Rules of Thumb for Chemical Engineers, 5th Edition - 2 Heat Exchangers.pdf

2 Heat Exchangers Shel Shelll Side Side Pr Pressu essure re Dr Drop ... ... ... ... ... ... ... ... ... ... . 46 Heat Transfer Coefficients ............................. 47 Fouling Resistances ...................................... 47 Inst Instal alla lati tion on Reco Recomm mmen enda dati tion onss .. .. .. .. .. .. .. .. .. .. .. .. 48 Ther Th erma mall Cond Conduc ucti tivi vity ty of Metal Metalss .. .. .. .. .. .. .. .. .. .. .. 49 Vac Vacuum Cond Condeenser sers ... .. ... ... ... ... ... ... ... ... ... ... . 50 Air-cooled Heat Exchangers: Forced vs. Induced Dra Draft .... ..... .... .... .... .... .... .... .... .... .... 51 AirAir-co cool oled ed Heat Heat Exch Exchan ange gers rs:: Air Air Data Data .. .. .. .. .. .. .. 52 Air-cooled Heat Exchangers: Thermal Design ..... 52 Air-cooled Heat Exchangers: Pressure Drop, Air Sid Side ........... ............ ............ ............ ..... 55 Air-cooled Heat Exchangers: Temperature Control ........... ........... ............ ............ ....... 55

Intro troduction . .... .... .... .... .... .... .... .... .... .... .... .. 28 TEMA ...... ............ ............ ............ ............ 28 Selection Guides .......................................... 33 Design Recommendations .............................. 35 Process Data .............................................. 37 Heat Exchanger Configuration and Area ........... 38 Determining the LMTD Configuration Correction Factor ........................................ 39 Tube Tu besi side de Pr Pressur ssuree Drop Drop ... ... ... ... ... .. ... ... ... ... ... 40 Tube Side Film Coefficient ............................. 40 Shell Dia Diameter .... .... .... .... .... .... .... .... .... .... .... 41 Ideal Shell Side Film Coefficient ...................... 42 Shell Shell Side Film Coefficie Coefficient nt Correction Correction Factors Factors . . . . 43 Overall Heat Transfer Coefficient .................... 45

Rules of Thumb for Chemical Engineers. DOI:  DOI:   10.1016/B978-0-12-387785-7.00002-5 Copyright      2012 Elsevier Inc. All rights reserved.

27

28

Rule Rules s of Thum Thumb b for for Chemi Chemical cal Engi Engine neers ers

Introduction Heat exchangers are critical elements in every process plant. While the majority of exchangers are the shell-andtube type, there are several additional important types. The major types of heat transfer equipment are: •

•

•

•

•

•

 Shell-and-tube   Finned tube   Bare tube  Plate-and-frame   Spiral   Plate coil

This   chapter chapter focuses focuses on shell-and shell-and-tube -tube exchange exchangers, rs, covering topics of interest to typical process engineers. Plate-and-frame and spiral exchangers are also discussed. Four factors impact impact the performan performance, ce, longevit longevity, y, and maintenance maintenance requirements for heat-transfer equipment and related components  [22]  [22]:: •

•

•

•

param paramete eters rs and appli applicat cation ion inform informati ation, on, proper  proper  sizing and selection of heat exchangers is impossible, and all aspects of performance will be compromised. Codes and design design speci�cations. Specifying a TEMA designation and an ASME pressure and temperature requirement will enhance all heat transfer selections. Instal Installat lation ion.. Follow Following ing approp appropria riate te instal installat lation ion recomm recommend endati ations ons can elimin eliminate ate most most prema prematur turee failures failures and greatly enhance enhance the performan performance ce and ef �ciency ciency of the heat transfer unit.   Evaluation. Always evaluate the selections in terms of a ten-ye ten-year ar operat operation ional al perio period, d, consid consideri ering ng all factors.

An Excel Excel workb workbook ook accom accompan panies ies this this chapte chapter. r. The workbook performs calculations for a liquid-liquid shelland-tube and-tube heat exchange exchangerr and complete completess the associate associated d TEMA datasheet.

Initia Initiall knowle knowledge dge and docume documenta ntatio tion n of all the operat operating ing param paramete eters. rs. Witho Without ut correc correctt opera operatin ting g

TEMA Desc Descri ribe be shel shelll-an andd-tu tube be heat heat exch exchan ange gers rs usin using g nome nomenc ncla latu ture re from from the the Stan Standa dard rdss of the the Tubu Tubula lar  r  Exchange Exchangerr Manufact Manufacturer urerss Associatio Association n (TEMA). (TEMA).   Figure 2-1 illustrates 2-1  illustrates the front head, shell, and rear head types and lists letter designations corresponding to each.  Figure 2-2 shows shows six typica typicall heat heat excha exchange ngerr con�guration gurations, s, with their corresponding TEMA designation (e.g., BEM). The various parts of the exchangers are called out with the key to the parts listed in  Table 2-1. 2-1. In additi addition on to the excha exchange ngerr con�guration gurations, s, TEMA provides design and construction standards for three major  clas classe sess of exch exchan ange ger, r, call called ed R, C, and and B.   Table Table 2-2 compares attributes of the three exchanger classes. The three classes are listed in order of decreasing cost (and mechanical performance). performance). Use datash datashee eets ts to tabula tabulate te the primar primary y proces processs and and mechanic mechanical al requirem requirements ents for a heat exchanger. exchanger. TEMA datasheets are recommended because they are well known by engineers and fabricators. Versions with SI and US units units are given in   Figure Figure 2-3 and   Figure Figure 2-4 2-4.. Similar  Similar  datashee datasheets ts from other sources, sources, such as heat exchanger  exchanger  manufacturers and engineering companies, may also be

used. Enter the TEMA designation (e.g., BEM) into the cell labeled  “ Type” on line 6. Enter the TEMA Class (e.g., R) on line 54. The process engineer usually works closely with the exchanger manufacturer to complete the datasheet. Heat  exchanger design is often a trial-and-error process, with different combinations of shell diameter, tube size, length, tube passes, and other other attribute attributess being being tested. tested. All heat  exchanger manufacturers use sophisticated software for  thermal and mechanical design, and they are usually more than happy to assist customers by running multiple design cases. Altho Although ugh comput computers ers solve solve the design design equati equations ons for  most new exchangers, engineers may want to do some preliminary work using the manual methods as described later in this chapter. Sophisticated software such as the HTRI Xchanger Suite [11] Suite  [11] performs  performs rigorous incremental calculations that account for the highly dynamic nature of  heat exchangers. exchangers. The manual manual calculat calculation ion methods methods use physical physical propertie propertiess averaged averaged across the exchange exchanger, r, and provide heat transfer and pressure drop approximations for various zones within the exchanger.

28

Rule Rules s of Thum Thumb b for for Chemi Chemical cal Engi Engine neers ers

Introduction Heat exchangers are critical elements in every process plant. While the majority of exchangers are the shell-andtube type, there are several additional important types. The major types of heat transfer equipment are: •

•

•

•

•

•

 Shell-and-tube   Finned tube   Bare tube  Plate-and-frame   Spiral   Plate coil

This   chapter chapter focuses focuses on shell-and shell-and-tube -tube exchange exchangers, rs, covering topics of interest to typical process engineers. Plate-and-frame and spiral exchangers are also discussed. Four factors impact impact the performan performance, ce, longevit longevity, y, and maintenance maintenance requirements for heat-transfer equipment and related components  [22]  [22]:: •

•

•

•

param paramete eters rs and appli applicat cation ion inform informati ation, on, proper  proper  sizing and selection of heat exchangers is impossible, and all aspects of performance will be compromised. Codes and design design speci�cations. Specifying a TEMA designation and an ASME pressure and temperature requirement will enhance all heat transfer selections. Instal Installat lation ion.. Follow Following ing approp appropria riate te instal installat lation ion recomm recommend endati ations ons can elimin eliminate ate most most prema prematur turee failures failures and greatly enhance enhance the performan performance ce and ef �ciency ciency of the heat transfer unit.   Evaluation. Always evaluate the selections in terms of a ten-ye ten-year ar operat operation ional al perio period, d, consid consideri ering ng all factors.

An Excel Excel workb workbook ook accom accompan panies ies this this chapte chapter. r. The workbook performs calculations for a liquid-liquid shelland-tube and-tube heat exchange exchangerr and complete completess the associate associated d TEMA datasheet.

Initia Initiall knowle knowledge dge and docume documenta ntatio tion n of all the operat operating ing param paramete eters. rs. Witho Without ut correc correctt opera operatin ting g

TEMA Desc Descri ribe be shel shelll-an andd-tu tube be heat heat exch exchan ange gers rs usin using g nome nomenc ncla latu ture re from from the the Stan Standa dard rdss of the the Tubu Tubula lar  r  Exchange Exchangerr Manufact Manufacturer urerss Associatio Association n (TEMA). (TEMA).   Figure 2-1 illustrates 2-1  illustrates the front head, shell, and rear head types and lists letter designations corresponding to each.  Figure 2-2 shows shows six typica typicall heat heat excha exchange ngerr con�guration gurations, s, with their corresponding TEMA designation (e.g., BEM). The various parts of the exchangers are called out with the key to the parts listed in  Table 2-1. 2-1. In additi addition on to the excha exchange ngerr con�guration gurations, s, TEMA provides design and construction standards for three major  clas classe sess of exch exchan ange ger, r, call called ed R, C, and and B.   Table Table 2-2 compares attributes of the three exchanger classes. The three classes are listed in order of decreasing cost (and mechanical performance). performance). Use datash datashee eets ts to tabula tabulate te the primar primary y proces processs and and mechanic mechanical al requirem requirements ents for a heat exchanger. exchanger. TEMA datasheets are recommended because they are well known by engineers and fabricators. Versions with SI and US units units are given in   Figure Figure 2-3 and   Figure Figure 2-4 2-4.. Similar  Similar  datashee datasheets ts from other sources, sources, such as heat exchanger  exchanger  manufacturers and engineering companies, may also be

used. Enter the TEMA designation (e.g., BEM) into the cell labeled  “ Type” on line 6. Enter the TEMA Class (e.g., R) on line 54. The process engineer usually works closely with the exchanger manufacturer to complete the datasheet. Heat  exchanger design is often a trial-and-error process, with different combinations of shell diameter, tube size, length, tube passes, and other other attribute attributess being being tested. tested. All heat  exchanger manufacturers use sophisticated software for  thermal and mechanical design, and they are usually more than happy to assist customers by running multiple design cases. Altho Although ugh comput computers ers solve solve the design design equati equations ons for  most new exchangers, engineers may want to do some preliminary work using the manual methods as described later in this chapter. Sophisticated software such as the HTRI Xchanger Suite [11] Suite  [11] performs  performs rigorous incremental calculations that account for the highly dynamic nature of  heat exchangers. exchangers. The manual manual calculat calculation ion methods methods use physical physical propertie propertiess averaged averaged across the exchange exchanger, r, and provide heat transfer and pressure drop approximations for various zones within the exchanger.

