Sie sind auf Seite 1von 17

S H E L L- A N D - T U B E H E AT E X C H A N G E R S

Effectively Design
Shell-and-Tube
Heat Exchangers
hermal design of shell-and-tube • baffles; and
To make the most
of exchanger T heat exchangers (STHEs) is
done by sophisticated computer
software. However, a good un-
derstanding of the underlying principles
of exchanger design is needed to use this
• nozzles.
Other components include tie-rods and
spacers, pass partition plates, impinge-
ment plate, longitudinal baffle, sealing
strips, supports, and foundation.
design software, software effectively. The Standards of the Tubular Ex-
This article explains the basics of ex- changer Manufacturers Association
one needs to changer thermal design, covering such (TEMA) (1) describe these various com-
topics as: STHE components; classifica- ponents in detail.
understand STHE tion of STHEs according to construction An STHE is divided into three parts:
and according to service; data needed for the front head, the shell, and the rear
classification, thermal design; tubeside design; shellside head. Figure 1 illustrates the TEMA
design, including tube layout, baffling, nomenclature for the various construction
exchanger and shellside pressure drop; and mean possibilities. Exchangers are described by
components, tube temperature difference. The basic equa- the letter codes for the three sections —
tions for tubeside and shellside heat for example, a BFL exchanger has a bon-
layout, baffling, transfer and pressure drop are well- net cover, a two-pass shell with a longitu-
known; here we focus on the application dinal baffle, and a fixed-tubesheet rear
pressure drop, and of these correlations for the optimum de-
sign of heat exchangers. A followup arti-
head.

mean temperature cle on advanced topics in shell-and-tube Classification


heat exchanger design, such as allocation based on construction
difference. of shellside and tubeside fluids, use of Fixed tubesheet. A fixed-tubesheet
multiple shells, overdesign, and fouling, heat exchanger (Figure 2) has straight
is scheduled to appear in the next issue. tubes that are secured at both ends to
Rajiv Mukherjee,
tubesheets welded to the shell. The con-
Components of STHEs struction may have removable channel
Engineers India Ltd.
It is essential for the designer to have a covers (e.g., AEL), bonnet-type channel
good working knowledge of the mechani- covers (e.g., BEM), or integral tubesheets
cal features of STHEs and how they in- (e.g., NEN).
fluence thermal design. The principal The principal advantage of the fixed-
components of an STHE are: tubesheet construction is its low cost be-
• shell; cause of its simple construction. In fact,
• shell cover; the fixed tubesheet is the least expensive
• tubes; construction type, as long as no expan-
• channel; sion joint is required.
• channel cover; Other advantages are that the tubes can
• tubesheet; be cleaned mechanically after removal of

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998 ©Copyright 1997 American Institute of Chemical Engineers. All rights reserved. Copying and downloading permitted with restrictions.
S H E L L- A N D - T U B E H E AT E X C H A N G E R S

Stationary Head Types Shell Types Rear Head Types

E L
A
One-Pass Shell Fixed Tube Sheet
Like "A" Stationary Head

Removable Channel and Cover

F M

Two-Pass Shell Fixed Tube Sheet


with Longitudinal Baffle Like "B" Stationary Head

N
G
Bonnet (Integral Cover) Fixed Tube Sheet
Like "C" Stationary Head
Split Flow

H Outside Packed Floating Head


C
Double Split Flow

Integral With Tubesheet S


Removable Cover

Floating Head with Backing Device


J

Divided Flow
T
N
Pull-Through Floating Head

Channel Integral With Tubesheet


K
and Removable Cover
U
Kettle-Type Reboiler
U-Tube Bundle

X W

Externally Sealed
Special High-Pressure Closures Cross Flow Floating Tubesheet

■ Figure 1. TEMA designations for shell-and-tube heat exchangers.

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


the channel cover or bonnet, and that free to “float” within the shell. This changer, the shell cover is removed
leakage of the shellside fluid is mini- permits free expansion of the tube first, then the split backing ring, and
mized since there are no flanged joints. bundle, as well as cleaning of both then the floating-head cover, after
A disadvantage of this design is the insides and outsides of the tubes. which the tube bundle can be re-
that since the bundle is fixed to the Thus, floating-head SHTEs can be moved from the stationary end.
shell and cannot be removed, the out- used for services where both the In the TEMA T construction (Fig-
sides of the tubes cannot be cleaned shellside and the tubeside fluids are ure 5), the entire tube bundle, includ-
mechanically. Thus, its application is dirty — making this the standard con- ing the floating-head assembly, can
limited to clean services on the shell- struction type used in dirty services, be removed from the stationary end,
side. However, if a satisfactory chem- such as in petroleum refineries. since the shell diameter is larger than
ical cleaning program can be em- There are various types of float- the floating-head flange. The floating-
ployed, fixed-tubesheet construction ing-head construction. The two most head cover is bolted directly to the
may be selected for fouling services common are the pull-through with floating tubesheet so that a split back-
on the shellside. backing device (TEMA S) and pull- ing ring is not required.
In the event of a large differential through (TEMA T) designs. The advantage of this construction
temperature between the tubes and The TEMA S design (Figure 4) is is that the tube bundle may be re-
the shell, the tubesheets will be un- the most common configuration in moved from the shell without remov-
able to absorb the differential stress, the chemical process industries (CPI). ing either the shell or the floating-
thereby making it necessary to incor- The floating-head cover is secured head cover, thus reducing mainte-
porate an expansion joint. This takes against the floating tubesheet by bolt- nance time. This design is particular-
away the advantage of low cost to a ing it to an ingenious split backing ly suited to kettle reboilers having a
significant extent. ring. This floating-head closure is lo- dirty heating medium where U-tubes
U-tube. As the name implies, the cated beyond the end of the shell and cannot be employed. Due to the en-
tubes of a U-tube heat exchanger contained by a shell cover of a larger larged shell, this construction has the
(Figure 3) are bent in the shape of a diameter. To dismantle the heat ex- highest cost of all exchanger types.
U. There is only one tubesheet in a U-
tube heat exchanger. However, the
Bonnet Bonnet
lower cost for the single tubesheet is (Stationary Stationary Support Stationary (Stationary
offset by the additional costs incurred Head) Tubesheet Bracket Tubesheet Head)
for the bending of the tubes and the
somewhat larger shell diameter (due
to the minimum U-bend radius), mak-
ing the cost of a U-tube heat ex-
changer comparable to that of a fixed-
tubesheet exchanger.
The advantage of a U-tube heat
exchanger is that because one end is
free, the bundle can expand or con- Baffles Tie Rods
tract in response to stress differen- and Spacers
tials. In addition, the outsides of the
tubes can be cleaned, as the tube bun- ■ Figure 2. Fixed-tubesheet heat exchanger.
dle can be removed.
The disadvantage of the U-tube Header Tubeplate Shell Tubes Baffles
construction is that the insides of the
tubes cannot be cleaned effectively,
since the U-bends would require flex-
ible-end drill shafts for cleaning.
Thus, U-tube heat exchangers should
not be used for services with a dirty
fluid inside tubes.
Floating head. The floating-head
heat exchanger is the most versatile
type of STHE, and also the costliest.
In this design, one tubesheet is fixed
relative to the shell, and the other is ■ Figure 3. U-tube heat exchanger.

