 Clayton, W., 147 Clegg, J., 449, 465, 531, 634, 736 Cleveland, R.G., 291, 326 Cobb, W.M., 346, 372 Coberly, C.J., 736 Colebrook, C.F., 277 Connally, C.A., Jr., 245, 246, 277, 771. 799 Conti, V.J., 252, 277 Cooke, C.E., Jr., 414 Coombs. C.E., 147 Corneil, D., 178, 210, 266, 277 Craft, B.C., 85, 108, 214, 277, 465, 561, 568. 573. 574, 579. 580, 582, 609, 633 Cragoe, C.S., 247, 277 Crawford, P.B., 357. 370 Curtis, M.R., 414 Dale, J.R., 414 Davenport, T.C., 252, 277 Davies, D.L.. 495, 496, 529 Day, J.J., 539, 569, 579, 580. 582, 593, 608, 609, 619, 633
* Prepared by Mehmet Parlar and George V. Chilingarian.
802 Deam, J.R., 83, 108 DeGroote, M., 147 DePriester, C.L., 52-54, 58 Dietz, D.N., 346, 355, 371 Dodd, H.V., 147 Doering, M.A., 414 Doherty, W.T., 277, 417, 465, 582, 634 Dolan, J.P., 371 Doty, D.R., 633 Dougherty, E.L., 270, 277 Dow, D.B., 147 Dukler, A.E., 291, 326 Duns, H., Jr., 287, 288, 326, 786, 799 Dyes, A.B., 355, 356, 371 Eakin, B.E., 258, 277 Earlougher, R.C., Jr., 326, 348, 351, 354, 367-369, 371 Eaton, B.A., 291, 326 Edmister, W.C., 261, 277 Eickmeier, J.R., 599, 619, 633 Einarsen, C.A., 371 Elenbaas, J.R., 178, 210, 266, 277 Erbar, J.H., 83, 108 Eubanks, J.M., 611, 633 Ezekial, M., 485, 530 Fagg, L.W., 609, 633 Falade, G.K., 368, 371 Ferguson, P.L., 467, 469, 529, 530 Fetkonch, M.J., 359-365, 371 Flanigan, O., 292, 326 Focazio, K.R., 634 Franks, B.L., 611, 633 Frick, T.C., 41, 108, 244, 277 Fuentes, A.J., 290, 326 Gardner, L.W., 326 Gatlin, C., 333, 371 Gibbs, S.G., 561, 633 Gibson, J.A., 347, 371 Gilbert, W.E., 294, 326, 458, 465 Gipson, F., 449, 465, 531, 634, 736 Gonzalez, M.H., 258, 277 Graves, E.D., Jr., 85, 108, 214, 277, 465, 561, 568, 573, 574, 579, 580, 582, 609, 633 Greenlee, R.W., 124 Gregory, A.R., 326 Griffin, F.D., 633 Gurwitsch, L., 147 Hagedorn, A.R., 286, 300, 326, 786, 799 Hammack, G.W., 414
Hansen, W.P., 233, 277 Hartnett, J.P., 242, 277 Hazebroek, P., 355, 371 Hein, W.G., 159, 170 Heinze, F., 266, 277 Hill, G.A., 371 Hirata, M., 233, 277 Hirschfelder, J.O., 259 Hohmann, E.C., 82, 108 Holden, W.R., 85, 108, 214, 277, 465, 561, 568, 573, 574, 579, 580, 582, 609, 633 Homer, D.R., 353, 356, 371 Horton, W.D., 518, 530 Hougen, O.A., 51, 58 Hoyle, W., 413 Hurst, W., 344, 350, 371, 372 Hutchinson, C.A., Jr., 355, 356, 371 Ikoku, C.U.,60, 71, 108
Jackson, B.R.,413 Johnson, J.L., 159, 170 hhnson, W., 414 Juch, A.H., 296, 326 Kamal, M., 371 Kato, H., 233, 277 Katz, D.L., 49, 58, 178, 210, 254-256, 266, 277 Kazemi, H., 347, 371 Kelly, H.L., 581, 633 Kelvin, Lord, Sir, W.T., 343, 371 Kempthome, O., 530 Kennedy, J.L., 199, 210 Kirkpatrick, C.V., 441, 465 Knowles, C.R., 291, 326 Knox, D.G., 528, 529 Kobayashi, R., 178, 210, 257-259, 266, 277 Lasater, J.A., 760, 799 Laurence, L.L., 108 Lawrence, D.K., 611, 633 Lea, J.F., 469, 530 Lebeaux, J.M., 530 Lee, A.L., 258, 277, 442, 465 Lenoir, J.M., 48, 51, 58 Lloyd, F.T., 530 Lockhut, F.J., 51, 58, 78-82, 108 Lockhart, R.W., 291, 326 Loeb, J., 414 Ludwig, E.E., 277 Lynch, E.J., 321, 329, 371
803 Mach, J., 460, 465, 539, 582, 609, 633 Maddox, R.N., 51, 58, 83, 108, 188, 210 Makowski, M.M., 234, 277 Marks, A., 277 Marsh, H.N., 611, 634 Martin, J.W., 799 Martinelli, R.C., 291, 326 Matthews, C.S., 350, 355, 371 Maxwell, T.E., 611, 633 McHenry, R.J., 78-81, 108 McMurry, E.D., 530 Mead, H., 799 Merryman, C.J., 611, 633 Meunier, D., 413, 414 Miller, C.C., 355, 356, 371 Miller, F.G., 369, 371 Miller, S.C., 371 Miller, W.C., 346, 371 Mills, K.N., 570, 578, 634 Mochlinski, K., 234, 277 Monson, L.T., 124, 147 Moody, L.F., 214, 215, 225, 277 Moms, B.P., 414 Momson, W.E., 210 Moulin, J., 414 Mueller, T.D., 344, 345, 369, 371 Muskat, M., 371 Myers, B.D., 414 Neely, B., 449, 465, 531, 634, 736 Nellensteyn, F.J., 147 Nelson, W.L., 21-23, 28, 38, 41, 44, 51, 56-58, 83, 86, 108 Ney, C., 290, 326 Nishiwaki, N., 233, 277 Oberfell, G.B., 254, 256, 277 Odeh, AS., 357, 372 O’Neil, R.K., 799 Orkiszewski, J., 771, 786, 799 Osborne, A., 83, 108 Papadopulos, I.S., 366, 372 Parent, J.D., 49, 58 Peacock, D.R., 414 Peck, R.E., 49, 58 Peng, D., 83, 108 Perrine, R.L., 369, 372 Perry, R.H., 69, 108, 219, 261, 277 Poettmann, F.H., 178, 210, 266, 277, 286, 326, 786, 799 Poupon, A., 414 Praisnar, A,, Jr., 465 Prats, M., 368, 372
Ramey, H.J., Jr., 346, 357, 366, 367, 370, 372 Rankine, A.O., 259 Reed, C.E., 485, 486, 530 Reistle, C.E., Jr., 147 Riley, H.G., 371 Ritter, R.B., 48, 58 Roberts, A.P., 373, 374, 413 Roberts, C.H.M., 124 Robinson, D.R., 83, 108 Robinson, J.R., 244, 246, 276 Rogers, W.F., 163, 170 Rohsenow, W.M., 242, 277 Ros, N.C.J., 287, 288, 326, 786, 799 Russel, D.G., 350, 371 Russel, J.H., Jr., 599, 634 Rzasa, M.J., 661, 736 Sackash, M.L., 799 Schaller, H.E., 413 Schmidt, Z., 633 Schultz, H.F., 799 Schweppe, J.L., 48, 58 Selig, F., 357, 372 Shaw, S.F., 432, 433, 465 Shea, G.B., 124 Shenvood, T.K., 485, 486, 530 Silberberg, I.H., 291, 326 Sivalls, C.R., 108 Smith, C.J., 259 Smith, D.P., 414 Smith, H.V., 59, 108 Smith, R.C., 414 Sorg, K.G., 259 Souders, M., 67, 108 Spotz, E.L., 259 Standing, M.B., 255, 277, 284, 285, 326, 662, 663, 736, 760, 771, 789, 790, 799 Steffensen, R.J., 414 Stenzel, R.W., 124, 147 Stoddard, J.H., 528, 529 Stoner, M.A., 277 Stormont, D.H., 166-1 70 Streeter, V.L., 277 Sudduth, L.F., 530 Swetnan, J.C., 799 Swigart, T.E., 124 Szilas, A.P., 214, 223, 227-229, 235, 239, 249, 250, 252, 264, 269, 277, 532, 634 Thomas, R., 243, 277 Thrash, P.J., 419, 439, 441, 442, 449, 456, 457, 459, 465 Tissot, B.P., 39, 41 Tixier, M.P., 414
804 Trammel, P., 465 Trantz, M., 259 Trader, R.N., 147 Treadway, R.B., 634 Turner, W.C., 243, 277 Uren, L.C., 108 Van Everdingen, A.F., 344, 350, 372 Van Poollen, H.K., 347, 372 Vary, J..4., 178, 210, 266, 277 Vasquez, M., 160,799 Vatalaro, F.J., 799 Vogel, J.V., 281, 283, 326, 759, 771, 799 Wade, R.T., 414 Waterman, L.C., 124 Watson, K.M., 57, 58 Wattenbarger, R.A., 357, 372 Watts, E.V., 611, 634
Weinaug, C.F., 178, 210, 266, 277 Weller, W.T., 369, 372 Welte, D.H., 39, 41 Whiney, K.F., 74-76, 85, 108 Wicks, M., 291, 326 WiLkins, R.B., 108 Williams, A.R., 147 Willis, R.M., 581, 633 Wilson, H.M., 199, 210 Wilson, P.M., 449, 465, 531, 634, 736 Winkler, H.W., 465 Witherspoon, P.A., 344, 345, 371 Witterholt, E.J., 414 Woodruff, W.E., 391-400, 413 Worley, M.S., I08 Wylie, E.B., 277 Wylie, M.R.T., 326 Zaba, J., 277, 417, 465, 582, 608, 634