Heat Heat Exch Exchan ange gers rs

Figure 2-1.  Nomenclature for shell-and-tube heat exchangers  [24]  [24]..

29

30

Rules of Thumb for Chemical Engineers

Figure 2-2.   Typical TEMA heat exchangers  [24].

Heat Exchangers

31

Table 2-1 Heat exchanger parts and connections (for Figure 2-2) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Stationary Head e Channel Stationary Head e Bonnet Stationary Head Flange e Channel or Bonnet Channel Cover Stationary Head Nozzle Stationary Tubesheet Tubes Shell Shell Cover Shell Flange e Stationary Head End Shell Flange e Rear Head End Shell Nozzle Shell Cover Flange Expansion Joint Floating Tubesheet Floating Head Cover Floating Head Flange Floating Head Backing Device Split Shear Ring

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

Slop-on Backing Flange Floating Head Cover e External Floating Tubesheet Skirt Packing Box Flange Packing Packing Follower Ring Lantern Ring Tie Rods and Spacers Transverse Baffles or Support Plates Impingement Baffle Longitudinal Baffle Pass Partition Vent Connection Drain Connection Instrument Connection Support Saddle Lifting Lug Support Bracket Weir Liquid Level Connection

Table 2-2 Comparison of TEMA class R, C, and B heat exchangers. Cost decreases from left to right  [23] Attribute

Class R

Class C

Class B

Application

Generally severe requirements such as petroleum and related processing applications 0.125 in. (3.2 mm)

Generally moderate requirements suchas commercial and general process applications 0.0625 in (1.6 mm)

General process service

0.0625 in (1.6 mm)

¾,

1, 1¼, 1½, and 2 in. 1.25 x tube OD ¼ inch lane

R R

R R

8 inch, tabulated ¼ inch minimum 1.3 x tube flow area

6 inch, tabulated ⅛ inch alloy, ¼ inch carbon steel Same as tube flow area

375  F maximum 300 psi up to 24 inch diameter shell 150 psi for 25 to 42 in. 75 psi for 43 to 60 in. Metal jacketed or solid metal for a) internal floating head cover, b) 300 psi and up, c) all hydrocarbons Flatness tolerance specified Outside diameter of the tube

600 psi maximum

Corrosion allowance on carbon steel Tube diameters, OD Tube pitch and minimum cleaning lane Minimum shell diameter Longitudinal baffle thickness Floating head cover cross-over area Lantern ring construction

Gasket materials

Peripheral gasket contact surface Minimum tubesheet thickness with expanded tube joints

Tube hole grooving

Two grooves

Length of expansion

Smaller of 2 inch or tubesheet thickness 3  /     16 16 inch deep grooves required

Tubesheet pass partition grooves

þ ¼, ⅜, ½, and ⅝ in. þ ⅜ tubes may be located 1.2 x tube OD

Metal jacketed or solid metal for a) internal floating head, b) 300 psi and up No tolerance specified 0.75 x tube OK for 1 inch and smaller ⅞ inch for 1¼ OD 1 inch for 1½ OD 1.25 inch for 2 OD Above 300 psi design pressure or 350  F design temperature: 2 grooves Small of 2 x tube OD or 2 inch Over 300 psi: 3 /    16 16  inch deep grooves required or other

þ ⅝ in. þ lane may be  /   

 16 16 inch in 12 inch and smaller shells for ⅝ and  ¾ in tubes 6 inch tabulated ⅛ inch alloy, ¼ inch carbon steel Same as tube flow area  3 3

375  F maximum 300 psi up to 24 inch diameter shell 150 psi for 25 to 42 in. 75 psi for 43 to 60 in. Metal jacketed or solid metal for a) internal floating head, b) 300 psi and up No tolerance specified 0.75 x tube OK for 1 inch and smaller ⅞ inch for 1¼ OD 1 inch for 1½ OD 1.25 inch for 2 OD Two grooves

Smaller of 2 inch or tubesheet thickness Over 300 psi: 3 /    16 16  inch deep grooves required or other (Continued )

32

Rules of Thumb for Chemical Engineers

Table 2-2 Comparison of TEMA class R, C, and B heat exchangers. Cost decreases from left to right  [23]dcont’d Attribute

Pipe tap connections

Class R

Pressure gage connections Thermometer connections Nozzle construction

6000 psi coupling with bar stock plug Required in nozzles 2 inch and up Required in nozzles 4 inch and up No reference to flanges

Minimum bolt size

¾

inch

Class C

Class B

suitable means for retaining gaskets in place 3000 psi coupling

suitable means for retaining gaskets in place 3000 psi coupling with bar stock plug Required in nozzles 2 inch and up Required in nozzles 4 inch and up All nozzles larger than one inch must be flanged ⅝ inch

Specified by purchaser Specified by purchaser No reference to flanges ½ inch recommended; smaller bolting may be used

Figure 2-3.  Data Sheet for shell-and-tube heat exchanger, SI units  [24].

Heat Exchangers

33

Figure 2-4.  Data Sheet for shell-and-tube heat exchanger, US units  [24].

Selection Guides The following factors should be considered when choosing the type of heat exchanger to use for a particular  application:

•

•

Operating conditions: service requirements (e.g., phase change), thermal duty, temperature approach   Cleanliness of the streams

34

Rules of Thumb for Chemical Engineers

•

•

•

•

 Maximum design pressure and temperature  Heating or cooling application  Maintenance requirements   Material compatibility with process �uids: wetted surfaces and gaskets

Shell-and-Tube Heat Exchangers This is the most common type of heat exchanger used in the chemical process industries. It is often the lowest cost  option, especially when made of carbon steel. Off-theshelf models are available in �xed tubesheet and U-tube design con�gurations in smaller sizes, and are usually used for liquid-liquid, reboiling, and gas cooling applications. TEMA Class exchangers are used for most custom  designs, with TEMA B (chemical industry service) being the most common. TEMA guidelines are limited to a shell diameter of 1524 mm (60 in.), working pressure of 207 bar  (3,000 psig), and product of shell diameter times pressure not exceeding 315,000 mm-bar (60,000 in.-psig).

Plate-and-Frame Heat Exchangers In appropriate circumstances, plate-and-frame heat  exchangers offer many advantages compared with

shell-and-tube designs. The plate-and-frame units have higher heat transfer coef �cients –   often three to four  times that of a shell-and-tube exchanger. They are compact, cost effective, and can handle certain fouling �uids. The most ef �cient design is achieved when the hot  and cold �uid �ow rates are approximately the same, resulting in similar velocities on both sides of the plates. This may require different process parameters (i.e., outlet  temperature) to a shell-and-tube exchanger that is speci�ed for the same service where the engineer speci �es a high shellside �ow rate to maximize the shellside �lm  coef �cient. The design of plate-and-frame exchangers is highly specialized and often proprietary. Manufacturers provide some curves and software for use by end users (for  example, see Ref  [10]), but detailed design is normally left  to the manufacturers.

Spiral Heat Exchangers Increased turbulent heat transfer, reduced fouling, easier maintenance, and smaller size characterize the performance of spiral heat exchangers when compared with shell-and-tube exchangers. These are true countercurrent units. Moretta has summarized the design calculations for heat transfer and pressure drop  [17].

Table 2-3 Shell-and-tube exchanger selection guide (cost increases from left to right)  [1]

Floating Head Outside Packed

Floating Head Split Backing Ring

Floating Head Pull-Through Bundle

Floating head

Floating head

Floating head

Yes Yes

Yes Yes

Yes Yes

Yes

Yes

Yes

Yes

Yes, mechanically or chemically

Yes, mechanically or chemically

Yes, mechanically or chemically

Yes, mechanically or chemically

Chemically only

Chemically only

Chemically only

Chemically only

Chemically only

Yes, mechanically or chemically Normally no limitations Yes

Yes, mechanically or chemically Normally no limitations No

Yes, mechanically or chemically Normally no limitations No

Type of Design

“U” Tube

Fixed Tubesheet

Provision for differential expansion Removable bundle Replacement bundle possible Individual tubes replaceable Tube interiors cleanable

Individual tubes free to expand Yes Yes

Expansion joint in shell No Not practical

Only those in outside row Difficult to do mechanically, can do chemically Chemically only

Tube exteriors with triangular pitch cleanable Tube exteriors with square pitch cleanable Number of tube passes Internal gaskets eliminated

Yes, mechanically or chemically Any practical even number possible Yes

Normally no limitations Yes

Heat Exchangers

35

Table 2-4 Compact heat exchanger attributes Exchanger Type

Attributes

Shell-and-tube

Up to 650  C (1200  F); 310 bar (4,500 psig) in the shell, 1380 bar (20,000 psig) in the tubes Up to 4650 m2 (50,000 ft2) heat transfer area Typical maximum sizes Floating Head Fixed Head or U-Tube Diameter 1524 mm (60 in.) 2000 mm (80 in.) Length 9m (30 ft) horizontal 12 m (40 ft) 25 m (75 ft) vertical Area 1270 m2 (13,650 ft2) 4310 m2 (46,400 ft2) Up to 180  C (350  F) and 20 bar (300 psig); fatigue characteristics of the metal plate may be limiting if temperature or pressure cycling is a process characteristic Up to 2800 m2 (30,000 ft2) heat transfer area in a single unit Typically designed with 70 kPa to 100 kPa (10 to 15 psi) pressure drop Maximum flow 2500 m3 /h (11,000 gpm) Minimum velocity 0.1 m/s (0.3 ft/s) Plates 0.5 to 1.2 mm (0.02 to 0.05 in.) thick 0.03 to 2.2 m 2 (0.32 to 23.7 ft 2) area per plate 1.5 to 5.0 mm (0.06 to 0.2 in.) spacing between plates Typically used in clean service (no particles larger than 2.5 mm), although “deep groove” or “wide gap” plate designs can tolerate up to 18 mm particles  [14]. Usually only used for liquid-liquid service. Operates efficiently with crossing temperatures and close approach temperatures Only the plate edges are exposed to atmosphere, so little or no insulation is required Consider when a high-grade, expensive construction material (e.g., tantalum) is required, when space is tight, or when enhanced energy recovery is important High turbulence High heat transfer coefficients High fouling resistance Not available in carbon steel Hotand cold side channels have nearly identical geometry, so hot andcold fluids should have roughly equivalent flow rates Significant size reduction and weight savings compared with shell-and-tube Gasketed exchangers may be unsuitable for use in highly aggressive media or when leakage is not tolerable Up to 450  C (850  F) and 40 bar (600 psig); fatigue characteristics of the metal plate may be limiting if temperature or pressure cycling is a process characteristic Other characteristics are similar to the gasketed plate-and-frame exchangers Up to 500  C (930  F) and 25 bar (360 psig); limits vary depending on size and material of construction Up to 350 m3 /h (1500 gpm); limited due to single channel 0.5 to 500 m2 (5 to 5400 ft 2) heat transfer area in one spiral body Countercurrent design allows for very deep temperature cross and close approach High turbulence reduces fouling and, especially, sedimentation (compared with shell-and-tube) Particularly effective in handling sludges, viscous liquids, and liquids with solids in suspension