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

There are also two types of packed Design data quently, nozzle sizes must sometimes
floating-head construction — outside- Before discussing actual thermal be one size (or even more in excep-
packed stuffing-box (TEMA P) and design, let us look at the data that tional circumstances) larger than the
outside-packed lantern ring (TEMA must be furnished by the process li- corresponding line sizes, especially
W) (see Figure 1). However, since censor before design can begin: for small lines.
they are prone to leakage, their use is 1. flow rates of both streams. 10. preferred tube size. Tube size
limited to services with shellside flu- 2. inlet and outlet temperatures of is designated as O.D. × thickness ×
ids that are nonhazardous and non- both streams. length. Some plant owners have a
toxic and that have moderate pres- 3. operating pressure of both preferred O.D. × thickness (usually
sures and temperatures (40 kg/cm2 streams. This is required for gases, based upon inventory considerations),
and 300°C). especially if the gas density is not and the available plot area will deter-
furnished; it is not really necessary mine the maximum tube length.
Classification for liquids, as their properties do not Many plant owners prefer to stan-
based on service vary with pressure. dardize all three dimensions, again
Basically, a service may be single- 4. allowable pressure drop for based upon inventory considerations.
phase (such as the cooling or heating both streams. This is a very important 11. maximum shell diameter. This
of a liquid or gas) or two-phase (such parameter for heat exchanger design. is based upon tube-bundle removal re-
as condensing or vaporizing). Since Generally, for liquids, a value of quirements and is limited by crane ca-
there are two sides to an STHE, this 0.5–0.7 kg/cm2 is permitted per shell. pacities. Such limitations apply only to
can lead to several combinations of A higher pressure drop is usually war- exchangers with removable tube bun-
services. ranted for viscous liquids, especially dles, namely U-tube and floating-head.
Broadly, services can be classified in the tubeside. For gases, the allowed For fixed-tubesheet exchangers, the
as follows: value is generally 0.05–0.2 kg/cm2, only limitation is the manufacturer’s
• single-phase (both shellside and with 0.1 kg/cm2 being typical. fabrication capability and the avail-
tubeside); 5. fouling resistance for both ability of components such as dished
• condensing (one side condens- streams. If this is not furnished, the ends and flanges. Thus, floating-head
ing and the other single-phase); designer should adopt values speci- heat exchangers are often limited to a
• vaporizing (one side vaporizing fied in the TEMA standards or based shell I.D. of 1.4–1.5 m and a tube
and the other side single-phase); and on past experience. length of 6 m or 9 m, whereas fixed-
• condensing/vaporizing (one side 6. physical properties of both tubesheet heat exchangers can have
condensing and the other side streams. These include viscosity, shells as large as 3 m and tubes
vaporizing). thermal conductivity, density, and lengths up to 12 m or more.
The following nomenclature is specific heat, preferably at both inlet 12. materials of construction. If
usually used: and outlet temperatures. Viscosity the tubes and shell are made of iden-
Heat exchanger: both sides single- data must be supplied at inlet and tical materials, all components should
phase and process streams (that is, outlet temperatures, especially for be of this material. Thus, only the
not a utility). liquids, since the variation with tem- shell and tube materials of construc-
Cooler: one stream a process fluid perature may be considerable and is tion need to be specified. However, if
and the other cooling water or air. irregular (neither linear nor log-log). the shell and tubes are of different
Heater: one stream a process fluid 7. heat duty. The duty specified metallurgy, the materials of all princi-
and the other a hot utility, such as should be consistent for both the pal components should be specified
steam or hot oil. shellside and the tubeside. to avoid any ambiguity. The principal
Condenser: one stream a condens- 8. type of heat exchanger. If not components are shell (and shell
ing vapor and the other cooling water furnished, the designer can choose cover), tubes, channel (and channel
or air. this based upon the characteristics of cover), tubesheets, and baffles.
Chiller: one stream a process the various types of construction de- Tubesheets may be lined or clad.
fluid being condensed at sub-atmo- scribed earlier. In fact, the designer is 13. special considerations. These
spheric temperatures and the other a normally in a better position than the include cycling, upset conditions, al-
boiling refrigerant or process stream. process engineer to do this. ternative operating scenarios, and
Reboiler: one stream a bottoms 9. line sizes. It is desirable to whether operation is continuous or
stream from a distillation column and match nozzle sizes with line sizes to intermittent.
the other a hot utility (steam or hot avoid expanders or reducers. Howev-
oil) or a process stream. er, sizing criteria for nozzles are usu- Tubeside design
This article will focus specifically ally more stringent than for lines, es- Tubeside calculations are quite
on single-phase applications. pecially for the shellside inlet. Conse- straightforward, since tubeside flow

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


represents a simple case of flow Viscosity influences the heat-trans- cooling water (thermal conductivity
through a circular conduit. Heat-trans- fer coefficient in two opposing ways of around 0.55 kcal/h•m•°C) has an
fer coefficient and pressure drop both — as a parameter of the Reynolds extremely high heat-transfer coeffi-
vary with tubeside velocity, the latter number, and as a parameter of Prandtl cient of typically 6,000 kcal/h•m2•°C,
more strongly so. A good design will number. Thus, from Eq. 1c: followed by hydrocarbon liquids
make the best use of the allowable (thermal conductivity between 0.08
pressure drop, as this will yield the h α (µ)0.33–0.8 (2a) and 0.12 kcal/h•m•°C) at 250–1,300
highest heat-transfer coefficient. kcal/h•m2•°C, and then hydrocarbon
If all the tubeside fluid were to h α (µ)–0.47 (2b) gases (thermal conductivity between
flow through all the tubes (one tube 0.02 and 0.03 kcal/h•m•°C) at
pass), it would lead to a certain veloc- In other words, the heat-transfer 50–500 kcal/h•m2•°C.
ity. Usually, this velocity is unaccept- coefficient is inversely proportional Hydrogen is an unusual gas, be-
ably low and therefore has to be in- to viscosity to the 0.47 power. Simi- cause it has an exceptionally high
creased. By incorporating pass parti- larly, the heat-transfer coefficient is thermal conductivity (greater than
tion plates (with appropriate gasket- directly proportional to thermal con- that of hydrocarbon liquids). Thus,
ing) in the channels, the tubeside fluid ductivity to the 0.67 power. its heat-transfer coefficient is to-
is made to flow several times through These two facts lead to some inter- ward the upper limit of the range
a fraction of the total number of tubes. esting generalities about heat transfer. for hydrocarbon liquids.
Thus, in a heat exchanger with 200 A high thermal conductivity promotes The range of heat-transfer coeffi-
tubes and two passes, the fluid flows a high heat-transfer coefficient. Thus, cients for hydrocarbon liquids is
through 100 tubes at a time, and the
velocity will be twice what it would
be if there were only one pass. The Pass Stationary Tie Rods Floating Shell
number of tube passes is usually one, Partition Tubesheet Shell and Spacers Tubesheet Cover
two, four, six, eight, and so on.

Heat-transfer coefficient
The tubeside heat-transfer coeffi-
cient is a function of the Reynolds
number, the Prandtl number, and
the tube diameter. These can be bro-
ken down into the following funda-
mental parameters: physical Support
Stationary-Head Saddles Floating-Head
properties (namely viscosity, ther- Channel Baffles Cover
mal conductivity, and specific heat);
tube diameter; and, very important-
ly, mass velocity. ■ Figure 4. Pull-through floating-head exchanger with backing device (TEMA S).
The variation in liquid viscosity is
quite considerable; so, this physical
Pass Tie Rods Floating-Head
property has the most dramatic effect Partition and Spacers Shell Floating Cover
on heat-transfer coefficient. Tubesheet Weir
The fundamental equation for tur-
bulent heat-transfer inside tubes is:

Nu = 0.027 (Re)0.8 (Pr)0.33 (1a)

or

(hD/k) =
0.027 (DG/µ)0.8 (cµ/k)0.33 (1b)
Support Support
Rearranging: Shell
Stationary-Head Saddle Baffles Saddle
Cover
Channel
h = 0.027(DG/µ)0.8(cµ/k)0.33(k/D) (1c)
■ Figure 5. Pull-through floating-head exchanger (TEMA T).