805
SUBJECT INDEX * Absorber, 194 -, bubble tray, 194 Absorption process, 193, 197, 208 - -, oil, 193, 197 - - in natural gasoline treatment, 208 Acceleration factor, 569 - - for rod string, 569 Acid gas, 178, 179, 182, 183, 188 - - content of natural gases, 179, 182, 183 - - removal from natural gas, 188 Adsorbent treater, 187 - -, solid, 187 Absorption process for, - - _ gas treating, 187 - - - liquid extraction from natural gas, 191 Agmg of emulsions, 112, 128 Agitation for demulsification, 133, 144 Air, - content of natural gases, 183 - lift, 415 Aikane. 39 Alkene, 40 Allowable working pressure, - - _ of pipeline, 212 Ammeter, 751 - charts for troubleshooting ESP systems, 775-784 -, functions of, 751 Anchnr, - pipe (perforated), 329 - shoe, 329 Anisotropic formations, 366, 367 - -, permeabilit,y matrix for, 366 - -, permeability nomenclature for, 367 - -, pressure gradient for, 366 API, 537. 538, 583-593, 611-618 - pump designation, 538 - recommended procedures for problem well analysis, 611-67 8 - recommended procedure for sucker rod pumping design, 583-593
- subsurface pump classification, 537 API gravity, 416-418 - - vs. specific gravity and pressure gradient, 700, 701 Aromatic hydrocarbons, 37 Automatic custody transfer, 14, 15 - - -, meter type, 14 - - -, volumetric dump type, 15 Baker’s correlation, 293 Bed’s correlation, 244-246 Begs-Brill, - - correlation, 286, 287 - - equation, 258 Begs-Robinson correlation, 244 Benzene series, 37 Bernouk’s equation, 26, 28 Blasius equation, 214 Blower, 12 Boiling loss, 150 Boiling point, - -, mean average, 48 - - of components of petroleum fluids, 43 Booster pump stations, 224, 228 - _ _ , design of, 228 Borden-Rzasa correlation, 661 Breathing loss, 149 Brown’s chart for specific heat of MC-vapors, 260 Brown-Katz-Oberfell-Alden correlation, 254 %S&W,116, 134 Bubble flow, 287, 288, 387 Bubble point, - - pressure correlation, 661 - - temperature of a mixture, 55 Bubble tray absorber, 194 Buildup (pressure), 327, 340, 353, 354, 355, 355 - analysis (example), 355 - following a long producing time, 356 -, idealized rate-pressure history for, 354
* Prepared by Mehmet P a r k and George V. Chilingarian.
806
- test, 327, 340, 353 - - in gas wells, 357 Buoyancy, 25, 146 - force on the rods, 577 Bureau of Mines Tester, 183 - _ _ _ for water content of natural gas, 183 Butane, 185 Cable, 748,149, 768, 769, 770, 773 - extensions (flat), 769, 770 - handling procedures, 773 -, motor flat, 749 -, power (ESP), 748 - selection, 767 - voltage drop losses, 768 Cadmium sulfide test, 183 Calibration, - of flowmeters, 385, 386, 389, 403, 408 - of spinner flowmeter (in-situ), 385, 386, 389 Caliper, -, through tubing, 378, 379 Capacity, -, gas, 67, 71, 88, 90, 95, 97, 106 -, liquid, 67, 71 - of Reda pumps, 763 -, pipeline flow, 224 -, valve flow, 165 -, venting, 171 -, water adsorption, 187 Carbon dioxide, 19, 179, 183, 188 - - content of natural gases, 179, 183 - - removal from natural gas, 19, 188 Carr- Kobayashi-Burrows correlation, 257259 Casing pump(s), 552 - -, advantages and disadvantages of, 553 - -, schematic diagram of, 553 - -, volumetric efficiencies of, 664 Cementing, 393 -, use of temperature survey in, 393, 394 Characterization factor, UOP, Watson, 21, 47 - -, definition of, 47 Chemelectric treaters, 120, 121, 136 - -, troubleshooting, 141 Chemical injection, - - for demulsification, 133, 137, 144 Chernikin’s model, - - for unsteady oil flow in buried pipelines, 250 Chew-Connally correlation, 245, 246 Choke(s), 294, 295, 323-325, 332, 452 -, adjustable, 323-325 -, fixed (positive), 323 -, flow through, 294 - for reducing kick-off pressure in gas lift, 452
-, functions of, 294, 323, 325 - in DST assembly, 332 -, orifice, 452 -, required size of, 294, 295 -, types of, 323 Christmas tree, 516, 469 - -, plunger lift, 516 - -, type M plunger lift, 469 Circle, some geometric properties of, 69 Claus plant, 188 Closed power fluid system, 671, 676, 683, 696 _ _ _ _ , power fluid tank in, 683 _ _ _ _ , pressure equation for, 696 - _ - -, surface facilities of, 683 Coalescence of droplets, 133 Coalescing force, - -, attractive, formula for, 116 Colebrook’s equation, 214 Compressibility, - factor ( Z ) ,44, 45, 254 - factor ( Z ) chart, 255 - of fluids (definition), 341 Compressible flow, 27, 29 -, formula for, 27, 29 Compressor, 12, 30, 154, 198 COMPSEL, 769, 788, 795 Condensate gas, 178 Conduction heat transfer, 231 -, steady-state, 231 -, unsteady-state, 231 Conductivity, -, thermal, 231, 234, 243, 247 Continuous gas lift, 432, 433, 434, 439, 441, 442, 444,446, 461 - _ _ ,dome (bellows) pressure determination in, 441,442 - - _, examples for designing, 444,446 - _ _ , flowing pressure gradient in, 434 - - - , gas output vs. production for, 433 _ _ _ ,graphical design, 439, 444,446 - - - , information needed to design, 439 - - - installation design, 434 _ _ _ , pressure gradient curves for, 433 _ _ _ , sample problem for, 461 Convection heat transfer, 232 -, forced, 233 -, natural, 232 Correlation, -, Baker’s, 293 -, Beal’s, 244-246 -, Beggs-Brill, 286-287 -, Begs-Robinson, 244 -, Borden-Rzasa, 661 -, Brown-Katz-Oberfell-Alden, 254
807
-, Cam-Kobayashi-Burrows, 257-259 -, chew-conndy, 245, 246 -, Cragoe’s, 247 -, Dukler’s, 291 -, Dukler-Wicks-Cleveland, 291 -, Duns-Ros, 287 -, Eaton et al., 291 -, Flanigm’s, 293 -, Gilbert’s, 294
-, Hagedorn-Brown, 286,289 -, Lockhart-MartineUi, 291 -, Standing’s, 284, 662, 663 -, Walther’s, 247 Corresponding states, - -, law of, 256 Corrosion, 622, 623, 625 - -erosion, 622 - fatigue, 623 -, galvanic, 623 - inhibitors, 625 -, pitting and concentrated cell, 622 -, stress, 622 - types encountered in subsurface pumps, 622 Counterbalance, - design, 578 - effect, 578, 579, 590, 608 CPF system, see Closed power fluid system Cragoe’s correlation, 247 Critical, - flow, 294 - pressure of components of petroleum fluids, 43 - temperature of components of petroleum fluids, 43 Crude oil enrichment, 190 Cycle-controlled expanding plunger, 468, 476 - - - -, well data for, 476-482 Cycle-controller, - -, electronic time, 475 - - for intermittent flowing, 516 - -, Taylor-type K, 474 - -, time, 475 Cycle-time, - -, minimum (in plunger lift), 487 Cyclone centrifugal separator, 691, 692 Dalton’s law, 85, 162 Darcy’s law - - for radial flow, 341 - - for radial steady-state flow, 346 Davenport-Conti model, 252 - - - for transient heat losses from a buried pipeline, 252 Dehydration> 4, 11, 113, 114, 115, 118, 121, 136, 186, 187