Gasketed plate-and-frame

Welded, brazed, or fusion-sealed plate-and-frame Spiral

Design Recommendations For conceptual and preliminary design work, engineers can easily model liquid-liquid shell-and-tube heat  exchangers. Where process �uids undergo a change in state (condensers and boilers), the design calculations are much more complex, and specialized software and training are recommended. Process engineers should start with a full understanding of the duty requirements. After collecting and tabulating thermodynamic properties for the major  �uid components, create heat and material balances for normal operating

conditions (including start-up and turndown scenarios). There may be design trade-off decisions and it is usually the process engineer ’s responsibility to address potential performance differences among alternative design solutions. Here are guidance questions for the process engineer: •

  Which of the following parameters can �oat? To close the heat balance, at least one parameter is determined from the other  �ve: hot and cold stream 

36

Rules of Thumb for Chemical Engineers

•

•

•

•

•

•

•

•

inlet temperature, outlet temperature, and �ow rate. The answer is often �exible, meaning that two or  three of the parameters may be safely varied within ranges. For example, if a liquid-liquid compact heat  exchanger is anticipated, the �ow rate of the two streams should be within about 20% of each other. What variation in temperature of the �uids is expected? This  is particularly pertinent for cooling tower water that has a temperature that varies with the outside dew point temperature.   What are the maximum allowable pressure drops through the equipment for the two streams? Be sure that unintended vaporization would not occur as pressure is reduced.   Are there conditions that could result in freezing, precipitation, or fouling? If the hot stream  �ow is stopped while continuing the cold stream  �ow, what  would happen as the temperature of the stagnant  �uid in the heat exchanger cools?   Similarly, what outcome is expected if the cold stream  �ow stops without interrupting the hot  stream?  Are thermodynamic properties for the hot and cold streams available, or can they be predicted from the properties of the pure components? There are many miscible liquids that behave rationally when mixed; for instance, the mixed liquid viscosity is a logarithmic average of the components ’  mass-weighted viscosities (see Equation 27-3 in Chapter 27). However, other mixtures deviate widely such as polar liquids (e.g., water, alcohols) and nonNewtonian emulsions and slurries.  Is a temperature cross expected and if so can it be avoided? A temperature cross occurs when the outlet  temperature of the hot  �uid is lower than the outlet  temperature of the cold  � uid. It is physically possible in true counter-current equipment such as a spiral heat exchanger, a double-pipe exchanger, and a single-pass type BEM shell-and-tube unit. In many instances, to use shell-and-tube equipment, multiple shells are required. Are there physical limitations? Consider the available space for installation (including logistics of rigging the exchanger into place), maintenance (with an allocation for removing tubes), and elevation requirements (the relationship with associated equipment such as columns and pumps).   Is this a batch or continuous process? Operating ef �ciency, in terms such as pumping cost and

•

maintenance, is usually more important for  exchangers that are in continuous operation for  months, or years, between shutdowns.   How will the �uid �ow rates be controlled? If it ’s planned to control the  � ow rate of cooling water, for  example, a reduction in duty due to process variations, or a lower than planned cooling water  � ow rate due to oversizing the exchanger may result in excessive fouling.

Evaluate the design problem using physical properties appropriate to the temperature of the �uids. This is especially important for viscosity which is highly temperature dependent, is a major contributor to the heat transfer  coef �cient, and plays a central role in pressure drop calculations. For preliminary design work, properties evaluated at the average temperature for each stream are �ne. Calculate the total duty for the exchanger in Watts, or  Btu/h. Add a safety factor of 10% which includes fouling and uncertainty (or another factor depending on the speci�c design problem). Then use the tabulated transfer coef �cients to compute “typical”   heat the required heat transfer area. This is conceptual. The actual required heat transfer area depends on the mechanical design of the exchanger and will be determined later. At this point the top part of the datasheet can be completed and sent to a vendor or heat exchanger engineer  to design an exchanger using one of the sophisticated computer programs they have at their disposal. However, the process engineer may also (or instead) use the approximate methods and procedure given below to come up with a reasonable design solution. The calculations can be solved with spreadsheets to provide a platform for  evaluating alternatives or rating existing exchangers without involving vendors or consultants. Pick either the hot or cold  � uid to  � ow inside the tubes (for a shell-and-tube exchanger). Assume a tube diameter  (usually start with  ¾  inch) and calculate the total length of  tubes to achieve the surface area based on the assumed overall heat transfer coef �cient. Then manipulate the exchanger length and number of tube passes, calculating the pressure drop through the tubes until a combination results in an acceptable pressure drop. Pick a shell type based on the process requirements. Determine its diameter by the tube layout and passes. Estimate the pressure drop through the shell using the method given in this chapter.

Heat Exchangers

Iterate the preceding two steps using different  assumptions (e.g., tube diameter, pressure drop, swapping the �uids between tube and shell side, etc.) to �nd a reasonable design. What is  “ reasonable?” There’s no one “correct ”  answer which is why experience and expertise are important characteristics for the designer. Calculate heat transfer  � lm coef �cients for the tube and shell side and combine with the tube resistance and assumed fouling factors to compute an overall heat  transfer coef �cient. Compare with the original assumption and iterate, using the newly computed coef �cient in place of the assumption, through the design steps if necessary.

37

The proper selection of a heat exchanger depends on interrelated factors; typically, many design solutions are compared before a  �nal design is accepted. Factors include: •

•

•

•

•

•

  Heat transfer rate ( “U”)  Cost (operating and maintenance over the expected life of the exchanger or 10 years)  Pumping power   Size and weight   Materials of construction   Miscellaneous factors such as leak-tightness, safety, reliability, and noise

Process Data The Excel spreadsheet accompanying this chapter steps through the design steps for a shell-and-tube exchanger in liquid-liquid service. The worksheet called “Fluid Data ” tabulates temperature-correlated coef �cients for vapor  pressure, viscosity, density, speci�c heat, and thermal conductivity. It also has point values for molecular weight, heat of vaporization, and �ash point. The fundamental process parameters – �ow and temperature –  are entered on the “Process Data ”  worksheet. There are input cells for all six �ow and temperature values even though at least one of these must be adjusted to satisfy the heat balance. There are also inputs for 

pressure, allowable pressure drop, and fouling resistance. See Figure 2-5. The change in enthalpy for each stream is evaluated using the equation: D H 

¼ W C   ðt   t  Þ  p

out 

in

 

(2-1)

Where:

 ¼  enthalpy change, kJ/h or Btu/h  ¼  mass �ow rate, kg/h or lb/h  ¼  speci�c heat, kJ/kg- C or Btu/lb- F   ¼  temperature at exchanger outlet, C or  F  ¼  temperature at exchanger inlet, C or  F

D H  W  C  p t out  t in

Figure 2-5.   Fundamental process data includes �ow and temperature information for the hot and cold streams. One of the values is calculated based on the other �ve to close the heat balance; the radio buttons identify the unknown.

38

Rules of Thumb for Chemical Engineers

Note that the speci �c heat is equal to the average of the values at inlet and outlet temperatures. The two results (for hot and cold streams) are added in a cell named “HeatBalance. ”   When the heat balance is satis�ed,  D H   for the hot side is a negative value and it is positive for the cold side. Therefore, HeatBalance has a zero value, and Excel ’s GoalSeek function is used to �nd the unknown variable. In this example the cold stream  temperature is found to be 10  C (50  F), and 79,000 W (270,000 Btu/h) are transferred. The heat balance can also be solved algebraically if the heat capacity is assumed to be constant (which is a good assumption). GoalSeek is used by the spreadsheet because

it is easy to implement and allows for changing of the heat  capacity variable with temperature. The stream properties are evaluated as follows. Density, speci�c heat, and thermal conductivity are evaluated for each component of the hot and cold streams at the inlet and outlet temperatures for each stream. They are multiplied by the mass fraction of the component in the stream then summed. This gives an estimate for the properties at the inlet and outlet of the exchanger; as the temperatures are changed during the design procedure, the properties are immediately updated. Viscosity is also tabulated and the logarithmic average is taken, weighted by the mass fraction of the components (see Equation 27-3 in Chapter 27).

Heat Exchanger Configuration and Area Pick either the hot or cold stream to �ow through the tubes. Rules of thumb to help decide include: •

•

•

•

 If one  � uid is highly corrosive, put it inside the tubes to reduce cost. Then only the tubes, tubesheets (sometimes just faced), tube channels, and piping need to be made of the corrosion-resistant alloy.   If one �uid is at a much higher pressure than the other, put it inside the tubes.   If one �uid is much more severely fouling than the other place it in the tubes. Tubes are easier to clean than shells, especially when mechanical means such as brushes are used.   If one �uid has a very limited allowable pressure drop, put it in the shell.

Characterize the tube side by assuming an overall heat  transfer coef �cient (see Table 2-8 on page 47) and a safety factor primarily to account for fouling. Select a tube size (Table 2-2), wall thickness (start with 14 BWG), length (typically 4 ft, 8 ft, 12 ft, 15 ft, or 20 ft), and number of  passes (either 1-pass or an even number up to about 14). After completing all of the calculations in the following sections, return to this step and update the assumed overall heat transfer coef �cient to equal that which was determined by the procedure. Iterate until the calculated overall coef �cient equals the assumed one. The heat transfer area is related to the heat duty, overall heat transfer coef �cient, and mean temperature difference:  A

 ¼ U  DQT 

mean

(2-2)

¼  ¼  ¼

 A  heat transfer area, usually calculated at the outside tube diameter, m 2 or ft 2 Q  heat transferred, W or Btu/h U   overall heat transfer coef �cient, W/m 2- C or Btu/  h-ft 2- F DT mean  mean temperature difference (MTD) between hot and cold streams,  C or   F

¼

Determine the mean temperature difference (MTD) by calculating the log-mean temperature difference (LMTD) then applying a correction factor that is based on the number of tube and shell passes. For a strict cocurrent �ow design (single pass shell and tube), there is no correction factor and this equation applies:

 ð

Þ ¼ ð T   t  ðÞT  ðT t  Þ t  Þ ln ðT   t  Þ

DT mean cocurrent 

in

in

out 

in

out 

out 

in

out 

(2-3)

Other designs use the following formula for LMTD and a correction factor read from graphs corresponding to different shell and tube con �gurations. F   1.0 for a true countercurrent exchanger (shell passes  tube passes). If  the correction factor is less than about 0.80 then consider  adding shells to achieve a result that is closer to countercurrent design.