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

rather large due to the large variation


in their viscosity, from less than 0.1
Table 1. Heat exchanger service for Example 1.
cP for ethylene and propylene to more
than 1,000 cP or more for bitumen. Shellside Tubeside
The large variation in the heat-transfer Fluid Crude oil Heavy gas oil circulating reflux
coefficients of hydrocarbon gases is Flow rate, kg/h 399,831 277,200
attributable to the large variation in Temperature in/out, °C 227 / 249 302 / 275
operating pressure. As operating pres- Operating pressure, kg/cm2 (abs.) 28.3 13.0
sure rises, gas density increases. Pres-
Allowable pressure drop, kg/cm2 1.2 0.7
sure drop is directly proportional to
the square of mass velocity and in- Fouling resistance, h•m2•°C/kcal 0.0007 0.0006
versely proportional to density. There- Heat duty, MM kcal/h 5.4945 5.4945
fore, for the same pressure drop, a Viscosity in/out, cP 0.664 / 0.563 0.32 / 0.389
higher mass velocity can be main- Design pressure, kg/cm2 (gage) 44.0 17.0
tained when the density is higher. This Line size, mm (nominal) 300 300
larger mass velocity translates into a
Material of construction Carbon steel Tubes: Type 410 stainless steel
higher heat-transfer coefficient. Other: 5Cr1Mo

Pressure drop
Mass velocity strongly influences that for a given number of tubes and permitted. If the pressure drop
the heat-transfer coefficient. For tur- two passes, the pressure drop is much through two of these exchangers turns
bulent flow, the tubeside heat-transfer lower than the allowable value, but out to be only 0.8 kg/cm2, the balance
coefficient varies to the 0.8 power of with four passes it exceeds the allow- of 2.7 kg/cm2 would be available for
tubeside mass velocity, whereas tube- able pressure drop. If in such circum- the other three.
side pressure drop varies to the square stances a standard tube has to be em-
of mass velocity. Thus, with increas- ployed, the designer may be forced to Example 1:
ing mass velocity, pressure drop in- accept a rather low velocity. However, Optimizing tubeside design
creases more rapidly than does the if the tube diameter and length may be Consider the heat exchanger ser-
heat-transfer coefficient. Consequent- varied, the allowable pressure drop can vice specified in Table 1. A TEMA
ly, there will be an optimum mass ve- be better utilized and a higher tubeside Type AES exchanger (split-ring pull-
locity above which it will be wasteful velocity realized. through floating-head construction)
to increase mass velocity further. The following tube diameters are was to be employed. Tubes were to
Furthermore, very high velocities usually used in the CPI: w, 1, e, 5, be either 25 mm O.D. (preferred) or
lead to erosion. However, the pres- 1, 14, and 11 in. Of these, 5 in. and 20 mm O.D., 2 mm thick, and 9 m
sure drop limitation usually becomes 1 in. are the most popular. Tubes long (but could be shorter).
controlling long before erosive veloc- smaller than 5 in. O.D. should not be A first design was produced using
ities are attained. The minimum rec- used for fouling services. The use of 25-mm-O.D. × 9-m tubes (Case A in
ommended liquid velocity inside small-diameter tubes, such as 1 in., Table 2). The tubeside pressure drop
tubes is 1.0 m/s, while the maximum is warranted only for small heat ex- was only 0.17 kg/cm2 even though
is 2.5–3.0 m/s. changers with heat-transfer areas less 0.7 kg/cm2 was permitted. Further,
Pressure drop is proportional to than 20–30 m2. the tubeside heat-transfer resistance
the square of velocity and the total It is important to realize that the was 27.71% of the total, which meant
length of travel. Thus, when the num- total pressure drop for a given stream that if the allowable pressure drop
ber of tube passes is increased for a must be met. The distribution of pres- were better utilized, the heat-transfer
given number of tubes and a given sure drop in the various heat exchang- area would decrease. However, when
tubeside flow rate, the pressure drop ers for a given stream in a particular the number of tube passes was in-
rises to the cube of this increase. In circuit may be varied to obtain good creased from two to four (keeping the
actual practice, the rise is somewhat heat transfer in all the heat exchang- shell diameter the same and decreas-
less because of lower friction factors ers. Consider a hot liquid stream flow- ing the number of tubes from 500 to
at higher Reynolds numbers, so the ing through several preheat exchang- 480 due to the extra pass-partition
exponent should be approximately ers. Normally, a pressure drop of 0.7 lanes), the tubeside pressure drop in-
2.8 instead of 3. kg/cm2 per shell is permitted for liq- creased to 1.06 kg/cm2, which was
Tubeside pressure drop rises steeply uid streams. If there are five such pre- unacceptable. (The shellside design
with an increase in the number of tube heat exchangers, a total pressure drop was satisfactory, with the allowable
passes. Consequently, it often happens of 3.5 kg/cm2 for the circuit would be pressure drop quite well utilized.)

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


m/s earlier), pressure drop (0.51
Table 2. Details of two designs for Example 1. kg/cm2 vs. 0.17 kg/cm2), and heat-
transfer coefficient (1,976 vs. 1,285
Case A Case B
kcal/h•m2•°C) were all much higher.
Shell I.D., mm 925 780 The overall heat-transfer coefficient
Tube O.D. × Number of tubes × 25 × 500 × 2 20 × 540 × 2 for this design was 398 kcal/h•m2•°C
Number of tube passes vs. 356 for Case A.
Heat-transfer area, m2 343 300
Tube pitch × Tube layout angle 32 × 90° 26 × 90° Stepwise calculations
Baffle type Single-segmental Single-segmental for viscous liquids
Baffle spacing, mm 450 400 When the variation in tubeside vis-
Baffle cut, percent of diameter 25 30
cosity is pronounced, a single-point
calculation for the tubeside heat-
Velocity, m/s
transfer coefficient and pressure drop
Shellside 1.15 1.52 will give unrealistic results. This is
Tubeside 1.36 2.17 particularly true in cases where a
Heat-transfer coefficient, kcal/h•m2•°C combination of turbulent (or transi-
Shellside 2,065 2,511 tion) flow and laminar flow exist,
Tubeside 1,285 1,976
since the thermal performance is very
different in these two regimes.
Pressure drop, kg/cm2
In such cases, it will be necessary
Shellside 0.86 1.2 to perform the calculations stepwise
Tubeside 0.17 0.51 or zone-wise. The number of steps or
Resistance, % zones will be determined by the vari-
Shellside film 17.24 15.84 ation in the tubeside viscosity and
Tubeside film 27.71 21.14 thus the Reynolds number.
Fouling 50.35 57.66
Example 2:
Metal wall 4.69 4.87 Stepwise calculations
Overdesign 8.29 4.87 The principal process parameters
for a kettle-type steam generator in a
refinery are shown in Table 3. The
viscosity of the heavy vacuum gas oil
Table 3. Process parameters for Example 2. varies from 1.6 cP at the inlet to 6.36
cP at the outlet.
Shellside Tubeside A design was produced without
Fluid Boiler feedwater, Steam Heavy vacuum gas oil performing the calculations stepwise
Flow rate, kg/h 23,100 (fully vaporized) 129,085 — that is, on the basis of a single av-
Temperature in/out, °C 154 / 154 299 / 165 erage temperature and corresponding
Allowable pressure drop, kg/cm2 Negligible 1.4 physical properties. Details of this de-
sign are shown in Table 4.
Fouling resistance, h•m2•°C/kcal 0.0002 0.0006
Performing the tubeside calcula-
Viscosity in/out, cP 0.176 / 0.176 1.6 / 6.36 tions stepwise, in ten equal heat duty
Design pressure, kg/cm2 (gage) 6.5 21.3 steps, revealed that the original ex-
Heat duty, kcal/h 11,242,000 11,242,000 changer was undersurfaced. The rele-
vant performance parameters for the
single-point and stepwise calculations
Since the overdesign in the four- still higher than that permitted. are compared in Table 5.
pass configuration was 28.1%, an at- Next, a design with 20-mm-O.D. The main reason for the difference
tempt was made to reduce the tube- tubes was attempted (Case B in Table was the variation in Reynolds num-
side pressure drop by decreasing the 2). The shell diameter and heat-trans- ber, from 9,813 in the first zone to
tube length. When the tube length fer surface decreased considerably, 2,851 in the last zone. In addition,
was reduced to 7.5 m, the overdesign from 925 mm to 780 mm, and from the mean temperature difference
was 5.72%, but the tubeside pressure 343 m2 to 300 m2, respectively. The (MTD) decreased drastically, from
drop was 0.91 kg/cm2, which was tubeside velocity (2.17 m/s vs. 1.36 138.47°C in the first zone to a mere