-, automated, 121 -, chemical, 11, 114 -, electrical, 11, 115, 118, 136 -, glycol, 186
-, gravity, 4, 113 -, sour gas, 187 -, thermal, 113 Dehydrators, 117-119, 120-123
-, chemelectric, 120-123 -, electric, 117-119 Deliverability, - of a gas field, 269-273 Demulsification, 133, 136, 137, 143, 144 -, agitation for, 133, 144 -, chemical, 132, 133, 137 -, chemical injection for, 133, 137 -, electrical treatment for, 136 -, heating for, 133, 143 -, operating procedures for, 133 -, troubleshooting in, 137 Demulsifiers, 114, 132 -, action of, 132 Densimeter, 376, 377 Density, -, average gas-liquid mixture, 287 -, average liquid-liquid mixture, 390 -, definition of, 212 -, effect of pressure and temperature on oil, 246 - ofcomponents of petroleum fluids (liquid), 43 - of gases, 44, 256 - of liquids, 47, 246 Desalting, 116 Desiccants, 186, 187 Design of, - - booster pump stations, 228 - - counterbalance (in sucker-rod pumping), 578 - - electric submergible pump system, 756-772, 785 - - flowing well systems, 279 - - gas-lift system, 434, 439, 444,446, 449 - - - - -, continuous, 434, 439, 444,446 - - - - -, intermittent, 449 - - pipeline systems, 222,224,225, 228,266,267 - - - _, gas, 266, 267 _ _ - _ , oil, 222, 224, 225, 228 - - production facilities for a well, 296, 297 - - separators, 67, 71 - - sucker-rod pumping system, 538, 539, 561, 568, 583, 597, 629 - - - - _-, API recommended procedure for, 583 - - - _ - -, theoretical analysis in, 560 - - vapor recovery systems, 158 - - vent valves, 162 - - vessels, 68
808 Destabilitization, emulsion, 114, 115 -, chemical, 114 -, electrical, 115 Dew point temperature of a mixture, 55 Differential, - double-acting pump, 656 - injection method, 383 - valves, 454, 455 - vaporization, 71 Diffusivi ty, - equation (radial form), 342 - -, line source solution of, 343 -, hydraulic, 343 -, thermal, 232, 234 Dimensionless, - numbers, see Number - pressure, 344 - radius, 343 - rime, 343 Diolefins, 35 Diphasic flow, 386, 387 - -, gas-liquid, 387 - -. in situ calibration of spinner flowmeter in, 389 - -, liquid-liquid, 387 - -. spinner response in, 389 Directional wells, multiphase flow in, 290 Distillation, - area flow diagram in natural gasoline plant, 208 -, Engler, 185 Dome pressure, determination in continuous gas-lift, 441, 442, 444,447 Double-acting pump, 656 - _ _ , differential, 656 Dougherty’s method, for deliverability of a gas field, 269, 273 Drag, 146 - coefficient for a sphere, 146 - force for VISCOUS flow, 146 Drainage radius, 346 Drawdown (pressure), 327, 340, 347, 351, 357 - anaiysis (example), 351 - test, 327, 340, 347 - - in gas wells, 357 - -, multiple rate, 350 Dnllstem test (DST), 327, 328, 329, 333, 334, 33.5-340 - - charts showing various problems and well conditions, 335-340 - -. component parts of a cca-.,entional tester, 329 - -, objectives of, 327 -- .-. procedure for, 333
- -, qualitative interpretation of, 336 - - tool assembly, 328 - -, typical chart, 334 Drizo process, 187 Dry, - box method, 188 - gas, 178 Dukler-Wicks-Cleveland correlation, 291 Duns-Ros correlation, 287 Dynamometer card(s), 593-595, 599, 600-611 - -, calculation of counterbalance effect using, 608, 610 - -, - - instantaneous torque using, 609 - -, - - polished rod horsepower using, 608 - -, - - pumping efficiency using, 611 - -, definition of, 593 - -, factors affecting the shape of, 599 - - for an ideal pumping system, 595, 599 - -, information obtained from, 593-595. - -, operating problems diagnosed from, 599, 611 - - showing areas and deflections needed for load calculations, 608 - - showing various well conditions and problems, 601-607 - -, typical, 599, 600 Eaton et al., correlation, 291 Edmister’s chart, for heat capacity difference vs. reduced properties, 261 Eductor tube, 425, 426 - -, average fluid density in, 426 - -, fluid velocity in, 425 Effective, - diameter, 212 - plunger stroke, 570, 584 Ei-function, 344 Ejector, -, first US patent issued for an oil, 427 -, oil, 427 Elasticity modulus, - -, steel, 570 - -, various plunger and rod sizes, 562-567, 572 - -, - tubing sizes, 571 Electric submergible pumps, - - _, applicability of, 737, 738 - - _, disadvantages of, 738 - - - , factors affecting the periormance of, 769 - - _ , multistage centrifugal type, 742, 743 - - -, performance curves for, 764, 794 - - -, selection of, 761, 763, 785 Electric submergible pump system, 738, 739, 740, 741, 742, 743, 748, 749, 750, 751, 752, 753, 772, 773, 775. 785
809
_ _ _ _ , components of, 740 _ _ _ _ , design of, 756, 785
_ _ - _ , handling of, 772 - - _ -, installation of, 773 - - - -,junction box in, 753
_ _ _ _ , motor in, 740, 741 - _ _ _ , motor flat cable in, 749 - _ _ _ , power cable in, 748 - _ _ -, pressure sensing instrument (PSI) in, 753
_ - _ _ , protector in, 743 _ - - _ , pump intake in, 743
____
selection data, 756
- _ - -, switchboard in, 750 _ _ _ _ , transformer in, 751, 752 _ _ - - , troubleshooting, 775 _ _ - _ , typical installation for, 738, 739 - _ _ -, variable speed drive in, 754 - - _-, wellhead for, 752 Elevation factor, 293 Elongation, -, net, 572 - of rods, 570 - of tubing, 571 Emulsifiers, role of, 126 Emulsion(s), 109, 110, 112, 125, 126, 128, 131, 136, 137 -, aging of, 112, 128 -, basic components of, 126 -, chemical resolution of, oil-in-water, 136 - - _ , _petroleum, 125 - - _ -, water-in-oil, 131 -, conditions for stable, 126 -, definition of, 125 - destabilizers, 114 -, electrical resolution of oilfield, 109 -, factors affecting the stability of, 126 - flow, 387 -, loose, 128 -, oil-in-water (inverse), 110, 125, 136 -, reduced temperature treating of, 130 - stabilizers, 112 -, theory of, 110 -, tight, 128 -, troubleshooting in treatment of, 137 -, water-in-oil, 110, 125, 126, 131 Energy, - consumption in U.S.A., 177, 179 - equation, 25 -, free, 110, 111 - loss between the p u n p and polished rcd, 582 - losses in gas lift, 421 - optimization in sucker-rod pumping, 611, 619 Engler distillation, 185