 ¼  ¼

Heat Exchangers

 ð

Þ ¼ F  ð T   t  ðT Þ ðT t  Þ  t  Þ ln ðT   t  Þ

¼

t   inlet and outlet temperatures of the cold stream, or   F

DT mean countercurrent  in

out 

out 

in

out 

in

(2-4)

out  in

¼ ¼

C

From the tube outside diameter, heat transfer area, and safety factor, calculate the total tube length:  L  All tubes

F    LMTD con�guration correction factor, dimensionless (see next section) T   inlet and outlet temperatures of the hot stream,  C or   F

ntubes

39

 ¼ A pF  d 

safety

(2-5)

o

Determine the minimum number of tubes by dividing the total length, L Alltubes, by tube length and rounding up to the next integer that is evenly divisible by the number of  tube passes. This Excel formula gives the answer:

¼   ROUNDðLengthOfAllTubes =ðTubeLength TubePassesÞÞ þ 0:5; 0ÞTubePasses

Determining the LMTD Configuration Correction Factor Many references present  F   factors in graphical form  (for example: Perry’s). Bowman compiled formulae that  accurately represent the graphs for every con �guration of  shell-and-tube exchanger system  [4]. Fakheri then collapsed the correlations into a single algebraic equation that is applicable to shell-and-tube heat exchangers with N shell passes and 2NM tube passes per shell (for example, with 2 shell passes there may be any multiple of 2N tube passes or 4, 8, 12, etc. tube passes) [6]. F 

¼

1 ln 1

S  ln  W  W  S 

 þ  þ

  þ W  þ S  

S W 

S 

1 1

W   R P

 ¼

T in

 ¼ t T 

out  in

’

 

1

’

W  ’

W 

 

þ p 2

 ffiffi  ffiffi  p  1

’

W 

Where: ’

(2-7)

¼  N  N  N PN þP

2

P

¼ F  ðT   t  Þ out 

in

Assumptions for the  F  factor equations and charts are: •

1= N 

•

•

•

out  in

•

•

in

•

in

For the special case when R cannot be evaluated):

1 1

DT mean

  P  T   t   t   t 

t out 

’

W 

W 

ln

1

PR

’

W 

And:

S W 

1

 R

 ¼

(2-6)

 ffiffi ffi ffi ffiþffi ffi ffi ffi  ¼       ¼  R2

 ffiffi

F 

W 

Where:

p 

p 2 1  

 ¼  1 (and the logarithms

•

  The overall heat transfer coef �cient, U, is constant  throughout the heat exchanger   The rate of  �ow of each �uid is constant    The speci�c heat of each �uid is constant  There is no condensation of vapor or boiling of liquid in a part of the exchanger   Heat losses are negligible  There is equal heat transfer surface area in each pass   The temperature of the shell-side �uid in any shellside pass is uniform over any cross section   There is no leakage of  �uid or heat across the transverse baf �e separating two shell passes

40

Rules of Thumb for Chemical Engineers

Tubeside Pressure Drop Calculate the pressure drop in two parts then add together: 1. Using the mass �ow rate per tube, use equations 2-9, 2-10, and 2-11 to compute pressure drop through the tubes. 2. From the velocity in the tubes and number of  tube passes, estimate the pressure drop for  turning the �ow through the heads or channels with   [15]: DPt 

¼ ð g 1Þ r  u 2 n p

2

(2-8)

c

Where:

 ¼ pressure drop through turns, Pa or psf (divide by 144 for psi) n  ¼  number of passes r ¼ density, kg/m  or lb/ft  u ¼ velocity in tubes, m/s or ft/s g  ¼  conversion factor, 1 m/s or 32.17 ft/s DPt   p

3

3

2

2

c

Compare the calculated and allowable pressure drops. Adjust physical parameters (tube size, exchanger length, and number of tube passes) and repeat the calculations for  heat exchanger area, total tube length, and pressure drop; iterate until a  “reasonable”  con �guration is attained. The “Tube Pressure Drop”  and “F Factor ”   worksheets do the calculations just described.

Tube Side Film Coefficient T   average temperature, shell side �uid,  C or   F U i  overall heat transfer coef �cient based on inside area, W/m 2- C or Btu/ft 2- F

 ¼ ¼

Compute the tube side �lm coef �cient from physical properties evaluated at the average �uid temperature. Use the correlation that corresponds to the �ow regime (laminar, transitional, or turbulent) for the tube side �lm  coef �cient. 1. Calculate the mean wall temperature, then evaluate the viscosity at that temperature. The formula uses the overall heat transfer coef �cient, expressed in terms of the surface area inside the tubes, and the inside �lm coef �cient. Neither of these values is known until the calculations for both the tube side and shell side are complete, so use an assumed value for both then iterate through all of the calculations until the assumed values match the calculated ones. The overall coef �cient was already assumed to estimate the heat transfer area; it was based on the outside area of the tubes (see page 38). A good initial guess for the �lm coef �cient is about 2,000 W/m 2- C or 400 Btu/ft 2- F. T w

 ¼ t  þ

U i hi

ðT   t Þ

 

(2-9)

Where: T w  average inside wall temperature,  C or   F t   average temperature, tube-side �uid,  C or   F

 ¼  ¼

d o

¼ U 

o

d i

hi  inside �lm coef �cient, W/m 2- C or Btu/ft 2- F 2. Use the Hausen correlation for laminar  �ow (Reynolds number  <   2000) [2]:

 ¼

hi

 ¼

k 

"

d i

¼

3:66

 þ

0:0668  N Re  N Pr  d i = L  1

þ 0:40 ½ N 

 ð Þ  ðd  = L Þ =

Re  N Pr 

i

2 3

#   m

0:14

mw

(2-10)

Where the properties are evaluated at the average �uid temperature and L is the length for  the tube pathway (e.g., if there are 10 tubes per  pass then L is the total length of tubing divided by 10).

 ¼  Prandtl Number  ¼ c k  m m ¼ viscosity, mPa-s or lb  /ft-h 3. Use the Sieder Tate equation for turbulent  �ow (Reynolds number  > ¼  10,000) [2]  N Pr 

 p

m 

Heat Exchangers

hi

k 

0:8 1=3 Re  N Pr 

¼ 0:023 d   N  i

0:14

  m

(2-11)

mw

4. Avoid the transition region if possible because the heat transfer coef �cient is very unpredictable and there is a possibility of  �ow oscillations. However, the transition coef �cient is bounded by the laminar  and turbulent coef �cients, and a plausible equation, based on the laminar and turbulent equations, is [2]:

ðh Þ ¼ h þ ðh  h i T 

i

i

i

Þ

 N Re

   2000

8000



41

  (2-12)

The “Tubes htc”  worksheet calculates the �lm coef �cient using the formulae in this section. Input an assumed value for the �lm coef �cient in Cell D7; the spreadsheet  uses this to calculate the wall temperature and evaluate the viscosity at that temperature. Note the calculated coef �cient in Cell D44 and make one or two iterations by changing the assumed value to equal the calculated result.

Shell Diameter The shell diameter is related to the number of tubes, tube passes, tube diameter, tube pitch, tube pitch layout, and tube omissions to allow space for impingement  baf �es or to decrease the number of tubes in the baf �e windows. TEMA and many others publish tables that list  the number of tubes that will �t into shells of standard diameters. For a quick estimation which should suf �ce for  preliminary design work, use this procedure (easily implemented in Excel): 1. Calculate the cross-sectional area occupied by each tube. For triangular pitch, draw the equilateral triangle with vertices at the center of three tubes. The area of the triangle is one-half of the area  required to accommodate one tube. Similarly, for  square pitch draw the square with corners at the center of four tubes. The area of the square is equal to the area required to accommodate one tube.

 Area1  tube;   square

2

¼ 2 ðPR d  Þ ¼ ðPR d  Þ

 Area1  tube;   triangular 

o

o

2

p 3

 ffiffi 4

 

(2-13) (2-14)

Where:

¼  ¼

PR tube pitch ratio (usually 1.25, 1,285, 1.33, or 1.5) d o  outside diameter of tubes, mm or in.

2. Calculate the diameter of a circle that equates to the area for all tubes in the shell.

 Dtight 

¼2



 N t  Areatube p

0:5



(2-15)

 ¼

nt   number of tubes in the shell 3. For each tube pass greater than one, add cross sectional area to account for the pass partition by multiplying the tube diameter by D tight .  Acorrected 

¼D

tight  d o

 ðn  1Þ þ ð N   Area Þ  p

t 

tube

(2-16)

 ¼

n p  number of tube passes in the shell 4. Calculate the minimum shell diameter by adding two tube diameters to the circle equating to Acorrected.

 Ds;min

¼2



 Acorrected  p

0:5



þ 2  d 

o

 

(2-17)

5. Finally, round up to the next standard shell size. For  example, if Ds, minimum   20.5 inches, use the next  standard size which is 21.25 inches (inside diameter)

 ¼

42

Rules of Thumb for Chemical Engineers

Ideal Shell Side Film Coefficient Use the Bell-Delaware method to compute the shell side �lm coef �cient, as described by Bejan and Kraus  [1] and many others. The Bell-Delaware method computes the heat transfer  � lm coef �cient for an ideal bank of tubes, then applies correction factors to account for baf �e cut and spacing, baf �e leakage effects, bundle bypass �ow, variable baf �e spacing in the inlet and outlet sections, and adverse temperature gradient build-up if laminar  �ow. ho

¼h

ideal  J c  J l  J b  J s  J r 

 

 ¼

¼ J 

ideal  c ps

 ¼  the Colburn factor for an ideal tube bank 

 J ideal 

The subscript  s   stands for physical properties at the average temperature of the shell side �uid; subscript  w  is at the wall temperature.