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

17.04°C in the last. Thus, while the Table 4. Design produced for Example 2
initial zones (the hot end) had both a
high heat-transfer coefficient and a
without stepwise calculations.
high MTD, these decreased progres-
Number of kettles 2 (in parallel)
sively toward the outlet (cold) end of
the exchanger. Consequently, while Kettle/port I.D., mm 1,825 / 1,225
the first zone required a length of Tubes per kettle 790 tubes
Type 316 stainless steel
only 2.325 m, the last zone required 25 mm O.D. × 2 mm thick × 9 m long
a length of 44.967 m, even though
Number of tube passes 12
the heat duties were the same. The
tubeside pressure drop was only Tube pitch 32 mm square (90°)
marginally higher by the stepwise Baffling Full support plates only
method, because the tubeside is en- Connections, mm (nominal) Shellside: inlet 75, outlet 3 × 200
tirely in the transition regime (Re be- Tubeside: 150
tween 2,851 and 9,813). Heat-transfer area, m2 1,104 (2 × 552)

Shellside design
The shellside calculations are far Table 5. Performance parameters for Example 2 using
more complex than those for the
tubeside. This is mainly because on single-point and stepwise calculations.
the shellside there is not just one flow
Single-point Stepwise
stream but one principal cross-flow Calculations Calculations
stream and four leakage or bypass Tubeside heat-transfer coefficient, kcal/h•m2•°C 347.9 229.2
streams. There are various shellside
Overall heat-transfer coefficient, kcal/h•m2•°C 244.7 179.3
flow arrangements, as well as various
tube layout patterns and baffling de- Tubeside pressure drop, kg/cm2 1.28 1.35
signs, which together determine the Overdesign, % 24.03 –9.11
shellside stream analysis.

Shell configuration the cold stream leaves at a tempera- changers with tube lengths greater
TEMA defines various shell pat- ture higher than the outlet tempera- than 3 m, since this would exceed the
terns based on the flow of the shell- ture of the hot stream. If a two-pass limit on maximum unsupported tube
side fluid through the shell: E, F, G, (F) shell has only two tube passes, length specified by TEMA — typical-
H, J, K, and X (see Figure 1). this becomes a true countercurrent ar- ly 1.5 m, though it varies with tube
In a TEMA E single-pass shell, the rangement where a large temperature O.D., thickness, and material.
shellside fluid enters the shell at one cross can be achieved. When a larger tube length is need-
end and leaves from the other end. A TEMA J shell is a divided-flow ed, a TEMA H shell (see Figure 1) is
This is the most common shell type shell wherein the shellside fluid en- used. An H shell is basically two G
— more heat exchangers are built to ters the shell at the center and divides shells placed side-by-side, so that
this configuration than all other con- into two halves, one flowing to the there are two full support plates. This
figurations combined. left and the other to the right and is described as a double-split config-
A TEMA F two-pass shell has a leaving separately. They are then uration, as the flow is split twice and
longitudinal baffle that divides the combined into a single stream. This is recombined twice. This construction,
shell into two passes. The shellside identified as a J 1–2 shell. Alterna- too, is invariably employed for hori-
fluid enters at one end, traverses the tively, the stream may be split into zontal thermosyphon reboilers. The
entire length of the exchanger two halves that enter the shell at the advantage of G and H shells is that
through one-half the shell cross-sec- two ends, flow toward the center, and the pressure drop is drastically less
tional area, turns around and flows leave as a single stream, which is and there are no cross baffles.
through the second pass, then finally identified as a J 2–1 shell. A TEMA X shell (see Figure 1) is
leaves at the end of the second pass. A TEMA G shell is a split-flow a pure cross-flow shell where the
The longitudinal baffle stops well shell (see Figure 1). This construction shellside fluid enters at the top (or
short of the tubesheet, so that the is usually employed for horizontal bottom) of the shell, flows across the
fluid can flow into the second pass. thermosyphon reboilers. There is only tubes, and exits from the opposite
The F shell is used for tempera- a central support plate and no baffles. side of the shell. The flow may be
ture-cross situations — that is, where A G shell cannot be used for heat ex- introduced through multiple nozzles