E.P.A., 153 Equation, -, Begs-Brill, 258 -, Bernoulli, 26, 28 -, Blasius, 214 -, Colebrook, 214 -, Cragoe, 247 -, DeGman-Andrade, 244 -, Fanning, 213 -, Hagen-Poiseuille (Reynolds number), 213 -, Hansen, 233 -, Heinze, 266 -, Kato-Nishiwaki-Hirata, 233 -, M a k ~ w ~ k i - M ~ ~ h l i n s234 ki, -, Prandtl-Karman, 214 -, Rohsenow-Hartnett, 242 -, Sieder-Tate, 233 -, Souders-Brown, 67-71 -, S U P ~ O214 , -, Vogel, 282 -, Weymouth, 262, 265, 268 Equilibrium, - flash calculations, 60, 77, 78 - ratios, vapor-liquid, 52 - -, generalized correlation for, 53, 54 - relations for complex mixtures, 60 Equivalent, - diameter for pipelines in parallel and in series, 267, 268 - length, formula, 219 - _ _for pipelines in parallel and in series, 267, 268 - producing time, 353, 356 Ethane, 185 Ethylene recovery plant, -__ , description of, 204 - _ -, sample problem, 204 Evaporation, - control, 150 - loss, 149 Exponential integral, - -, formula, 344 - - solution, 344, 345 F a g ' s method, for calculating torque from dynagraphs, 609 Fall-off tests, 368 Fanning, - equation, 213 - friction factor, 213 Ferrox process, 188 Filling loss, 150 Fianigan's, - correlation, 293
810
- method, 292 Flash, -, adiabatic, 60 - equilibrium calculations, 60, 77, 78 - -, three-phase, 83 -, isothermal, 60, 77 - vaporization, 60, 71 Flocculation of droplets, 132 Flotation cell, 13, 14 Flow-after-flow tests, 359, 360 - - - -, US. Bureau of Mines procedure for, 361 Flow in pipes, 211, 224, 230, 237, 261, 266, 267, 286-288, 291, 386-389 - - _, bubble, 287, 288, 387 - - - , capacity, 224 - - -, critical, 294 - - -, diphasic, 386, 387, 389 - - -, efficiency, 284 - - -, emulsion, 387 - - -, froth, 288 - - -, gas, 261, 266, 267 - - -, mean pressure evaluation for -, 263 - - -, mist, 287, 288, 387 - - -, monophasic, 384 - - -, multiphase, 286, 288, 291, 386-389 - - -, -, horizontal, 291 - - - , -, vertical, 286, 288 - - _, oil, 211, 230, 237 - - _, nonisothermal -, 230 - - -, steady-state -, 237 - - -, polyphasic, 386, 388 - - -, regime map, 288 - - -, slug, 287, 288, 387 - - -, triphasic, 386, 388 Flow in reservoir, 280 - - -, pseudo-steady state, 344 - - -, pressure distribution in - - -, 346 - - -, steady-state radial, 346 Flowline treatment system, 129 Flowmeter, -, calibration of, 385, 386, 389, 403, 407 -, continuous, 377, 378 -, fullbore spinner, 377, 379 -, inflatable packer, 376 - logs, field applications, 401, 402, 404, 405 - -, interpretation, 384, 401 -, surveys, field examples, 402, 404-406, 408, 409, 411 Fluid level, - -, static (definition), 429 - -, working (definition), 429 Fluid load on polished rod, 517 Fluid sampler, in DST tool assembly, 333
Fluid statics, fundamental equation of, 23 Fluid weight conversion table, 417 Footpiece (removable), 474 Formation shrinkage factor, 664, 665 Formation volume factor, - _ -, Standing’s correlation for, oil, 662 - - - - - -, total, 663 Fourier’s, - law for heat conduction, 231 - number, 250 Fractionation, 197, 198 Fracturing, 395 -, temperature anomalies after, 395 Free-cycling plunger, 468 Free energy, 110, 111 Free water knockout, 134, 139 Frequency factor, - - for various combinations of rod strings, 563-567 - -, percentage increase in, 592-596 Friction, 213 - factor@), 213 - -, Dukler’s correlation for, 291 - -, Fanning’s, 213 - -, for various commercial pipes, 215 - loss in tubing (chart), 762 -, pump (hydraulic), 697, 699 Froth flow, 288 Gas (see also Natural gas) - column, pressure of, 416, 442 - -, average temperature in, 420 - -, weight of, 442 - compressibility, 44, 45, 254-256 - deliverability, 269, 273 - density, 44, 256 - expansion, 420 - -, adiabatic, 421, 461 - -, isothermal, 420, 461 - -, polytropic, 421 - flow in pipes, 261, 263, 266, 261, 269 - - rate (Weymouth equation), 263 - gathering, 198 - gravity, 184, 258 - hydrates, 199 - injection, 200 - measurement, 184 -, natural, see Natural gas -, physical properties of, 254 - processing and conditioning, 18, 186 - - - -, reasons for, 18 - - - -, plants, 181 -, sales specifications, 182 - separator intake, 743
811
- - - , purpose of, 743 - - -, static type, 746, 748 - -, rotary, 747, 748 test methods, 183 transmission systems, 266 treating, 18, 181, 186, 187 - -velocity number, 289 - vent lines, pressure drop through, 674,675 - viscosity, 257, 258 - well testing, 357 - - -, buildup, 357 - - -, drawdown, 357 - - -, flow-after-flow,359,360 - - -, isochronal, 359, 362 - - -, modified isochronal, 359, 364 - wells, plunger equipment for, 469, 471 Gas lift, - -, advantages of, 459 - -, continuous, 432,433,439,444,461 - -, design of -, 434 - -, disadvantages of, 432 - -, efficiency of, 421 - -, energy utilized in, 420 - - equipment, 452 - -, fundamentals of, 416 - -, gas output vs. production in, 425, 433 - -, gas volume, necessary for, 421,424 - -, - - vs. bottomhole tubing pressure in, 424 - -, history Of, 415,426 - -, horsepower requirement for gas compression in, 458 - -, installation, rules of thumb for, 459 - -, intermittent, 428 - -, paraffin accumulation in wells on, 426 - -, plunger equipment for wells on, 469, 470 - -, principles and methods of, 426 - -, straight, 432 - - terminology, 429 - -, types of, 431 - -, U-tube type of, 426,427,432 - -, unloading operations in, 436,437 - - valves, 428,429, 450, 453-457 - - -, spacing, 435,438,441 ’ - - -, specifications (Maw), 443 Gas-liquid-ratio gradient, - - - - in plunger lift, 484 Gas/liquid ratio vs. - - - - depth, 499, 502 - - - - minimum tubing pressure, 500, 503 - - - - net operating pressure, 513, 514 Gasoline (natural), 181, 184, 185, 189,206 Gathering system, 1, 2 - -, gas, 198,199 Geothermal gradient, 396, 397,420 -
Gilbert’s correlation, 294 Glycol absorption-dehydration process, 19, 20, 186 Gradiomanometer, 373-375, 388 - surveys, interpretation of, 390 - -, field examples, 401, 402, 404,406,408-411 Graetz number, 233 Grashof number, 232 Guiberson, - Powerlift pumps, 642,650,651,656,660 - - -, nominal sizes of, 660 - - -, specifications for, 650,651 - wellsite power plants, 688,690,691 Gun barrels, 135 - -, troubleshooting, 140 Hagedom- Brown, - - correlations, 286,289 - - pressure traverse curves, 300 Hagen-Poiseuille equation, 213 Hansen’s equation, 233 Heat, - conduction, 231 - convection, 232 -, cragoe’s correlation for oil specific heat, 247 - exchangers, 21,22, 140 -, temperature vs. hydrocarbon vapor specific heat, 260 - transfer, coefficients for pipes, 242, 248 - - concepts for buried pipelines, 234 - - fundamentals, 230 Heater treaters, 6-11, 136,140 - -, horizontal, 6,7, 136 - -, troubleshooting, 140 - -, vertical, 10, 11, 136 Heating value of natural gases, 184 Heinze equation, 266 Henry’s law, 61 Holdup factor, - -, corrected, 290 - -, liquid, 286 - -, no-slip, 286 - -, water, 375, 388, 389 Hollow rods, 641 Hooke’s law, 570 Horizontal two-phase flow in pipes, 291 - - - - - -, correlations for, 291 - - - - - -, pressure traverse curves for, 317322 - - _ _ _ _ , technique of using - - - -, 291 Homer plot, 284,353, 356 Horsepower, 31,213, 221 -, brake, 221, 581 -, frictional, 619