 ¼  ¼

W s  mass �ow rate of shell side �uid across the tube bank   As  bundle cross�ow area at the centerline of the shell between two baf �es

(2-18)

Implied by the nature of the correction factors, many geometrical properties of the shell such as baf �e cut, baf �e spacing, shell diameter, and outside diameter of the tube bundle must be known or estimated. The procedure uses the geometrical properties to calculate each factor. If the geometrical properties are unknown, then a total correction of 0.60 may be used ( ho 0.6 hideal ) since this has “long been used as a rule of thumb ”  [16]. Calculate the ideal heat transfer coef �cient for pure cross�ow in an ideal tube bank from  [13]: hideal 

Where:

For 30  and 90   tube layout bundles, 45   layout with  pt  / d o >  1.707, and 60   layout with  p t  / d o >   3.732:

¼

ws

k s

 As

c ps  m s

ms

bc

 Ds

 D þ ð D  d  otl 

otl 

o

Þ

 pn

¼

 d 

o

 pn



For a 45  and 60  layouts with ratios less than 1.707 and 3.732 respectively, the equation is:

¼ L 

 As

bc

0:14

    ! 2=3

¼ L 

 As

 

 Ds

 D þ ð D  d  otl 

o

otl 

Þ

 pt 

 d 

o

 pn



¼  ¼  ¼

 pt  PR d o , Pitch, which is the Pitch Ratio x tube OD  pn  pitch normal to the  � ow direction (see Table 2-6)  L bc  baf �e spacing

ms;w

(2-19)

Table 2-5 Correlation coefficients for  J ideal  and  f ideal  [13] Pitch Layout

30 30 30 30 30 45 45 45 45 45 60 60 60 60 60 90 90 90 90 90

Reynolds Number

a1

a2

a3

a4

b1

b2

b3

b4

0e10 10e100 100e1000 1000e10000 10000 0e10 10e100 100e1000 1000e10000 10000 0e10 10e100 100e1000 1000e10000 10000 0e10 10e100 100e1000 1000e10000 10000

1.4 1.36 0.593 0.321 0.321 1.55 0.498 0.73 0.37 0.37 1.4 1.36 0.593 0.321 0.321 0.97 0.9 0.408 0.107 0.37

 0.667  0.657  0.477  0.388  0.388  0.667  0.656  0.5  0.396  0.396  0.667  0.657  0.477  0.388  0.388 0.667  0.631  0.46  0.266  0.395

1.45 1.45 1.45 1.45 1.45 1.93 1.93 1.93 1.93 1.93 1.45 1.45 1.45 1.45 1.45 1.187 1.187 1.187 1.187 1.187

0.519 0.519 0.519 0.519 0.519 0.5 0.5 0.5 0.5 0.5 0.519 0.519 0.519 0.519 0.519 0.37 0.37 0.37 0.37 0.37

48 45.1 4.57 0.486 0.372 32 26.2 3.5 0.333 0.303 48 45.1 4.57 0.486 0.372 35 32.1 6.09 0.0815 0.391

1  0.973  0.476  0.152  0.123 1  0.913  0.476  0.136  0.126 1  0.973  0.476  0.152  0.123 1  0.0963  0.602  0.022  0.148

7 7 7 7 7 6.59 6.59 6.59 6.59 6.59 7 7 7 7 7 6.3 6.3 6.3 6.3 6.3

0.5 0.5 0.5 0.5 0.5 0.52 0.52 0.52 0.52 0.52 0.5 0.5 0.5 0.5 0.5 0.378 0.378 0.378 0.378 0.378

þ

 

þ

 

þ

 

þ

 

Heat Exchangers

43

Table 2-6 Tube geometry as a function of tube pitch,  p t  Pitch Normal to Flow, p n 

Tube Layout

30  Triangular Staggered Array

 ffiffi

p t 

 ffiffi

The Colburn factor is a function of the shell side Reynolds number: d o  W s

(2-20)

ms  A s

¼a

1

t 

p 2

 ffiffi

Where: a

a3

 ¼ 1  þ

0:14 N Re;sa4

The coef �cients, listed in Table 2-5, depend on the tube pitch layout and Reynolds number.

Calculate  J ideal  from the following relationship:  J ideal 

t 

p 2 p 

45  Rotated Square Staggered Array

a

  1:33 PR=d o

3  p t  2 p t  2 p t  p t 

p 3 p 

90  Square Inline Array

¼

p ffi ffi!

p t 

60  Rotated Triangular Staggered Array

 N Re;s

Pitch Parallel to Flow, p p 

 N Re;sa2

(2-21)

Shell Side Film Coefficient Correction Factors Where:

This section describes each of the �ve Bell-Delaware correction factors. Some of the equations require additional information about the construction of the heat  exchanger, as noted.

¼ p1  ½p  þ 2 f  sinðarccos  fÞ   2 arccos  f D  2  l  f  ¼ F c

s

c

 Dotl 

Baffle Cut and Spacing,

Jc 

This factor takes into account the heat transfer rate that occurs in the baf �e window where the shell side �uid �ows more longitudinally, deviating from the ideal cross-�ow arrangement. It is related to the shell diameter, tube diameter, and baf �e cut. The value ranges from about 0.53 for a large baf �e cut up to 1.15 for small windows with a high window velocity. If  there are no tubes in the window J c 1.0   [13]. It is expressed as a fraction of the number of tubes in cross �ow, F c [1]; the equation assumes single segmental baf �es:

¼

 J c

 ¼ 0:55  þ

0:72 F c

 

¼  ¼

¼

l c  baf �e cut   distance from the baf �e to the inside of  the shell, mm or in.  Dotl   outside diameter of the tube bundle, mm or in.

(2-22)

Baffle Leakage Effects,

J L 

This factor includes tube-to-shell and tube-to-baf �e leakage, where the shell �uid bypasses the normal �ow path. If baf �es are too closely spaced, the fraction of  �ow in the leakage stream increases compared with cross �ow. It is typically between 0.7 and 0.8  [13]. Use this formula  [1]:  J l 

¼ 0:44 ð1   r  Þ þ ½1  0:044 ð1  r  Þ expð2:2 r  Þ a

a

b

(2-23)

44

Rules of Thumb for Chemical Engineers

Where:

¼  A Aþ A ¼ A  Aþ A

0.0156 in to reduce the leak stream between tube and baf �e hole [19]),

sb

r a r b

sb

tb

sb

tb

 Aw

w

¼ 12 ðp  q Þ D  d s

1

(2-24) q2

2  l c  Ds

¼   arccos 1 d  ¼  D    D , shell-to-baf �e spacing. See Table 2-7.  D  ¼  baf �e diameter  p d   ð1   F  Þ  N   d ;  A ¼

ntw

o

w

t  tb

4

(2-25)

tube-to-baffle leakage area  Where: q3 F w

 ¼

 sin q , fraction of the total number of  3

2 p tubes in one window Ds 2  l c q3  2 arccos  Ds C 1 C 1 Ds Dotl , shell-to-outer tube limit distance baf �e-hole diameter  tube OD (usually 0.8 dtb mm or 0.03125 in., but may be reduced to 0.4 mm or 

¼ ¼ ¼

2

2



tw

w

c

s



o

t 

b

Bundle and Partition Bypass Effects,

b

tb

s

 Awt 

   

s

(2-26)

¼ D8 ðq  sin q Þ, gross window area  ¼   arccos 1  D 2 l  ¼ p4 n  d  , area occupied by tubes in one window  ¼ F   n , number of tubes in the window

 Awg

sb;

Where:

sb

wt ;

Where:

shell-to-baffle leakage area 

q1

wg

free area for fluid flow in one window section

Calculate A sb, Atb, and Aw  as follows:  Asb

¼A A



 



Jb 

This factor corrects for  �ow that bypasses the tube bundle due to clearance between the outermost tubes and the shell and pass dividers. For exchangers with very small clearances the factor is about 0.9, but larger clearances are required for a pull-through �oating head where the factor is about 0.7. Sealing strips can increase the value [13]. A rule of thumb is to use one pair of sealing strips for  approximately every six tube rows [2]. Use these formulae to calculate  J b  [1]:  J b

 ¼

exp

Or  J b

½C r   ð1  c

 ¼

1

for   z

2  z 1=3



Þ

for  z <

1   (2-27) 2

1 2

Table 2-7 Diametric shell-to-baffle clearance, based on TEMA class R [24] Nominal Shell Diameter

DN 200 to 325 350 to 425 450 to 575 600 to 975 1000 to 1350 1375 to 1500

Inches 8 to 13 14 to 17 18 to 23 24 to 39 40 to 54 55 to 60

Shell Type

Pipe Pipe Pipe Rolled Rolled Rolled

Difference in Shell-to-Baffle Diameter

Millimeters 2.540 3.175 3.810 4.445 5.715 7.620

Inches 0.100 0.125 0.150 0.175 0.225 0.300

This parameter strongly influences the calculation of J l . The clearance may be reduced to 0.0035 to 0.004 times the shell diameter limit the baffle-to-  shell leak stream, but only for rolled shells and only if necessary since it is hard to guarantee compliance   [19] .

Heat Exchangers

Where:

¼  1.35 for N <¼ 100 or 1.25 for N > 100 A r   ¼  A n (API Standard 660 requires a seal device z  ¼ n; C 

RE,s

RE,s

bp

 J s

 ¼

nb

(2-28)

bo bc

3/ 5

1/ 3

And  L bi ,  L bo , and  L bc  are baf �e spacing at inlet, outlet, and central respectively

 p

dp

 

o

i

bc

o

 p

otl 

Þ

bi

i

c

s

 n

 ¼  number of baf �es in the exchanger  L   L  ¼  L  L   L   ¼  L  n ¼ for turbulent  �ow or  for laminar  �ow

r  cc

bc

o

nb

r  cc

from 25 mm to 75 mm, 1 in to 3 in., from the baf �e tips and for every 5 to 7 tube pitches thereafter  [19], leading to the rule of thumb of 0.17 for this parameter) nss  number of sealing strip pairs

bp

1

 n

Where:

s ss

s

1

i

b

c

 ¼ D  2  l  n; ¼  p  p  ¼  longitudinal tube pitch  A ¼ L   ð D  D þ 0:5 n  w Þ  L   ¼  central baf �e spacing, mm or in. n ¼ number of bypass divider lanes that are parallel to the cross�ow stream  w  ¼   width of the bypass divider lane (if unknown,

 1  þ ð L  Þð   Þ þ ð L Þð   n  1  þ ð L  Þ þ ð L  Þ

45

 p

bc

dp

 p

assume 2 x Tube OD)

Variations in Baffle Spacing,

Js 

Temperature Gradient for Laminar Flow Regime,

The �nal correction factor is used when the Reynolds number on the shell side is less than 100. It is equal to 1.0 for NRE,s >  100. If N RE,s <  20:  J r 

When baf �e spacing is increased at the ends of the exchanger to accommodate the nozzles, local decreases in �ow velocity occur. This factor accounts for the consequent decrease in heat transfer, and typically ranges from  0.85 to 1.0 [13]. Calculate  J s  with [25]:

Jr 

¼   ¼ 10 nr ;cc

0:18

¼

(2-29)

Where nr,cc is the number of effective tube rows crossed through one cross�ow section. For 20 < NRE,s <   100, perform a linear interpolation between the two extreme values  [1].