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


located strategically along the length tern. Furthermore, a triangular pat- Tube pitch
of the shell in order to achieve a bet- tern produces high turbulence and Tube pitch is defined as the shortest
ter distribution. The pressure drop therefore a high heat-transfer coeffi- distance between two adjacent tubes.
will be extremely low — in fact, cient. However, at the typical tube For a triangular pattern, TEMA
there is hardly any pressure drop in pitch of 1.25 times the tube O.D., it specifies a minimum tube pitch of
the shell, and what pressure drop does not permit mechanical cleaning 1.25 times the tube O.D. Thus, a 25-
there is, is virtually all in the noz- of tubes, since access lanes are not mm tube pitch is usually employed
zles. Thus, this configuration is em- available. Consequently, a triangular for 20-mm O.D. tubes.
ployed for cooling or condensing va- layout is limited to clean shellside For square patterns, TEMA addi-
pors at low pressure, particularly services. For services that require tionally recommends a minimum
vacuum. Full support plates can be mechanical cleaning on the shellside, cleaning lane of 4 in. (or 6 mm) be-
located if needed for structural in- square patterns must be used. Chemi- tween adjacent tubes. Thus, the mini-
tegrity; they do not interfere with the cal cleaning does not require access mum tube pitch for square patterns is
shellside flow because they are par- lanes, so a triangular layout may be either 1.25 times the tube O.D. or the
allel to the flow direction. used for dirty shellside services pro- tube O.D. plus 6 mm, whichever is
A TEMA K shell (see Figure 1) is vided chemical cleaning is suitable larger. For example, 20-mm tubes
a special cross-flow shell employed and effective. should be laid on a 26-mm (20 mm +
for kettle reboilers (thus the K). It A rotated triangular pattern sel- 6 mm) square pitch, but 25-mm tubes
has an integral vapor-disengagement dom offers any advantages over a should be laid on a 31.25-mm (25
space embodied in an enlarged shell. triangular pattern, and its use is mm × 1.25) square pitch.
Here, too, full support plates can be consequently not very popular. Designers prefer to employ the
employed as required. For dirty shellside services, a minimum recommended tube pitch,
square layout is typically employed. because it leads to the smallest shell
Tube layout patterns However, since this is an in-line diameter for a given number of tubes.
There are four tube layout pat- pattern, it produces lower turbu- However, in exceptional circum-
terns, as shown in Figure 6: triangular lence. Thus, when the shellside stances, the tube pitch may be in-
(30°), rotated triangular (60°), square Reynolds number is low (< 2,000), creased to a higher value, for exam-
(90°), and rotated square (45°). it is usually advantageous to em- ple, to reduce shellside pressure drop.
A triangular (or rotated triangular) ploy a rotated square pattern be- This is particularly true in the case of
pattern will accommodate more tubes cause this produces much higher a cross-flow shell.
than a square (or rotated square) pat- turbulence, which results in a high-
er efficiency of conversion of pres- Baffling
sure drop to heat transfer. Type of baffles. Baffles are used to
As noted earlier, fixed-tubesheet support tubes, enable a desirable ve-
construction is usually employed for locity to be maintained for the shell-
clean services on the shellside, U- side fluid, and prevent failure of tubes
tube construction for clean services due to flow-induced vibration. There
on the tubeside, and floating-head are two types of baffles: plate and rod.
construction for dirty services on Plate baffles may be single-segmental,
Triangular Rotated
both the shellside and tubeside. (For double-segmental, or triple-segmen-
(30 ˚) Triangular clean services on both shellside and tal, as shown in Figure 7.
(60˚) tubeside, either fixed-tubesheet or Baffle spacing. Baffle spacing is
U-tube construction may be used, al- the centerline-to-centerline distance
though U-tube is preferable since it between adjacent baffles. It is the
permits differential expansion be- most vital parameter in STHE design.
tween the shell and the tubes.) The TEMA standards specify the
Hence, a triangular tube pattern may minimum baffle spacing as one-fifth
be used for fixed-tubesheet exchang- of the shell inside diameter or 2 in.,
ers and a square (or rotated square) whichever is greater. Closer spacing
pattern for floating-head exchangers. will result in poor bundle penetration
Square Rotated
For U-tube exchangers, a triangular by the shellside fluid and difficulty in
(90˚) Square pattern may be used provided the mechanically cleaning the outsides of
(45˚) shellside stream is clean and a the tubes. Furthermore, a low baffle
square (or rotated square) pattern if spacing results in a poor stream dis-
■ Figure 6. Tube layout patterns. it is dirty. tribution as will be explained later.

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

Single Segmental Double Segmental Triple Segmental


Baffles Baffles Baffles

No-Tubes-in-Window Segmental Baffles Rod Baffle

■ Figure 7. Types of baffles.

The maximum baffle spacing is the shell inside diameter. Although the case of a two-pass shell (TEMA
the shell inside diameter. Higher baf- this, too, is an important parameter for F), a vertical cut is preferred for ease
fle spacing will lead to predominantly STHE design, its effect is less pro- of fabrication and bundle assembly.
longitudinal flow, which is less effi- found than that of baffle spacing. Baffling is discussed in greater de-
cient than cross-flow, and large un- Baffle cut can vary between 15% tail in (2) and (3).
supported tube spans, which will and 45% of the shell inside diameter.
make the exchanger prone to tube Both very small and very large Equalize cross-flow
failure due to flow-induced vibration. baffle cuts are detrimental to effi- and window velocities
Optimum baffle spacing. For tur- cient heat transfer on the shellside Flow across tubes is referred to as
bulent flow on the shellside (Re > due to large deviation from an ideal cross-flow, whereas flow through the
1,000), the heat-transfer coefficient situation, as illustrated in Figure 9. It window area (that is, through the baffle
varies to the 0.6–0.7 power of veloci- is strongly recommended that only cut area) is referred to as window flow.
ty; however, pressure drop varies to baffle cuts between 20% and 35% be The window velocity and the
the 1.7–2.0 power. For laminar flow employed. Reducing baffle cut cross-flow velocity should be as close
(Re < 100), the exponents are 0.33 for below 20% to increase the shellside as possible — preferably within 20%
the heat-transfer coefficient and 1.0 heat-transfer coefficient or increas-
for pressure drop. Thus, as baffle ing the baffle cut beyond 35% to de-
spacing is reduced, pressure drop in- crease the shellside pressure drop
creases at a much faster rate than usually lead to poor designs. Other
does the heat-transfer coefficient. aspects of tube bundle geometry
This means that there will be an should be changed instead to achieve Baffle
Cut
optimum ratio of baffle spacing to those goals. For example, double-
shell inside diameter that will result segmental baffles or a divided-flow
in the highest efficiency of conver- shell, or even a cross-flow shell,
sion of pressure drop to heat transfer. may be used to reduce the shellside
This optimum ratio is normally be- pressure drop.
tween 0.3 and 0.6. For single-phase fluids on the
Baffle cut. As shown in Figure 8, shellside, a horizontal baffle cut (Fig-
baffle cut is the height of the segment ure 10) is recommended, because this Baffle
that is cut in each baffle to permit the minimizes accumulation of deposits
shellside fluid to flow across the baffle. at the bottom of the shell and also
This is expressed as a percentage of prevents stratification. However, in ■ Figure 8. Baffle cut.

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


Eddies
Main Flow
Baffle

Shell Main
Eddies Flow
Diameter Baffle
Baffle

a. Small Baffle Cut b. Large Baffle Cut c. Ideal Baffle Cut and Baffle Spacing

■ Figure 9. Effect of small and large baffle cuts.

of each other. If they differ by more transfer, the other streams are not as pressure drop of each stream is iden-
than that, repeated acceleration and effective. The A stream is fairly effi- tical, since all the streams begin and
deceleration take place along the cient, because the shellside fluid is end at the inlet and outlet nozzles.
length of the tube bundle, resulting in in contact with the tubes. Similarly, Subsequently, based upon the effi-
inefficient conversion of pressure the C stream is in contact with the ciency of each of these streams, the
drop to heat transfer. peripheral tubes around the bundle, overall shellside stream efficiency
and the F stream is in contact with and thus the shellside heat-transfer
Shellside stream analysis the tubes along the pass-partition coefficient is established.
On the shellside, there is not just lanes. Consequently, these streams Since the flow fractions depend
one stream, but a main cross-flow also experience heat transfer, al- strongly upon the path resistances,
stream and four leakage or bypass though at a lower efficiency than the varying any of the following con-
streams, as illustrated in Figure 11. B stream. However, since the E struction parameters will affect
Tinker (4) proposed calling these stream flows along the shell wall, stream analysis and thereby the shell-
streams the main cross-flow stream where there are no tubes, it encoun- side performance of an exchanger:
(B), a tube-to-baffle-hole leakage ters no heat transfer at all. • baffle spacing and baffle cut;
stream (A), a bundle bypass stream The fractions of the total flow rep- • tube layout angle and tube
(C), a pass-partition bypass stream resented by these five streams can be pitch;
(F), and a baffle-to-shell leakage determined for a particular set of ex- • number of lanes in the flow di-
stream (E). changer geometry and shellside flow rection and lane width;
While the B (main cross-flow) conditions by any sophisticated heat- • clearance between the tube and
stream is highly effective for heat exchanger thermal design software. the baffle hole;
Essentially, the five streams are in • clearance between the shell I.D.
parallel and flow along paths of vary- and the baffle; and
ing hydraulic resistances. Thus, the • location of sealing strips and
flow fractions will be such that the sealing rods.