812
-, hydraulic, 221,581,619
-
requirement for gas compression in gas lift, 458 - - -, motor in ESP system, 765 - - -, prime mover in sucker-rod pumping system, 581 -, total polished rod, 581,589, 590, 608 Huff-and-puff method, 400 Hydrate point, - - of hydrocarbon gases, 265 - - of nitrogen containing natural gases, 266 Hydrates, -, formation of gas, 199, 265, 266 -, conditions for - - -, 265, 266 -, preventing - -, 199 - formation temperature, 266 Hydraulic, - diffusivity, 343 - gradient, 222 - jars, 333 - radius, 212 Hydraulic pumping system, - - -, calculation of bottomhole pressure in, 720 - - -, - - surface horsepower in, 669, 693,705, 708 - - -, control manifolds in, 680 - - _, power fluid cleaning system in, 681 - - - , power plants in, 687-691 - - -, schematic diagram of a complete, 637 - - -, surface pumps in, 681 - - -, tubing arrangements in, 670 - - -, types of power fluid systems in, 682 Hydraulic subsurface pumps, 635 - - -, Guiherson Powerlift, 642,650,651, 656 - - -, Kobe, 638-649, 695 - - -, Oilmaster, 652,655-657 - - -, operation of parallel free, 636 - - -, piston-type, 635,638 - - -, pressure and friction losses affecting, 694, 695 - - -, selection of, 660,667,669,708 - - specifications for, 643-655, 657 - - -, wellheads for, 680 Hydrocarbon content of natural gases, 182, 183 Hydrocarbons, -, aromatic, 37 - in natural gas, 180 -, naphthene, 35 -, rules of nomenclature for, 39 -, saturated, 32-34 -, unsaturated, 32, 35, 36 Hydrogen embrittlement, 622,623 Hydrogen sulfide,
-
-.
- - content of natural gases, 179, 182, 183 - -, removal of, 19,188 Ideal gas equation, 44 Inert gases, 178 Inflatable combination tool (ICT), 376, 383 Inflow performance relationship (IPR),281 - - - for solution-gas-drive reservoirs, 282,283, 285,665,666,759 - - -, Vogel's reference curve for, 665,666,759 Inhibitor, -, corrosion, 625 - injection methods, 625 Injection, -, natural gas, 200 - of fluids, 397 -, reasons for natural gas, 200 - wells, temperature survey in, 397, 398 - -, testing, 368 Interface stabilization, 112 Interfacial tension, 112 Interference tests, 365, 366, 368 - - for permeability anisotropy, 366 - -, nomenclature for, 368 - -, vertical, 368 Intermittent flowing and plunger system, 515 ---__ , advantages of, 516 Intermittent gas-lift, 428,449 - - - cycle of operation for tubing operated valves, 450 - - - design, 449 - - - valves, 428 IPR curves, see Inflow performance relationship Iron oxide method, 188 Isochronal test, - -, modified, 359,363,364 - -, normal, 359, 362 - -, Brown-Begs procedure for -, 363 Jet collar, 452 Joule-Thomson effect, 192 Junction box (in ESP), 753 Kato-Nishiwaki-Hirata equation, 233 Kick-off pressure, 422, 428,432 - -, derivation of formula for, 423 - - for gas-lift through casing, 424. - - - - - - tubing, 423 - -, types of chokes to reduce initial, 452 Kick-off valves, 428,453 Kilowatt-hour meters, 621 Kobe, - pumps, 638-642 - -, nominal sizes of, 660
813
- -, pressures acting on type-A, 695 - -, schematic diagram showing operation
of
type-A, 658
- -, specifications for, 643-649 - wellsite power plants, 687, 689, 690 KVA requirement for transformer bank, 767 Laminar flow, 211
- -, criteria for, 212 - -, Fanning friction factor for, 213 Last-stroke method, 720 Law, -, Dalton’s, 85 -, Darcy’s, 341 -, Fourier’s, 231 -, Henry’s, 61, 85 -, Hooke’s, 570 - of corresponding states, 256 -, Raoult’s, 85, 162 -* Stokes’, 113, 145, 146 Least-squares equations for plunger lift, 476, 483 Lift, 429 -, air, 415 -, gas, see Gas-lift -, net, 582, 667 -, plunger, see Plunger lift -, total, 431 -, working, 430 Line source solution, 343 Liquid extraction from natural gas by, 189 - - - _ _ _adsorption, 191 - - - _ _ _ crude enrichment, 190 - _ - _ _ _ refrigeration, 191 LNG, 181 Lockhart method, 81 Lockhart-Martinelli correlation, 291 Lockhart-McHenry method, 78 Logging (production)devices, 373 Logs, -, flowmeter, 384 -, production, 373 -, field examples of -, 401 -, troubleshooting using -, 393, 394, 401, 402, 408, 411 -, temperature, see Temperature survey -, thermal decay time (TDT), 401, 404 Lost circulation, 393 - - zones, detection by temperature survey, 394 LPG, 181, 184 Makowski-Mochlinski equation, 234 Manometer, 380 -, gradio-, 373-375, 388, 399 - surveys, field examples, 401, 402 Maximum,
-
production rate in plunger lift, 484 reliable flow, 28 Mercaptans, 178, 183 Mills acceleration factor, 570 Mist, - extraction, 17, 18, 64 - flow, 287, 288, 387 Molecular weight of, - _ -components of petroleum fluids, 43 - - - gas mixtures, 256 Monel, 624 Monophasic flow, 384 - _, correction factor vs. Reynolds number in, 384 - _, spinner in, 384 Moody’s friction factor chart, 215 Multiphase flow in, - _ -directional wells, 290 - _ _ horizontal flowlines, 291 - _ -pipelines on hilly terrains, 292 - - - vertical wells, 286 Multiphase well test analysis, 369 Multiwell testing, 365 - -, interference, 365 - -, pulse, 367 Naphthene hydrocarbons, 35, 36 Natural gas, 177 - -, acid gas content of, 179, 182, 183 _ _ , air content of, 179, 183 - -, carbon dioxide content of, 179, 183 - -, composition of, 178, 179 - -, compressibility factor for, 256 - -, density of, 256 - - engineering department, functions of, 201 _ - , heating value of, 184 - -, hydrocarbon content of, 180, 182, 183 - - injection, 200 - -, liquefied, 181 - -, liquid extraction from, 189 - - liquids, 184 - - _ , specifications for, 184 - - -, testing, 185 - - -, value of, 189 - -, physical properties of, 254 - - processing plants, 181 - -, removal of impurities from, 185 - - sales specifications, 182 - -, specific gravity of, 184, 258 - - test methods, 183 - -, treating, 185 - -, viscosity of, 258 Natural gasoline, 181, 184, 185 - -, composition of, 181 - - plant, 206
.