Overall Heat Transfer Coefficient Given the tube (inside) and shell (outside) �lm coef �cients, fouling factors, and tube wall thermal conductivity, calculate the overall heat transfer coef �cient for both the clean and fouled conditions. The clean coef �cient is: U o;clean

¼

d o d i  h i

þ

1 d o  ln d o =d i 2 k 

ð

Þþ

(2-30)

1 ho

And the coef �cient in the fouled condition is: U o; fouled 

¼

d o d i  h i

þ

d o  R f ;i d i

þ

1 d o  ln d o =d i 2  k 

ð

Þ þR þ ;  f  o

1 ho

(2-31)

Where:

¼

  overall heat transfer coef �cient based on the U o outside area of the tubes   outside and inside tube diameter, d o   and d i respectively   outside and inside �lm coef �cients, ho   and hi respectively  R f ;o   and R f ;i  fouling factors on the shell and tube side, respectively k   thermal conductivity of the tube material (see Table 2-9)

¼ ¼  ¼

¼

It is good practice to limit the reduction in heat transfer  due to fouling to about 80% of the clean heat transfer 

46

Rules of Thumb for Chemical Engineers

coef �cient. This is done by instituting a cleaning schedule that removes accumulations before they become too severe.

Use this calculated overall heat transfer coef �cient to update the assumed coef �cient (page 18) and iterate the calculations until the values are in reasonable agreement.

Shell Side Pressure Drop The Bell-Delaware method accounts for tube bundle bypass and baf �e leakage effects. It computes a pressure drop that is 20% to 30% of that calculated without the bypass and leakage effects. 1. The crossflow section between the interior baffles. Use the b coef �cients in Table 2-5 to compute the friction factor for an ideal tube bank, which depends on the tube layout and Reynolds number:

 f ideal 

¼b

b

  1:33 PR=d o

1

 N Re;sb2

(2-32)

2. The baffle windows. For an ideal window, calculate the pressure drop using the equation corresponding to the �ow regime.

For  N Re >

¼   100:

DPw;ideal 

0:6 n Þ ¼ W 2 ð g2  þ  A  A  r s

c

 ¼ 1  þ

DPw;ideal 

¼ 26

b3

  ffiffi ffi ffi ffi ffi ffi ffi p  ms  W s

 As  A w  r s

nr ;tw

 pt 

L bc

 d  þ  D o

w2



w

(2-37)

0:14 N Re;sb4

4  f ideal  W s2

nr ;cc

2 r s  g c  A s

mw m

¼

(2-33)

s

¼

¼

¼ exp½C   r   ð1   bp c

p  ffiffi ffi ffiÞ 3

2 z

 

¼

(2-34)

The baf �e leakage correction factor is a function of r a  and  r b  (see page 28); it typically ranges from 0.4 to 0.5. r a  r bc

¼ exp½1:33 ð1 þ Þ  c  ¼ 0:15 ð1  þ r  Þ þ 0:8 a

 Dw

0:14



The bundle bypass correction factor uses parameters determined for  J b, the  � lm coef �cient correction factor for  bundle and partition bypass effects; it typically ranges from 0.5 to 0.8  [13]. For a Reynolds number  <   100, C bp   4.5; Reynolds number  >   100, C bp   3.7. The limit for  R b  is 1.0 for  z>   0.5.

 Rl 

(2-36)

s

s

 ¼

 Rb

w

þ  A  AW   r

The pressure drop for one ideal cross �ow section is: DPb;ideal 

s

If  N Re < 100:

Where: b

tw

 

(2-35)

¼ p d   n 4 Aþ D  q =2 ¼ 0:8 ½l   0:5 ð D  D þ d  Þ w

o

tw

s

2

c

nr ;tw

s

o

otl 

 p p

3. The entrance and exit sections, from the nozzle to the first baffle window. Combined with the cross �ow and baf �e window  � ndings, the total pressure drop through the exchanger (excluding the nozzles) is: DPs

¼ ½ðn  1Þ ðDP ; Þ R þ n  DP ;  R þ 2 DP ;  R 1  þ nn ;; b

b ideal 

b ideal 

b



b

b

r  tw r  cc



w ideal 

l 

(2-38)

Heat Exchangers

47

Heat Transfer Coefficients Table 2-8 Approximate overall heat transfer coefficients [21]

Cold Fluid

U W/m2-  C

U Btu/h-ft2-  F

Water Water Water Water Water Light oil Brine Brine Brine Organic solvents Heavy oils

850e1700 280e850 20e280 340e900 60e280 110e400 570e1140 170e510 20e280 110 e340 50e280

150e300 50e150 3e50 60e160 10e50 20e70 100e200 30e90 3e50 20e60 8e50

Water Light oils Heavy oils Organic solvents Gases Gases Heavy oils Aromatic HC and Steam

1400e4300 280e850 60e450 570 e1140 30 e280 20e230 50e340 30e85

250e750 50e150 10e80 100e200 5e50 4e40 8e60 5e15

Water Light oils Heavy oils (vacuum) Organic solvents Refrigerants Refrigerants

2000e4300 450e1000 140 e430 570 e1140 430e850 170 e570

350e750 80e180 25e75 100e200 75e150 30e100

Water Water Water Water, brine Water, brine Water Water, brine Water Water

2000e4300 1700e3400 570 e1140 280 e680 280 e680 30 e170 60 e280 450 e1140 60 e170

350e750 300e600 100e200 50e120 50e120 5e30 10e50 80e200 10e30

Hot Fluid Sensible Heat Transfer (No Change of Phase) Water Organic solvents Gases Light oils Heavy oils Organic solvents Water Organic solvents Gases Organic solvents Heavy oils Heaters Steam Steam Steam Steam Steam Dowtherm Dowtherm Flue gas Evaporators Steam Steam Steam Steam Water Organic solvents Condensers Steam (pressure) Steam (vacuum) Saturated organic solvents near atmos. Saturated organic solvents with some non-cond Organic solvents, atmospheric and high non-condensable Aromatic vapors, atmospheric with non-condensables Organic solvents, vacuum and high non-condensables Low boiling hydrocarbon, atmospheric High boiling hydrocarbon, vacuum

Fouling Resistances The following are the more common fouling mechanisms  [5]: •

  Crystallization. Certain salts commonly present in natural waters have a lower solubility in warm water  than in cold. Therefore, when cooling water is heated, particularly at the tube wall, these dissolved salts will

•

crystallize on the surface in the form of scale. Common solution: reducing the temperature of the heat transfer surface often softens the deposits.   Sedimentation. Depositing of dirt, sand, rust, and other   small particles is also common when fresh water is used. Common solution: velocity control.

48

Rules of Thumb for Chemical Engineers

•

•

•

•

Biological growth. Common solution: material selection. Smooth surfaces (e.g., chrome plated) and copper or copper alloys reduce biological growth. Chemical reaction coking. This appears where hydrocarbons deposit in a high temperature application. Common solution: reducing the temperature between the �uid and the heat transfer surface.  Corrosion. Common solution: material selection.   Freezing fouling. Overcooling at the heat transfer  surface can cause solidi �cation of some of the �uid stream components. Common solution: reducing the temperature gradient between the �uid and the heat  transfer surface.

Plate-and-frame heat exchangers are usually less prone to fouling than shell-and-tube units. Also, because they have much higher overall heat transfer coef �cients, using the same fouling resistance values as for a shell-and-tube exchanger has a proportionally greater effect on the calculated overall  U . This is a common engineering error  that leads to oversizing the plate-and-frame exchanger. The general practice is to specify plate-and-frame exchangers with no fouling factor, but to specify a percent  of excess surface area instead. Also, select a frame size that will accommodate additional plates in the event that  more surface is needed because of a loss of performance due to fouling. Recent research by HTRI   [11]   shows that fouling in crude oil preheat service depends primarily on velocity, surface temperature, and the composition of the stream. Nesta outlined a  “ no foul design method ” that is applicable to medium through high boiling point liquid hydrocarbon mixtures with API gravity less than 45 [19]. By increasing the velocity of the hydrocarbon above threshold values and providing little or no excess surface area (that normally is allocated for fouling), the method provides much longer  run time than traditional designs. Here is a summary of the no-foul design method from Nesta:

1. Tube side: minimum velocity 2 m/s (6.6 ft/s) for 19 mm (0.75 in.) and 25.4 mm (1 in.) tubes; minimum  velocity 2.2 m/s (7.2 ft/s) for 31.75 mm (1.25 in.) and 38.1 mm (1.5 in.) tubes. 2. Shell side: minimum cross-�ow stream velocity 0.6 m/s (2 ft/s). 3. Maximum temperature at the tube wall: 300  C (570 F). 4. Shell design should use single segmental baf �es with 20% cut, oriented horizontally for TEMA Type E and J shells. Where impingement protection is required, use impingement rods, not plates. 5. Provide up to 20% excess surface area when both streams are within the scope of this design practice, but do not apply a fouling factor. 6. Provide pressure drop as required to achieve the minimum velocities. Building on the no-foul design method, Bennett, et.al. provided this “most basic”  design algorithm  [3]: 1. Check company experience with the heat exchanger  to be designed 2. Decide on fouling factors. If a stream is determined to be non-fouling, do not use a fouling factor for that  stream. If a stream is known to foul, use a fouling factor in accordance with the company ’s best  practices. 3. Place the most heavily fouling stream on the tubeside to facilitate cleaning, if necessary, and to avoid the areas of low velocity that occur on the shellside 4. Design for high velocities within erosion and vibration limits (per the no-foul design method). Exceptions to this general high-velocity rule for  fouling mitigation include corrosion, geothermal brines, and slurries that present an erosion limit. 5. Keep overdesign between 0% and 20%.

Installation Recommendations Here are some installation tips for typical shell-andtube heat exchangers  [12] and  [22]): •

  Provide suf �cient clearance for removing the tube bundle at the head end of the exchanger. For  exchangers with �xed tube sheets, allow enough room to remove the heads and clean the tubes

•

(consider the possibility of using brushes that would be at least as long as the tubes). Provide valves and bypasses in the piping system for  both the shell and tube sides. Ball valves with locking handles are recommended if available for the pipe sizes.