Horizontal Cut A

A
B C C
F
C C

B B
B
B B
A
Vertical Cut E
E

■ Figure 10. Baffle cut orientation. ■ Figure 11. Shellside flow distribution.

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

Using a very low baffle spacing


tends to increase the leakage and by-
Table 6. Process parameters for Example 3.
pass streams. This is because all five
Shellside Tubeside
shellside streams are in parallel and,
therefore, have the same pressure Fluid Crude oil Heavy gas oil circulating reflux
drop. The leakage path dimensions Flow rate, kg/h 367,647 105,682
are fixed. Consequently, when baffle Temperature in/out, °C 209 / 226 319 / 269
spacing is decreased, the resistance of Heat duty, MM kcal/h 4.0 4.0
the main cross-flow path and thereby Density in/out, kg/m3 730 / 715 655 / 700
its pressure drop increases. Since the
Viscosity in/out, cP 0.52 / 0.46 0.27 / 0.37
pressure drops of all five streams must
be equal, the leakage and bypass Specific heat in/out, kcal/kg•°C 0.63 / 0.65 0.78 / 0.73
streams increase until the pressure Thermal conductivity in/out, kcal/h•m•°C 0.087 / 0.085 0.073 / 0.0795
drops of all the streams balance out. Allowable pressure drop, kg/cm2 1.0 0.7
The net result is a rise in the pressure Fouling resistance, h•m2•°C/kcal 0.0006 0.0006
drop without a corresponding increase Design pressure, kg/cm2 (gage) 36.6 14.0
in the heat-transfer coefficient.
Design temperature, °C 250 340
The shellside fluid viscosity also
affects stream analysis profoundly. In Line size, mm (nominal) 300 150
addition to influencing the shellside Material of construction Carbon steel 5Cr1Mo
heat transfer and pressure drop per-
formance, the stream analysis also
affects the mean temperature differ- Table 7. Effects of varying baffle spacing for a constant 25%
ence (MTD) of the exchanger. This baffle cut for Example 3.
will be discussed in detail later. First,
though, let’s look at an example that Design A Design B Design C
demonstrates how to optimize baffle Baffle spacing, mm 300 350 400
design when there is no significant Tube-to-baffle-hole leakage (A), fraction 0.157 0.141 0.13
temperature profile distortion.
Main cross-flow stream (B), fraction 0.542 0.563 0.577
Example 3: Bundle bypass stream (C), fraction 0.113 0.116 0.119
Optimizing baffle design Baffle-to-shell leakage stream (E), fraction 0.12 0.109 0.1
Consider the heat exchanger ser- Pass-partition bypass stream (F), fraction 0.069 0.072 0.075
vice specified in Table 6. Since there Overall shellside heat-transfer efficiency, % 71.3 73.4 74.9
are two independent variables — baf- Shellside velocity, m/s
fle spacing and baffle cut — we will
Cross-flow 2.5 2.15 1.87
first keep the baffle cut constant at
Window flow 2.34 2.34 2.34
25% and vary the baffle spacing
(Table 7). Later, the baffle spacing Shellside pressure drop, kg/cm2 1.34 1.03 0.79
will be kept constant and the baffle Heat-transfer coefficient, kcal/h•m2•°C
cut varied (Table 8). In real practice, Shellside 2,578 2,498 2,372
both parameters should be varied si- Tubeside 1,402 1,402 1,402
multaneously, but keeping one pa- Overall 401.8 399.8 396.5
rameter constant and varying the
Overdesign, % 7.58 7.08 6.21
other will more vividly demonstrate
the influence of each parameter.
The first design developed is des- increased from 300 mm to 400 mm, less, since the shellside velocity and
ignated Design A in Table 7. Here, the main cross-flow, bundle bypass, the Reynolds number decrease, both
the baffle cut is 25% and the baffle and pass-partition bypass streams in- the shellside heat-transfer coefficient
spacing is 300 mm. In Designs B and crease progressively, whereas the and the shellside pressure drop de-
C, the baffle spacing was changed to tube-to-baffle-hole leakage and baf- crease, but the former at a much
350 mm and 400 mm, respectively. fle-to-shell leakage streams decrease lower rate than the latter. Since the
There is no temperature profile dis- progressively. The overall heat-trans- allowable shellside pressure drop is
tortion problem with these designs. fer efficiency of the shellside stream 1.0 kg/cm2, Design A is ruled out, as
Notice that as the baffle spacing is increases progressively. Neverthe- its shellside pressure drop far ex-

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


Table 8. Effects of varying baffle cut for a constant 400-mm baffle spacing for Example 3.

Design D Design E Design F Design G Design H


Baffle cut, percent of diameter 25 30 33 36 20
Tube-to-baffle-hole leakage (A), fraction 0.13 0.106 0.093 0.08 0.159
Main cross-flow stream (B), fraction 0.577 0.612 0.643 0.674 0.54
Bundle bypass stream (C), fraction 0.119 0.122 0.118 0.117 0.126
Baffle-to-shell leakage stream (E), fraction 0.1 0.091 0.085 0.078 0.114
Pass-partition bypass stream (F), fraction 0.075 0.069 0.062 0.052 0.061
Overall shellside heat-transfer efficiency, % 74.9 73.0 75.7 78.6 72.7
Shellside velocity, m/s
Cross-flow 1.87 1.87 1.87 1.87 1.87
Window flow 2.34 1.86 1.65 1.48 3.09
Shellside pressure drop, kg/cm2 0.79 0.69 0.65 0.6 0.98
Heat-transfer coefficient, kcal/h•m2•°C
Shellside 2,372 2,200 2,074 1,929 2,406
Tubeside 1,402 1,402 1,402 1,402 1,402
Overall 396.5 391.4 387.3 381.9 397.4
Overdesign, % 6.21 4.86 3.76 2.33 6.43

ceeds this limit. Designs B and C are pears to be the best choice, since De- streams. This greatly reduces the
both acceptable. The overdesign sign D cannot be accepted because of cross-flow pressure drop. However,
varies marginally. Thus, it would be the excessive shellside pressure drop. the window velocity and therefore the
prudent to adopt Design C, since it window pressure drop cannot be re-
has a lower pressure drop and a bet- Reducing ∆P duced appreciably (assuming that the
ter stream analysis. by modifying baffle design maximum recommended baffle cut
Now consider the effect of varying Single-pass shell and single-seg- was already tried with single-segmen-
the baffle cut while keeping the baffle mental baffles. The first baffle alter- tal baffles before switching to double-
spacing constant at 400 mm, as native is the single-segmental baffle segmental baffles). Nevertheless,
shown in Table 8. As the baffle cut is in a single-pass (TEMA E) shell. since cross-flow pressure drop is in-
progressively increased from 25% in However, in many situations, the variably much greater than window
Design D to 36% in Design G, the shellside pressure drop is too high pressure drop, there is an appreciable
following changes are observed: with single-segmental baffles in a sin- reduction in the total pressure drop.
• the main cross-flow stream (B) gle-pass shell, even after increasing There is also a decrease in the shell-
fraction increases appreciably; the baffle spacing and baffle cut to the side heat-transfer coefficient, but this
• the tube-to-baffle-hole (A), baf- highest values recommended. Such a is considerably less than the reduction
fle-to-shell (E), and pass-partition (F) situation may arise when handling a in the pressure drop. The use of dou-
stream fractions decrease steadily; very high shellside flow rate or when ble-segmental baffles is covered in
• the bundle bypass (C) stream the shellside fluid is a low-pressure depth in (3).
fraction remains steady; gas. In these cases, the next alterna- Divided-flow shell and single-seg-
• the overall heat-transfer effi- tive that should be considered is the mental baffles. If the allowable shell-
ciency of the shellside stream first de- double-segmental baffle (Figure 7). side pressure drop cannot be satisfied
creases and then increases; and Single-pass shell and double-seg- even with double-segmental baffles at
• as the window velocity decreas- mental baffles. By changing the baf- a relatively large spacing, a divided-
es, the shellside heat-transfer coeffi- fling from single-segmental to double- flow shell (TEMA J) with single-seg-
cient falls; the pressure drop also de- segmental at the same spacing in an mental baffles (Figure 1) should be in-
creases, but not as fast as the heat- otherwise identical heat exchanger, vestigated next. Since pressure drop is
transfer coefficient. the cross-flow velocity is reduced ap- proportional to the square of the veloc-
These observations are reflected in proximately to half, because the shell- ity and to the length of travel, a divid-
the overdesign values. Design E ap- side flow is divided into two parallel ed-flow shell will have approximately