814
- -, sales specifications for, 185 - -, value of, 189 Net lift, 582, 667 Net operating pressure vs., _ _ - _ depbh, 501, 504 - _ _ _ GLR, 513, 514 Net plunger stroke, 570 Ney’s method, for multiphase flow in directional wells, 290 Nitrogen removal from natural gas, 189 Nomographs (in plunger lift), -, constructing, 488 - for 2-in. plunger, 491, 499-501, 513 - for 2-1/2-in. plunger, 492, 502-504, 514 -, how to use, 496, 497, 505 -, information needed to estimate plunger lift performance in a well from, 497 -, mathematical operations by, 490 -, need for, 488 Number, -, Fourier, 250 -, gas-velocity, 289 -, Graetz, 233 -, Grashof, 232 -, liquid-velocity, 289 -, Nusselt, 232, 241 -, Prandtl, 233 -, Reynolds, 211 Oilmaster, - pumps, 656 - -, nominal sizes of, 660 - -, specifications for, 652-655, 657 - wellsite power plants, 689, 691 Olefins, 35 Open power fluid (OPF) system, 670-673 _ _ _ _ , general pressure equation for, 696 _ _ _ _ , power fluid tank in, 685, 686 _ - _ _ , surface facilities of, 684, 685 Open flow potential, 362, 403, 406, 407 _ _ _ from flow-after-flow test, 362 Operating line, 489, 490, 493, 520-527 Optimum pressures for two- and three-stage separations, 73 Orifice meter, 3, 184 - -, derivation of formula for flow through, 26 Orsat analysis, 183
P over E, 667, 668, 696 Packer, - assembly, 330 -, inflatable, 376 Paraffin accumulation, - -, fighting, 426
- - in wells operated by gas lift, 426 Paraffins, 33-36, 178 -, cyclic, 35, 36 -, iso-, 33, 34 Partial pressure, 163 - -, Dalton’s law of, 162 Permeability, - calculation from buildup test, 355 _ _ _ drawdown test, 350, 352 - - _multiple-rate drawdown test, 351 -, vertical interference and pulse test for vertical, 368 Petroleum, -, classification of, 38 -, mixed base, 39 -, naphthenic or asphaltic base, 39 -, paraffin base, 39 Physical properties of hydrocarbons, - - - -, fluids, 43 _ _ _ -, gases, 254 _ - - _ , interrelationship among various, 56 Pipes, fluid flow in, 211 PipelinHs), -, advantages of, 211 - design, 222, 225, 228, 266-269 -, exit temperature in buried, 238 - flow capacity, increasing, 224 -, gas, in parallel, 266 -, _, in series, 267 -, - transportation in, 266-269 -, heat transfer concepts for buried, 234 - insulators, thermal conductivities of, 243 -, nonisothermal flow in, 230 -, oil, branching, 228 -, -, looped, 225 -, - transportation in, 221, 225, 228, 249, 252 - pressure profile over a hilly terrain, 223, 227 -, transient flow of oil in buried, 249, 252 -, transient time for heat losses in buried, 252 Plant, -, ethylene recovery, 204 -, natural gasoline, 206 -, wellsite power, 687-690 Plunger(s), -, brush-seal, 473 -, cycle-controlled expanding, 468, 476 - _ _ _ , well data for, 476-482 -, effects of various well conditions on operation of gas-lift, 495 - equipment for, gas-lift wells, 469, 470 - - - gas wells, 471 - - _high GOR wells, 472 -, extended, 474 -, free-cycling, 468
815
-, GLR vs. depth for, 2-in., 499 - - - - - 2-1/2411., 502 -, GLR vs. minimum tubing pressure for 2-in., 500 - - - - - - -, 2-1/2-in., 503 -, GLR vs. net operating pressure for 2-in., 513 - ----_, 2-1/2-in., 514 - load (gross), 584 -, net operating pressure vs. depth for 2-in., 501 - - - - - - - 2-1/2411., 504 -, nomograph for 2-in., 491 -, - - 2-1/2-in., 492 -, operating line (derivation) for 2-in., 489 - - _ - -2-1/2-in., 490 - overtravel, 572 -, retractable segmented, 470, 473, 474 - stroke, effective, 570, 584 - -, factor, 585 -, supplementary operating line (derivation) for, 2-in., 493 -, - - - - - 2-1/2411., 494 -, tandem, 474 -, teflon-seal, 473 -, turbulent-seal, 473 Plunger lift, - - application curves based on gas and pressure requirements, 513, 514 - - applicability, 510, 512, 515 - -, average pressure at point where gas enters tubing in, 488 - - gas well applications, 517 - -, history of, 468 - -, how to determine candidate wells for, 512 - -, least squares equations for, 476, 483, 484 - -, accuracy of (in performance prediction) - - - -, 507-509 - -, examples of using - - - -, 510 - -, methods of obtaining - - - -, 485-487 - -, nomenclature for equations used in, 527 - - nomographs, see Nomographs - -, operation of, 470, 473, 476 - -, predicting performance of, 475, 517 - - system, intermittent flowing and, 515, 516 - - -, Christmas tree in, 469, 516 - -, types of wells suitable for, 510 Polar molecules, 112 Polished rod, 532-534 - - horsepower, 589, 590, 608 - - load calculations, 576 --__ , minimum load, 578, 586, 587 - - - -, peak load, 577, 584, 586, 588, 589 Polyphasic flow, 386 - -, flow parameters in, 388 - -, flow patterns in, 387
Potentiometer, 380 Pour point, 253 Power, - cable, 748, 749 - consumption in sucker-rod pumping, 611,619, 620 - - - - - _, of prime mover, 611 ----__ , testing, 621 - plants in hydraulic pumping, 687-691 Power fluid (in hydraulic pumping), - - flow rate, 669, 693 - -, selection of, 682 - - surface pressure, 666, 693 - - systems, 682 - - _, closed (CPF),671, 676, 683 - - -, open (OPF), 670-673, 684-686 - - tank, 683, 685, 686 Powerlift (hydraulic) pumps, 642, 656 - - -, nominal sues of, 660 _ - -, specifications for, 650, 651 Prandtl number, 233 Prandtl-Karman equation, 214 Pressure, - at point where gas enters tubing in plunger lift, 488 -, allowable working (of pipeline), 212 -, bottomhole, -, -, calculation from productivity index, 518 -, -, in gas-lift, 434, 518 -, -, in hydraulic pumping, 720 - buildup in plunger lift, 484 - buildup test, see Buildup -, dimensionless, 344 - drawdown test, see Drawdown -, gas column, 416 -, kick-off, 422, 428 -, last-stroke, 721 -, net-operating (in plunger lift), 484 -, partial, 162, 163 -, pseudo-, 357 -, pseudo-critical, 46, 254, 256 - ratio ( P / E ) , 667, 668, 696 - recorders, 329 -, reduced, 45, 46, 256 -, storage, 161, 163, 165 Pressure drop in, - - - API tubings, 705-711, 736 - - - gas-vent lines, 674, 675 - - - standard pipes, 702-704, 736 - - - tank roof fittings and pipe bends, 160 - - - tubing-casing annulus, 720-736 - - - tubing-pipe annulus, 712-713, 736 - - - tubing-tubing annulus, 714-719, 736 Pressure gradient, 416
816
- - curves for continuous gas-lift, 433-435 - - vs. % salt water in water-oil mixtures, 419 Problem wells, API recommended procedures for
-, electric submergible, see Electric submergible
analysis of, 611-618 Production combination tool (PCT),383 - - -, completion evaluation using TDT and, 401, 404, 406 _ - - , components of, 383 - - - survey in a gas well (field example), 401 Production logging, 373 - - devices, 373 - -, field examples, 401 - -, purposes of, 373 - -, troubleshooting using, 393, 394, 401, 402, 408, 411 Production system, overall, 279, 296 Productivity index, - - change with depth, 518 - - consistency with operating conditions, 518 - -, definition of, 280 - -, reference depth for, 517 - -, specific, 280 - -, use of, for derivation of operating line in plunger lift, 489, 490, 493, 494 _ - vs. flow rate for various drive mechanisms, 281 Propane, 185 - recovery vs. lean oil/gas ratio, 195 _ _ _ butane and pentane recovery, 196 Protector (in ESP), -, functions of, 743 -. labyrinth-path, 745 -, positive-seal, 744 -, selection of, 769 Pseudo-, - critical pressure and temperature, 46, 254, 256 - pressure, 357 - ieduced pressure and temperature, 45, 46, 256 - steady state flow, 344 - - - _, dimensionless pressure during, 346 - - - _, pressure distribution during, 346 PSI unit (in ESP), 753, 754 Pulse testing, 367 - -, vertical, 368 Pump(s), -, API classification of subsurface, 537 -, API designation for subsurface, 538 - booster stations, 228 -, casing, 552 - constants for various plunger sizes, 548, 574 - displacement, hydraulic, 660, 665, 668 - -, sucker-rod, 574, 584 -, double-acting, 656 - efficiency, 221. 575