Heat Exchangers

•

•

•

•

•

•

  Provide thermowells and pressure gage connections in the piping at each inlet and outlet, located as close to the unit as practicable. Some exchangers are designed with these features, in which case they can be omitted from the piping. Provide valves to allow venting of gas vapor from the exchanger, and vacuum breakers for exchangers in steam service. The normal locations are close to the steam inlet or on the top portion of the shell.   Ensure that foundations are adequately sized. In concrete footings, foundation bolts set in pipe sleeves of larger size than the bolt size will allow for  adjustment after the foundation has set.  Loosen foundation bolts at one end of unit to allow free expansion and contraction of the heat exchanger  shell.   Exchangers in condensing steam duty should be installed at a 3   to 4  slope, toward the shell outlet, to facilitate drainage of condensate. Heat exchangers should be installed to promote gravity drainage with no vertical lift before or after steam traps. Condensate accumulating in the exchanger results in water  hammer and poor temperature control; corrosion problems may also occur.   Condensate drainage pipes should have a vertical drop-leg of at least 18 inches from the exchanger to the trap.

•

•

•

•

•

•

•

 For condensate capacities of 3,500 kg/h (8,000 lb/h) or less, use a steam trap; for capacities higher than that use a control valve with level controller. If the steam supply is modulated with a control valve, all condensate drains must  �ow by gravity to a collection tank or pumping system to return the condensate to the boiler. Install a condensate drip pocket with a steam trap in front of the steam control valve. Install a strainer in front of the control valve. Locate the valve at least 10 pipe diameters away from  the exchanger, and use a pipe size equal to or larger  than the inlet connection to the unit.  Do not pipe drain connections to a common closed manifold. Install a gage glass in a vapor or gas space to indicate possible �ooding due to faulty trap operation.  Quick-opening and closing valves controlling �uids to or from an exchanger may cause water-hammer, and care should be taken for proper selection of such equipment.   Re-torque all external bolted joints after installation and again after the exchanger has been heated to prevent leaks and blowing out of gaskets.  Insulate all heat-transfer-exposed surface areas.

Thermal Conductivity of Metals Use the values in  Table 2-9   when computing overall heat transfer coef �cients (page 45). Thermal conductivity is the quantity of heat transferred through a unit thickness. Table 2-9 Thermal conductivity of metals used in heat exchangers Heat Exchanger Tube Material

Aluminum Brass, Admiralty Brass, Red Carbon steel (0.5% C) Carbon steel (1.5% C) Copper Hastelloy C Inconel Monel Nickel Tantalum Titanium Type 316 stainless steel Type 410 stainless steel

49

k, W/m-K

147 111 159 54 @ 20  C 36 @ 20  C 33 @ 400  C 386 8.7 14.5 26 90 54 21 16.3 24.9

k, Btu/h-ft- F

85 64 92 31 @ 68  F 21 @ 68  F 19 @ 750  F 223 5 8.4 15 52 31 12 9.4 14.4

50

Rules of Thumb for Chemical Engineers

Vacuum Condensers This section provides tips for designing overhead condensers for vacuum distillation  [20]. Outlet Temperature and Pressure . It is important to have proper subcooling in the vent end of the unit to prevent large amounts of process vapors from going to the vacuum system along with the inerts. Control. It is necessary to have some over-surface and to have a proper baf �ing to allow for pressure control during process swings, variable leakage of inerts, etc. One designer adds 50% to the calculated length for the oversurface. The condenser must be considered part of the control system (similar to extra trays in a fractionator) to allow for process swings not controlled by conventional instrumentation.

The inerts will “blanket ”  a portion of the tubes. The blanketed portion has very poor heat transfer. The column pressure is controlled by varying the percentage of the tube surface blanketed. When the desired pressure is exceeded, the vacuum system will suck out more inerts, and lower the percentage of surface blanketed. This will increase cooling and bring the pressure back down to the desired level. The reverse happens if the pressure falls below that desired. This is simply a matter of adjusting the heat transfer coef �cient to heat balance the system. Figure 2-6   shows typical baf �ing. The inerts move through the �rst part of the condenser as directed by the baf �es. The inerts then pile up at the outlet end lowering heat transfer as required by the controller. A relatively

Figure 2-6. Baf�ing and inlet bathtub  are shown in this typical vacuum condenser design. The vapor inlet nozzle is expanded to �ve times its area. “

”

Heat Exchangers

large section must be covered by more or less stagnant  inerts which are subcooled before being pulled out as needed. Without proper baf �es, the inerts build up in the condensing section and decrease heat transfer until the pressure gets too high. Then the vacuum valve opens wider, pulling process vapor and inerts into the vacuum  system. Under these conditions pressure control will be very poor. Pressure Drop . Baf �ing must be designed to keep the pressure drop as low as possible. The higher the pressure drop the higher the energy consumption and the harder the

51

 job of attaining proper vent end subcooling. Pressure drop is lower at the outlet end because of smaller mass �ow. Bypassing. Baf �es should prevent bypass of inlet  vapor into the vent. This is very important. Typical Condenser.   Figure 2-6   illustrates an inlet  “bathtub”  used for low vacuums to limit pressure drop at  entrance to exchanger and across  � rst rows of tubes. Note the staggered baf �e spacing with large spacing at inlet, and the side-to-side (40% cut) baf �es. Enough baf �es must be used in the inlet end for minimum tube support. In the last 25% of the outlet end a spacing of  1/ 10  of a diameter is recommended.

Air-cooled Heat Exchangers: Forced vs. Induced Draft Air-cooled heat exchangers are classi �ed as forced draft when the tube section is located on the discharge side of the fan, or induced draft when the tube section is located on the suction side of the fan. Forced draft units are more common. Typically, 25.4-mm (1-in.) OD carbon steel tubes are �tted with aluminum  �ns, 12.7 to 15.9 mm high (½ to ⅝

inch), providing outside surface area about 14 to 21 times greater than the area of the bare tubes. The process stream,   �owing inside the tubes, can be cooled to about 10 C to    15 C (20 F to 30 F) above the dry-bulb temperature of the air. Air  � ows at a velocity of 3 to 6 m/s (10 to 20 ft/s).

Table 2-10 Comparison of forced draft and induced draft air-cooled heat exchangers  [8] Attribute

Forced Draft

Induced Draft

Distribution of air across section

Poor distribution of air over the section Greatly increased possibility of hot air recirculation due to low discharge velocity and absence of stack Total exposure of tubes to sun, rain, and hail Easily adaptable for warm air recirculation during freezing conditions Low natural draft capability on fan failure due to small stack effect Slightly lower fan power because the fan is located in the cold air stream (air has higher density) No limit

Better

Effluent air recirculation to intake Influence of weather conditions Freezing conditions Result of fan failure

Power requirement

Temperature limit e discharge air stream Temperature limit e tubeside process fluid

Limited by tube components

Maintenance

Better access to mechanical components

Lower possibility because fan discharges air upward, away from the tubes, at about 2½ times the intake velocity, or about 450 m/min (25 ft/s) Less effect from sun, rain, and hail because 60% of face is covered Warm discharge air not recirculated Natural draft stack effect is greater than forced draft type Slightly higher fan power because the fan is located in the hot air stream (air has lower density) Limited to about 95  C (200  F) to prevent potential damage to fan blades, bearings, belts, and other components in the air stream Limited to 175  C (350  F) because fan failure could subject fan blades and bearings to excessive temperatures Mechanical components are more difficult to access because they are above the tubes

52

Rules of Thumb for Chemical Engineers

Air-cooled Heat Exchangers: Air Data The overall heat transfer coef �cient is governed by the air  � lm heat transfer, which is generally in the order of 60 W/m 2- C (10 Btu/h-ft 2- F). Air-cooled exchangers transfer less than 10% of that of water-cooled shell-and-tube units. Also, the speci �c heat of air is only 25% that of  water (on a mass basis). As a result, air coolers are very large relative to water coolers. On the other hand, the �nned tubes partially offset the poor thermal performance because they provide an external surface area about 20 times that of plain tubes. The performance of air-coolers is tied to the dry-bulb air temperature, which varies considerably throughout the year. Assume a design temperature that is exceeded during 2% to 5% of the annual time period, but calculate the performance of the cooler at the higher end of the temperatures that are known to occur at the plant site, in order to obtain a feel for the performance range to expect.

Obtain the following data to get a realistic estimate of  the design air temperature  [7]: •

•

•

 Annual temperature-probability curve  Typical daily temperature curves  Duration-frequency curves for the occurrence of the maximum dry-bulb temperature

The air density affects fan design ( �ow, head, and power). Table 2-11 gives values for correction factors for  altitude and temperature. Air data should include environmental characteristics. Marine air or sulfur dioxide content can be corrosive to fans, �ns, tubes, and structures. Dusty atmospheres may lead to increased fouling, indicating incorporation of  fouling factors in the design and possibly suggesting design accommodations such as increased tube pitch. Wind and rain patterns should also be considered  [7].

Table 2-11 Approximate correction factor for air density as a function of altitude and temperature Air Temperature Altitude, m (ft)

0 300 (1,000) 600 (2,000) 900 (3,000) 1,200 (4,000) 1,500 (5,000) 1,800 (6,000) 2,100 (7,000) 2,400 (8,000)

L 20  C (0  F)

20  C (70  F)

40  C (100  F)

90  C (200  F)

1.15 1.11 1.07 1.03 0.99 0.96 0.92 0.89 0.86

1.00 0.96 0.93 0.90 0.86 0.83 0.80 0.77 0.74

0.92 0.91 0.88 0.85 0.82 0.79 0.76 0.73 0.70

0.80 0.77 0.75 0.72 0.69 0.67 0.64 0.62 0.60

Air-cooled Heat Exchangers: Thermal Design Thermal performance calculations are analogous with those for shell-and-tube exchangers. The process �uid �ows inside the tubes, and the inside heat transfer  �lm  coef �cient is calculated exactly the same way as with shell-and-tube units. The air  �ows on the outside of the tubes; calculation of the air side �lm coef �cient is complicated; some guidance is given later in this section. For the heat balance, Q  U A MTD, the corrected logmean temperature difference is determined from charts

¼

(Figure 2-7 and Figure 2-8). For four or more tube passes the correction factor is 1; it is slightly less than 1 for threepass units. Use the charts for one- and two-pass coolers. If  the factor is less than 0.8 then strongly consider changing the design temperatures or number of passes to obtain a good design. Engineers can juggle at least nine variables when optimizing the design of an air-cooled heat exchanger. Mukherjee discussed each of these variables in terms of 

Heat Exchangers

Figure 2-7.  MTD correction factors for air-cooled heat exchangers (1-pass, cross- �ow, both �uids unmixed) [8].

Figure 2-8.  MTD correction factors for air-cooled heat exchangers (2-pass, cross- �ow, both �uids unmixed) [8].