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

one-eighth the pressure drop in an oth- its cost, typically by about 10%. The
erwise identical single-pass exchanger. higher cost is offset to some extent by
The advantage of a divided-flow the higher shellside heat-transfer co-
shell over double-segmental baffles is efficient, since pure cross-flow is
that it offers an even larger reduction more efficient than the combination

Temperature
in pressure drop, since not only cross- of cross-flow and window flow in
flow velocity but even window veloc- conventional designs.
ity can be reduced. The disadvantage Cross-flow shell. There are some
is the increase in cost due to the addi- services where the pressure drop limi-
tional piping required. tation is so severe that none of the
Divided-flow shell and double- above shell/baffling configurations can
segmental baffles. If even a divided- yield a satisfactory design. A steam
flow shell with single-segmental baf- ejector condenser operating at a pres-
fles is unable to meet the allowable sure of 50 mm Hg and having an al- Exchanger Length
shellside pressure drop limit, it will lowable pressure drop of 5 mm Hg is
be necessary to adopt a combination an example. Such situations require the ■ Figure 12. Countercurrent flow.
of a divided-flow shell and double- use of a cross-flow shell (TEMA X).
segmental baffles. With such a com- Here, pure cross-flow takes place at
bination, a very large reduction in a very low velocity, so there is virtually
shellside pressure drop is possible — no pressure drop in the shell. Whatever
to as low as 4% of the pressure drop pressure drop occurs is almost entirely
in a single-pass exchanger with the in the nozzles. Support plates will be

Temperature
same baffle spacing and baffle cut. In needed to meet TEMA requirements
sharp contrast, the heat-transfer coef- and prevent any possible flow-induced
ficient will reduce to about 40%. tube vibration. Since the shellside flow
No-tubes-in-window segmental is parallel to these support plates, shell-
baffles. As baffle spacing is increased side pressure drop is not increased.
to reduce the shellside pressure drop,
an exchanger becomes more prone to Increasing tube pitch
tube failure due to flow-induced vi- For a given number of tubes, the
bration. Exchangers with double-seg- smaller the tube pitch, the smaller the Exchanger Length
mental baffles are less likely to expe- shell diameter, and therefore the
rience such problems than those with lower the cost. Consequently, design- ■ Figure 13. Cocurrent flow.
single-segmental baffles. ers tend to pack in as many tubes as
However, a vibration problem may mechanically possible. However, in the case of X shells, it
persist even with double-segmental As noted earlier, designers gener- may be necessary to increase the tube
baffles. In such cases, a no-tubes-in- ally set the tube pitch at 1.25 times pitch above the TEMA minimum to
window design (Figure 7) should be the tube O.D. For square or rotated meet pressure drop limitations, since
adopted. Here, each tube is supported square pitch, a minimum cleaning there are no other parameters that can
by every baffle, so that the unsupport- lane of 4 in. or 6 mm is recommend- be modified.
ed tube span is the baffle spacing. In ed by TEMA.
exchangers with normal single-seg- As far as thermal-hydraulics are Mean temperature difference
mental baffles, the unsupported tube concerned, the optimum tube-pitch- Temperature difference is the driv-
span is twice the baffle spacing. to-tube-diameter ratio for conversion ing force for heat transfer.
Should it become necessary to use of pressure drop to heat transfer is When two streams flow in op-
a very large baffle spacing to restrict typically 1.25–1.35 for turbulent flow posing directions across a tube wall,
the shellside pressure drop to the per- and around 1.4 for laminar flow. there is true countercurrent flow
mitted value, intermediate supports Increasing the tube pitch to re- (Figure 12). In this situation, the
may be used to increase the natural duce pressure drop is generally not only limitation is that the hot
frequency of the tubes, thus produc- recommended for two reasons. First, stream should at all points be hotter
ing a design that is safe against tube it increases the shell diameter and, than the cold stream. The outlet
failure due to flow-induced vibration. thereby, the cost. Second, reducing temperature of the cold stream may
The no-tubes-in-window design pressure drop by modifying the baf- be higher than the outlet tempera-
requires a larger shell diameter for a fle spacing, baffle cut, or shell type ture of the hot stream, as shown in
given number of tubes. This esclates will result in a cheaper design. Figure 12.

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


Since the temperature difference ality, a very small temperature differ- er, there are some services where this
varies along the length of the heat ence is possible, but this represents an is not true. An example of this is the
exchanger, it has to be weighted to area of uncertainty and the credit is cooling of a viscous liquid — as the
obtain a mean value for single-point very small, so it is usually ignored. liquid is cooled, its viscosity increas-
determination of heat-transfer area. When there is a temperature cross es, and this results in a progressive
The logarithmic mean temperature (that is, the outlet temperature of the reduction in the shellside heat-trans-
difference (LMTD) represents this cold stream is higher than the outlet fer coefficient. In this case, the sim-
weighted value. temperature of the hot stream), and plistic overall MTD approach will be
If the hot and cold streams flow in pure countercurrent flow is not possi- inaccurate, and the exchanger must
the same direction, flow is cocurrent ble, multiple shells in series must be be broken into several sections and
(Figure 13). The mean temperature used. This will be discussed in detail the calculations performed zone-wise.
difference is still represented by the in the followup article scheduled to
LMTD. However, the LMTD for be published in the next issue. Temperature profile distortion
cocurrent flow is lower than that for An F shell has two passes, so if there An important issue that has not
countercurrent flow for the same ter- are two tube passes, this is a pure coun- been considered so far is the tempera-
minal differences. This is because al- tercurrent situation. However, if an F ture profile distortion. As noted earli-
though one terminal temperature dif- shell has four or more tube passes, it is er, the leakage and bypass streams are
ference is very high, the other is far no longer a true countercurrent situation less efficient for heat transfer than the
too low — that is, the temperature and, hence, the Ft correction has to be main cross-flow stream.
differences along the path of heat applied. An F shell having four or more Consider a case where the shellside
transfer are not balanced. tube passes is represented as a 2-4 shell. stream is the cold fluid. Since the
What is even more serious with The Ft factor for a 2-4 shell is identical main cross-flow stream encounters a
cocurrent flow is that the outlet tem- to that for two 1-2 shells in series or two very large fraction of the total heat-
perature of the cold stream must be shell passes. The TEMA Ft factor chart transfer surface, it has to pick up a
somewhat lower than the outlet tem- for three shell passes really represents very large part of the total heat duty.
perature of the hot stream, which is a three shells in series, that for four shell Assume that the cross-flow stream is
serious limitation. Consequently, passes four shells in series, and so on. 58% of the total shellside stream, but
countercurrent flow is always pre- It is important to realize that the that it comes in contact with 80% of
ferred to cocurrent flow. LMTD and Ft factor concept assumes the tubes. As a result, its temperature
These principles apply only to sin- that there is no significant variation in rises more rapidly than if the entire
gle-pass exchangers. However, as the overall heat-transfer coefficient shellside stream were to pick up the
noted earlier, shell-and-tube heat ex- along the length of the shell. Howev- entire heat duty. Therefore, its temper-
changers invariably have two or more
tube passes. Since the shellside fluid
flows in one direction, half the tube
passes experience countercurrent
flow and the other half experience
Tubeside
cocurrent flow. The MTD for this sit-
uation is neither the LMTD for coun- m
Strea
tercurrent flow nor that for cocurrent ross-flow
flow, but a value between the two. Main C am
C Stre
Temperature