-, Guiberson Powerlift, see Guiberson powerlift pumps -, hydraulic subsurface, see Hydraulic subsurface pumps - intake (in ESP), - -, gas separator, 743, 746, 747 - -, standard, 742 -, Kobe, see Kobe pumps -, Oilmaster, see Oilmaster pumps -, Reda, see Reda pumps -, rod, 537, 555-559 -, sucker-rod, see Sucker-rod pumps -, tubing, 537, 554 Pump-off control, 621 Pumping, - efficiency calculation using dynagraphs, 611 - mechanisms, 219 -, principles of, 219 - speed, dimensionless, 591 - -, production vs., 540-547 - system, conditions for idealized, 595 - -, schematic diagram of a complete hydraulic, 637 - unit (sucker-rod), - - , air-balanced, 535 - - design calculations, 628-629 - -, Mark 11, 535 - -, nomenclature for conventional, 536
pumps
- friction, 697, 699
Radioactive tracer survey, 380, 331 - - -, differential injection method, 383 - - -, timed runs method, 381, 382 - - -, velocity-shot method, 380, 381 Ranarex, 184 Raoult’s law, 85 Real gas equation, 44 Reda, - cable types and applications, 749 - motors, 740, 766 - pumps, 737 - -, request for information form for, 757 - _ , typical performance curve for, 794 Reduced pressure and temperature, 46, 256 Refrigeration process, 191 - -, For liquid extraction from natural gas, 191 - -, turboexpander, 192 - -, vapor rectification, 193 Regeneration, 187 Regulators, 154, 155, 157 -, typical installation of pressure and vacuum, I55
817
-, vapor control, 154 Reid vapor pressure, 181, 185 Removal of, - - acid gases from natural gas, 188 - - carbon dioxide - - -, 19,188 - - hydrogen sulfide - - -, 19,188 - - impurities - - -, 19,185 - - liquids - - -, 189 - - nitrogen - - -, 189 - - water - - -, 19,185 - - paraffin deposits from gas-lift wells, 426 - - solids from oil-treating system, 145 - - solids from power fluid, 691,692 Resistance coefficients, - -, definition of, 219 - - for bends and sudden reducers, 218 - - for valves and fittings, 216, 217 Resistance of valves and fittings to fluid flow, 220 Retrograde condensation, 178 Reynolds number, 211 - - for two-phase flow, 287 Rod(s), -, hollow, 641 - motion analysis, 569 -, polished, 532-534, 576 - pumpwith, - - - stationary barrel and bottom hold-down, 556,558 - - - stationary barrel and top and bottom hold-down, 559 - - - stationary barrel and top hold-down, 559 - - - travelling barrel, 555, 556 - weight, calculation of, 573,577 - - for various combination strings, 563-567 - - in fluid, 585 Rohsenow-Hartnett equation, 242 Roughness, relative, 213 Safety joint (in DST), 332 Sales gas specifications, 182 Saturated hydrocarbons, 32-34 Separation, -, factors affecting, 66 - of oil and gas, 59 - of immiscible liquids, 145,146 -, stage, 71 Separator@), -, comparison of different types of, 62 -, cyclone centrifugal, 691,692 - design, 67,71 -, functions of, 18,60 -, - - wellstream, 60 -, horizontal, 15-17, 94-103
-, internal parts of, 63 -, spherical, 16,18,59, 104-107 -, types of, 61 -, vertical, 16,17,87-93 Settling, 134 - tank, 135 - tank in OPF system, 686 Shape factor, 346, 348, 349 Shrinkage factor, formation, 664,665 Sieder-Tate equation, 233 Skin, - effect, 284 - - due to non-Darcy flow,358, 359 - factor from pressure, - - - - buildup test, 353, 355 _ - - _ drawdown test, 350 - - - - - -, in gas-wells, 358, 359 - - - - - -, multiple-rate, 351 Slippage velocity, 375,388
- - vs.
density difference at various holdups, 389 Slug flow, 287,288,387 Sonic velocity, 29, 30 Souders-Brown equation, 67,71 Sour gas, 179 - - dehydration, 187 Spacing, - between gas-lift valves, 435,438,441 - equations for tubing pressure operated valves, 438 -, graphical determination of continuous gas-lift valve, 441 Specific, - gravity, - - of natural gases, 184 - - of petroleum liquids, 47,48 - - vs. API gravity, 416-418, 700-701 - - vs. pressure gradient, 700-701 - heat, - - capacity difference vs. reduced pressure, 261 - - of hydrocarbon vapors, 259,260 - - of petroleum oils, 247 - productivity index, 280 - weight, 212, 417,418 Specifications for, - - gas-lift valves, Macco, 443 - - electric submergible pumps, Reda, 763 - - hydraulic pumps, 643-655 - - - -, Guiberson Powerlift, 650-651 - _ - -, Kobe, 643-649 - - - -, Oilmaster, 652-655 - - sales gas, 182 - - separators, 88,91-93, 95,98-103, 107 - - -, horizontal high-pressure, 98-103
818
- - -, horizontal low-pressure, 95 - - -, spherical, 107 - - -, vertical high-pressure, 91-93 - - -, verticd low-pressure, 88 Spinner, -, full-bore flowmeter, 379 - in monophasic flow, 384 -, in situ calibration of, 385, 386 - response characteristics, 385 - response in diphasic flow, 389 - type velocimeter, 376, 377 Stability, - condition for undisturbed drops, 116 - of emulsions, 126 Stabilization time, 346 Stage separation, 71 - -, three-, 72,73 - -, two-, 72, 73 Standing’s correlation for, - - - oil formation volume factor, 662 - - - solution-gas-drive reservoirs, 285 - - - total formation volume factors, 663 Standing-Katz compressibility chart, 255 Static, - fluid level, 429 - head, 429 - submergence, 429 Steady-state flow, radial, 346 Steam injection, - -, huff-and-puff, 400 - - system, 3 Stokes’ law, 113, 145 - -, derivation of, 146 Storage-shipping section, 2 Storage tank, 11 - - pressure, 161, 163, 165 - - temperature, 163 Stretford process, 188 Stripper wells, 531 Submergence, -, percentage working, 431 -, static, 429 -, working, 430 Subsurface pumps, - -, API classification of, 537 - -, API designation for, 538 - -, common materials used in manufacture of, 624 - -, common types of corrosion encountered in, 622 - -, electric, see Electric submergible pumps - -, hydraulic, see Hydraulic pumps - -, installation and operation of, 626 - -, types of, 551-560
Sucker-rod pumps, - - -, corrosion types encountered in, 622 - - -, material selection, 622 - - -, minimum tubing sizes for standard types of, 549 - - -, production vs. SPM charts for various bore sizes and net plunger travels for, 540-547 - - -, selection of, 539, 551 - - -, selection of setting depth for, 551 - - - size determination, 574 - - -, types o f , 551-560 Sucker-rod pumping, 531 - - _design, 538, 539,560, 561,568,583, 597 - - - -, API recommended procedure for, 583 - - - - , minimum information needed for, 538, 539, 583 - - -, energy optimization in, 611,619 - - -, installation and operation of equipment in, 626 - - -, limitations in using, 531 - - -, nomenclature for, 536 - - -, power consumption testing in, 619 - - -, selection of materials in, 622 - - - system, 531-536 - - - -, surface and subsurface efficiencies of, 619 - - - unit, 535 Suction head, 221 Superficial, - gas velocity, 289, 293 - liquid velocity, 289 Superposition principle, 353 Supersonic velocity, 29, 30 Supino equation, 214 Surface, - production equipment, 2,20 - pumps in hydraulic pumping system, 681 - tension, 111 - - of an emulsion system, 127 Switchboard (ESP), 750 -, electromechanical, 751 - selection, 767 -, solid-state, 751 Szilas’ model, 249 Tailpipe effect on operation of gas-lift plunger, 495 Temperature, -, factors affecting storage tank, 163 -, hydrate formation, 266 - in gas column, average, 420 -, maximum and minimum liquid-surface, 163, 164