53

54

Rules of Thumb for Chemical Engineers

economic impact; highlights are given in   Table 2-12 [18]. Ganapathy has described a procedure for designing an air-cooler  [7]: 1. Identify all process and site data. 2. Assume the layout of the tube bundle, air temperature rise or mass �owrate, and �n geometry. 3. For the assumed values, calculate �lm coef �cients and overall heat transfer coef �cient, effective

temperature difference, and surface area; check this surface against the assumed layout. 4. When the required surface �ts the assumed layout, calculate the tube-side pressure drop and check this against the allowable pressure drop. 5. When surface and tube-side pressure drop are veri�ed, calculate the air-side pressure drop and fan horsepower.

Table 2-12 Variables that must be optimized for air-cooled heat exchanger design [18] Variable

Considerations

Air flow rate

Rule of thumb for face velocity approaching the tube bundle (total flow divided by total area of bundle): e 3 row coil: 240 to 275 m/min (800 to 900 ft/min) e 4 row coil: 150 to 210 m/min (500 to 700 ft/min) e 5 row coil: 140 to 180 m/min (450 to 600 ft/min) e 6 row coil: 100 to 150 m/min (350 to 500 ft/min) Air-side film coefficient varies to the 0.5 power of air mass velocity Air-side pressure drop varies to the 1.75 power of air mass velocity Length is established in conjunction withthe bundle width.There are usually two bundles in a section,and two fansper section. Bundle width normally limited to 3.2 m to 3.5 m (10 ft to 11.5 ft); fans are commonly 3.6 m to 4.3 m (12 ft to 14 ft) in diameter. API 661 specifiesminimumfancoverage of40%. Therefore,tubesare typically inthe range of8 m to10 m long(26 ftto 33ft). Cost of exchanger is lower with smaller diameter tubes Cleaning is more difficult with smaller diameter Minimum recommended (and most common) tube size is 25 mm (1 in) OD Optimize with pressure drop by adjusting the number of passes and tube size Usual fin heights are 9.5 mm, 12.7 mm, and 15.9 mm ( 3/ 8  in.,  1 / 2  in., and  5 / 8  in.) Selection depends on relative values of air-side and tube-side film coefficients With higher fins, fewer tubes can be accommodated per row Typically, use higher fins for steam condensers and water coolers Typically, use lower fins for gas coolers and viscous liquid hydrocarbon coolers Spacing usually varies between 276 to 433 fins/m (7 to 11 fins/in.) Typically, use higher density for steam condensers and water coolers Typically, use lower density for gas coolers and viscous liquid hydrocarbon coolers Staggered pattern almost invariably employed Designers tend to use the following combinations of bare-tube OD, finned-tube OD, and tube pitch: 25 mm / 50 mm /60 mm (1 in / 2 in / 2.375 in.) 25 mm / 57 mm / 67 mm (1 in / 2.25 in / 2.625 in.) As tube pitch is decreased, air-side pressure drop and power consumption increase more rapidly than the air-side heat transfer coefficient Most exchangers have four to six tube rows, but can range from three to ten Air-side film coefficient varies inversely with number of tube rows More rows advantage: more heat transfer area in the same bundle width, reducing number of bundles and sections More rows disadvantage: increases fan horsepower for the same air velocity and lowers the Mean Temperature Difference Typically, four or five tube rows for steam condensers and water coolers Typically, six or seven tube rows for gas coolers and viscous liquid hydrocarbon coolers Distribution of tubes in the various passes need not be uniform; especially useful in condensers where the flow area in each pass can be gradually reduced as the liquid fraction increases progressively Optimize to obtain uniform pressure drop in each pass Power varies directly with volumetric air flow rate and pressure drop Fan horsepower varies to the 2.75 power of the air mass velocity Optimum air mass velocity is higher when air-side heat transfer coefficient is highly controlling (e.g., steam condensers and water coolers) Exchangers are usually designed with a pressure drop between 0.3 in. H 2O and 0.7 in. H 2O

Tube length

Tube outside diameter

Fin height

Fin spacing

Tube pitch

Number of tube rows

Number of tube passes Fan power consumption

Heat Exchangers

Air-Side Heat Transfer Coefficient The Briggs and Young correlation (as reported in [2]) solves for the air-side �lm coef �cient, ho. It was developed empirically using data from tube diameters from 11 mm to 41 mm (0.44 in. to 1.61 in.) and �n heights from 1.4 mm to 16.6 mm (0.056 in. to 0.652 in.). Fin spacings ranged from  0.9 mm to 3 mm (0.035 in. to 0.117 in.); the tubes were in equilateral triangular pitch tube banks with pitches up to 4.5 in. ho

 ¼

C 

k air  d o



d o  r air  umax mair 



0:68

ð N  Þ Pr 

1=3

0:2

     H  s

Y 

0:12

s

(2-39)

Where:

55

 ¼ ¼  ¼  ¼  ¼

C   coef �cient (includes units conversion), 0.000231 (SI) or 0.134 (US) k air    thermal conductivity of air, 0.026 W/m-C or  0.015 Btu/h-ft-F d o  outside diameter of tube (without  �ns), m or ft  rair   density of air, 1.23 kg/m 3 or 0.0765 lb/ft 3 (see Table 2-11) umax  maximum velocity of air, m/h or ft/h umax   is related to the face velocity of the air  approaching the tube bundle by the ratio of total face area to open area between tubes. mair  viscosity of air, 0.0000181Pa-s or 0.0438lb m  /ft-h c p  m air   N Pr   Prandtl number, dimensionless k air 

¼  ¼  ¼ c  ¼ heat capacity of air, 1005 J/kg-C or 0.24 Btu/lb-F  H  ¼  height of  �n, mm or in. s ¼ spacing between �n centers, mm or in. Y  ¼  thickness of  �n, mm or in.  p

2

 ¼

ho  air-side heat transfer  �lm coef �cient, W/m  -C or  Btu/h-ft 2-F

Air-cooled Heat Exchangers: Pressure Drop, Air Side Calculate the air side pressure drop with the Robinson and Briggs correlation (as reported in [2]). Exchangers are usually designed with a pressure drop between 75 Pa and 175 Pa (0.3 in H 2O and 0.7 in H 2O). First, calculate the friction factor in consistent units:  f 

¼ 9:47



d o  r air  u max mair 

0:32

    pt  d o

0:93

(2-40)

DPair 

¼

 ð

 ¼  ¼  ¼

 pt   tube pitch, m or ft  n  number of tube rows in the bundle gc  conversion factor, 1 m/s 2 or 32.17 ft/s2

The other variables are the same as for Equation  2-39, but be sure the units are consistent, especially for u max. Results will be kg/m 2 (x 9.81  Pa) or lbf  /ft 2 (x 0.192  in. H2O).

¼

Then: 2  f n  r air  umax gc

Where:

¼

2

Þ

(2-41)

Air-cooled Heat Exchangers: Temperature Control Various methods are used to control the process �uid outlet temperature: switching fans on and off, use of two-speed or variable-speed motors, use of variable pitch fan blades, and adjustable shutters mounted above

the tube sections. The manufacturer of the heat  exchanger will normally recommend the best  solution after consulting with the buyer and designing the unit.

56

Rules of Thumb for Chemical Engineers

Nomenclature  A

¼ heat transfer area, usually calculated at the outside tube diameter, m  or ft  ¼ tube bundle bypass area  ¼   free �ow area through one cross �ow section evaluated at centerline ¼ shell to baf �e leakage area for a single baf �e ¼ tube to baf �e leakage area for a single baf �e ¼ area available for  �ow through a single baf �e window ¼ �ow area through a single baf �e window with no tubes ¼ window area that is occupied by tubes  ¼ heat capacity, kJ/kg- C or Btu/lb- F ¼ baf �e diameter  ¼ outside diameter of the tube bundle, mm or in. ¼ inside diameter of the shell ¼ effective diameter of a baf �e window ¼ inside tube diameter, consistent units ¼ outside tube diameter, consistent units ¼ LMTD con�guration correction factor, dimensionless ¼ fraction of cross sectional area in the cross�ow section ¼ fraction of cross sectional area in the baf �e window ¼ friction factor  ¼ conversion factor, 1 m/s or 32.17 ft/s ¼ height of  �n, mm or in.  ¼ �lm coef �cient, W/m  - C or Btu/h-ft  - F ¼ Bell Delaware correction factor (various subscripts) ¼ thermal conductivity, W/m- C or Btu/ft- F ¼ tube length ¼ central baf �e spacing ¼ baf �e spacing at inlet  ¼ baf �e spacing at outlet  ¼ baf �e cut  ¼ distance from the baf �e to the 2

 Abp  As  Asb  Atb  Aw  Awg  Awt  c p  Db  Dotl   Ds  Dw d i d o F  F c F w  f  gc  H  h  J  k   L   L bc  L bi  L bo l c  N Pr   N Re

2

2

¼ ¼

inside of the shell, mm or in. c p  m Prandtl number  k  d  r  u Reynolds number  m

 ¼  ¼

nb nr ;cc

2

2

2

nr ;tw nt  nss ntw n p ndp DPt 

PR  pt   pn  p p Q  R f  r a r b r c s T  t 

¼ number of baf �es in the exchanger  ¼ effective tube rows crossed through one cross�ow section ¼ effective tube rows crossed in the window section ¼ number of tubes ¼ number of sealing strip pairs ¼ number of tubes in a baf �e window ¼ number of passes ¼ number of bypass dividers parallel to cross�ow stream  ¼ pressure drop through turns, Pa or psf (divide by 144 for psi) ¼ pitch ratio ¼ tube pitch ¼ tube pitch normal to the �ow direction ¼ tube pitch parallel to the �ow direction ¼ heat transferred, W or Btu/h ¼ fouling factor  ¼ A =ð A þ A Þ ¼ ð A þ A Þ= A ¼ A = A ¼ spacing between �n centers, mm or in. ¼ inlet and outlet temperatures of the hot stream, C or   F ¼ inlet and outlet temperatures of the cold stream, C or   F ¼ mean temperature difference between hot and  cold streams, C or   F ¼ overall heat transfer coef �cient, W/m  - C or  Btu/h-ft  - F ¼ velocity in tubes, m/s or ft/s ¼   mass �ow rate ¼ width of bypass divider lanes that are parallel to the cross�ow stream  ¼ thickness of  �n, mm or in. ¼ density, kg/m  or lb/ft  ¼ viscosity, cP ¼ ratio of sealing strip pairs to tube rows in sb

sb

sb

bp

tb

w

tb

s

DT mean

U 

2

2

u W  w p Y  r m z

3

cross�ow section

3