819
-, maximum and minimum vapor space, 163, 164 -, pseudo-critical, 46, 254, 256 -, reduced, 45, 46, 256 Temperature survey (logging), 391 - - after fracturing, 395 - -, applications of, 393 - - during steam soak and production periods, 400 - - field examples, 401, 402, 406, 409-411 - - in injection wells, gas vs. water, 397 - - in production wells, gas vs. oil, 396 - - showing cement level during cementing, 394 - - - effect of fluid loss through lubricator, 391 _ - _ importance of reaching equilibrium with surroundings, 392 - _ -leak around casing shoe, 399 - - _zones of lost circulation, 394 - - under dynamic vs. static conditions, 393 _ - -static conditions, production well vs. injection well, 398 Thermal, - conductivity, 231 - - of petroleum oils, 247 - - of pipeline insulators, 243 - - of soil, 234 - diffusivity, 232 - - of soil, 234 - expansion coefficient, 233 Thermal decay time (TDT) surveys, 401 - - _ _ for completion evaluation, 401, 404 Thermometer, high resolution, 373, 374 Time lag (in pulse testing), 367 Timed-runs method, 381, 382 Tool, -, inflatable combination, 376, 383 -, production combination, 383 Torque, - calculation using dynamometer cards, 609 -, definition of, 580 - on gear reducer, 580 -, peak, 581, 586 Total dynamic head (TDH), 761 Tracer surveys, radioactive, 380 Transformer (ESP), - configurations, 752 -, single-phase, 751 Transmissivity, - calculation from pressure drawdown test, 351 -, definition of, 351 Transportation, -, heavy oil pipeline, 253 -, natural gas pipeline, 254 -, oil and gas, 211
-, oil pipeline, 221 Traverse curves, - - for horizontal flow, 291, 317-322 - - for vertical flow, 289, 301-316 - -, technique of using, 289-291 Treating section, surface operations equipment, 2, 3 Trigger, -, magnetically operated, 469 - use in plunger lift, 469 Triphasic flow, 386, 388 Troubleshooting, - chemelectric treaters, 141 - chemical injectors for demulsification, 137 - electric submergible pump system, 775-784 - flow splitters, 139 - gas separators, 138 - gun barrels, 140 - heat exchangers, 140 - heater treaters, 140 - in treatment of emulsions, 137 - sucker-rod pump system, 599, 601-607, 611 - using ammeter charts, 775-784 - - DST charts, 335-340 - - dynamometer cards, 599, 601-607, 611 - - production logs, 393, 394, 401, 402, 408, 411 Tubing arrangements (in hydraulic pumping), 670 - -, casing-free type system, 670-672 - - - - - - with gas vent, 671, 673 - -, closed power fluid (CPF) systems, 671, 676, 683 - -, fixed type, 679, 680 - - for dual-zone wells, 671, 677 - -, open power fluid (OPF) systems, 670-673, 684 - -, parallel free type system, 671, 672 - -, parallel reverse-circulation type system, 671, 673 Tubing friction losses, 761, 762 Tubing pump, 537, 554 - -, advantages and disadvantages of, 555 - -, schematic diagram of, 554 Turbulent flow, 211 - -, criteria for, 212 - -, Fanning friction factor for, 213, 214 Turboexpander process, 192 Tutweiler method, 184 U-tube technique, 426, 427, 432
- - -, advantages of, 432 Unloading operations (in gas-lift), 436, 437 Unsaturated hydrocarbons, 32, 35, 36 Vacuum conditioning, 190
820 Valves, -, automaic diverter, 3 -, bellows, 456, 457 -, bleeder, 769 -, check, 769 -, circulating, 332 -, constant-flow control, 680 -, differential, 454 -, Bryan -, 455 -, pressure -, 455 -, specific gravity -, 455, 456 -, equalizing, 331 -, flapper type spring, 453 -, flow type, 453, 454 -, 4-way wellhead, 680 -, gas-lift, 428, 429, 450, 453-457 -, opening pressure of - -, 444,447 -, spacing of - -, 435, 441 -, hydraulic, 333 -, intermittent type, 428 -, kick-off, 428, 453 -, mechanically controlled, 429, 454 -, multiple-orifice, 324, 325 -, pressure-charged, 456 -, resistance (to fluid flow) of, 220 -, shut-in, 332 -, subsurface safety, 671, 678 -, tester, 332 -, tubing pressure operated, cycle of, 450 -, spacing equations for - - -, 438 -, unbalanced, 439 -, vent, design of, 162 -, flow capacity of -, 165 Vapor losses, 149 - - through vents, causes of, 151 - -, types of, 149 Vapor pressure, - - of components of petroleum fluids, 43 - -, Reid, 181, 185 Vapor recovery, 12, 149 - -, equipment required for, 154 - -, factors affecting amount of, 153 - -, fundamentals of, 152 Vapor recovery system, 12, 149, 151 - - -, design of, 158 - - -, fast payouts from, 166 - - - , formula for pipe sizes in, 161 - - _, functions of, 152, 154 - - _, gas pressure drop in, 160 - _ - , pressure balance for a typical, 159 - _ - , venting in, 166 Vapor rectification process, 193 Vaporization, -, differential, 71
-, flash, 60, 71 Variable speed drive (VSD),754-756 Velocity-shot method, 380, 381 Vent(s), - lines (gas), pressure drop through, 674, 675 -, pressure and vacuum, 154, 160 - valve design, 162 Venting, 166, 171 -, atmospheric and low-pressure storage tanks, 171 -, capacity requirements, 171 -, means of, 173 -, testing of devices, 174 Venturi meter, 27 Vertical two-phase flow (in pipes), 286 - - - - correlations, 286 - - - - patterns, 287 - - - -, pressure loss in, 286 - - - - regime map, 288 - - - -, traverse curves, 301-316 - - - -, technique of using - -, 289 Vertical test, - -, interference, 368 - -, pulse, 368 Vertipig, 470, 473, 474 Viscosity, -, dynamic, 212 -, gas, 49, 257, 258 -, effect of pressure and temperature on -, 49 -, kinematic, 212, 247 -, liquid, 50, 51, 244, 245 -, oil, 50, 51, 244-247, 698 -, effect of temperature on -, 50, 244, 247, 698 -, effect of pressure on -, 51 -, gas-free -, 245 -, gas-saturated -, 245 -, pressure above bubble point vs. -, 246 -, water, vs. temperature, 699 Vogel’s, - equation, 282, 284 - method for solution-gas-drive reservoirs, 282 - reference IPR curve for solution-gas-drive reservoirs, 665, 666, 759 Volume ratio (hydraulic pumps), 656, 660 Walther’s correlation, 247 Well testing, 327 - - analysis in presence of gas phase, 369 - -, common assumptions in models, 340 - -, gas, 357, 359 - -, multiwell, 365 Weymouth equation, 262, 265, 268 - -, modified for gas flow in hilly terrain, 269 Whinery-Campbell technique, 73
821 Washtanks, 4, 5, 16, 23 Waste oil treatment, 142-145 - - -, agitation for, 144 - - - , chemical addition for, 144 - - - , heating for, 143 - - -, removal of solids during, 145 - _ -, settling time in, 144 - - - systems, 142, 143 Waste water treatment, 13, 14 Water, - adsorbents, 187 - content of gases, 182 - _ _ ,-test for, 183 - cushion (in DST),333, 339 -, free, 134 - holdup, 375, 388, 389
-, removal of, from natural gas, 19, 185 Watercut meter, 376, 377 Wellsite power plant, - _ -, central vs. individual, 688 - - - , functions of, 688 - _ -, Guiberson, 688, 690, 691 - _ -, Kobe, 687, 689, 690 - - -, Oilmaster, 689, 691 Working, - fluid level, 429, 667 - head, 430 - lift, 430 - submergence, 430, 431 Yield strength of material vs. depth for RW type pumps, 551, 552
This page intentionally left blank This page intentionally left blank This Page Intentionally Left Blank