A correction factor, Ft, which de- re Profi


le
pends on the four terminal tempera- nt Te mperatu
Appare
tures and the shell style can be deter-
mined from charts in the TEMA stan- E Stream
dards. The LMTD for countercurrent
flow is multiplied by this factor to ob- Shellside
Last Baffle

tain the corrected MTD.


An important limitation for 1-2 C = Bundle-to-Shell Bypass
shells (one shell pass and two or more E = Baffle-to-Shell Leakage
tube passes) is that the outlet tempera-
ture of the cold stream cannot exceed
the outlet temperature of the hot
Exchanger Length
stream. This is because of the presence
of one or more cocurrent passes. In re- ■ Figure 14. Temperature profile distortion factor due to bypass and leakage.

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS


S H E L L- A N D - T U B E H E AT E X C H A N G E R S

ature profile will be steeper than that


of the total stream (the apparent tem-
Table 9. Process parameters for Example 4.
perature profile) without considering
the various flow fractions (Figure 14). Shellside Tubeside
The temperature profiles of the Fluid Naphtha Cooling water
baffle-hole-to-tube leakage, shell-to- Flow rate, kg/h 9,841 65,570
bundle leakage, and pass-partition by-
Temperature in/out, °C 114 / 40 33 / 40
pass streams will depend on their re-
Heat duty, MM kcal/h 0.46 0.46
spective flow fractions and the frac-
tional heat-transfer area encountered. Specific gravity in/out 0.62 / 0.692 1.0 / 1.0
However, since the shell-to-baffle Viscosity in/out, cP 0.254 / 0.484 0.76 / 0.66
leakage stream does not experience Average specific heat, kcal/kg•°C 0.632 1.0
any heat transfer, the remaining four Thermal conductivity in/out, kcal/h•m•°C 0.092 / 0.101 0.542 / 0.546
streams must pick up the entire heat Allowable pressure drop, kg/cm2 0.7 0.7
duty, so that these four streams to-
Fouling resistance, h•m2•°C/kcal 0.0002 0.0004
gether will have a temperature profile
steeper than that of the apparent Design pressure, kg/cm2 (gage) 12.0 6.5
stream. Consequently, the temperature Design temperature, °C 150 60
difference between the hot and the Material of construction Carbon steel Admirality brass
cold streams will be lower all along
the length of the heat exchanger,
thereby resulting in the reduction of
the MTD. This reduction in the MTD Table 10. Construction parameters for Example 4.
is known as the temperature profile
distortion (or correction) factor. Shell I.D. 500 mm
The temperature profile distortion Tubes 188 tubes, 20 mm O.D. × 2 mm thick × 6 m long
factor is more pronounced when the
Number of tube passes 2
leakage and bypass streams are high,
especially the shell-to-baffle leakage Tube pitch 26 mm square (90°)
stream, and the ratio of shellside tem- Baffling Single-segmental, 140 mm spacing, 21% cut (diameter)
perature difference to the temperature Connections 75 mm on shellside, 150 mm on tubeside
approach at the shell outlet is high. Heat-transfer area 70 m2
The latter is because the closer the
temperature approach at the shell out-
let, the sharper the reduction in MTD. The leakage and bypass streams In many situations, a temperature
tend to be high when the shellside profile distortion factor is unavoid-
viscosity is high and when the baffle able, such as when cooling a viscous
Nomenclature spacing is very low. Thus, care has to liquid over a large temperature
c = stream specific heat, kcal/kg•°C
be exercised in the design of viscous range, and there is no alternative to
D = tube inside diameter, m liquid coolers such as a vacuum the use of multiple shells in series.
Ft = LMTD correction factor, residue cooler in a crude oil refinery. However, in many other situations,
dimensionless The minimum recommended tem- improper baffle spacing unnecessar-
G = stream mass velocity, kg/m2•h perature profile distortion factor is ily imposes such a penalty where it
h = stream heat-transfer coefficient, 0.75. Below this, two or more shells is easily avoidable. Designers nor-
kcal/h•m2•°C in series must be employed. By using mally tend to pack baffles as close
k = stream thermal conductivity, multiple shells in series, the ratio of as possible to get the maximum
kcal/h•m•°C
Nu = Nusselt number = hD/k,
shellside temperature difference to shellside heat-transfer coefficient,
dimensionless the temperature approach at the shell pressure drop permitting. In many
Pr = Prandtl number = cµ/k, outlet is reduced. The mixing of the such cases, the use of somewhat
dimensionless main cross-flow stream with the by- higher baffle spacing will reduce the
Re = Reynolds number = DG/µ, pass and leakage streams after each shell-to-baffle leakage stream (the
dimensionless shell reduces the penalty due to the principal culprit) and hence improve
Greek Letter distortion of the temperature profile the MTD correction factor appre-
µ = stream viscosity, kg/m•h and hence increases the temperature ciably, thereby producing a much
profile distortion factor. better design.

CHEMICAL ENGINEERING PROGRESS • FEBRUARY 1998


Table 11. Detailed results of Example 4 iterations.

Existing Design Alternative No. 1 Alternative No. 2 Alternative No. 3 Alternative No. 4
Baffle spacing, mm 140 160 175 190 210
Stream analysis, fraction of stream
Baffle-hole-to-tube leakage (A) 0.189 0.173 0.163 0.154 0.143
Main cross-flow (B) 0.463 0.489 0.506 0.521 0.539
Shell-to-bundle leakage (C) 0.109 0.113 0.116 0.118 0.121
Shell-to-baffle leakage (E) 0.24 0.225 0.215 0.207 0.196
Pass-partition bypass stream (F) 0 0 0 0 0
Overall shellside heat-transfer 62 64.7 66.4 67.9 69.7
efficiency, %
Temperature profile distortion factor 0.6 0.692 0.735 0.766 0.794
Shellside velocity, m/s 0.15 0.14 0.13 0.13 0.12
Shellside heat-transfer coefficient, 614 570 562 550 512
kcal/h•m2•°C
Shellside pressure drop, kg/cm2 0.034 0.029 0.027 0.026 0.023
Overall heat-transfer coefficient,
kcal/h•m2•°C 380 362 359 354 338
Mean temperature difference, °C 13.73 15.9 16.87 17.58 18.22
Overdesign, % –21.1 –12.8 –8.26 –5.73 –6.61

Example 4: Temperature
distortion and baffle spacing R. MUKHERJEE is assistant chief consultant in Literature Cited
the Heat and Mass Transfer Dept. of Engineers
Consider an existing naphtha India Ltd., New Delhi (011-91-11-371-6171; 1. Tubular Exchanger Manufacturers
Fax: 011-91-11-371-5059l; e-mail: Association, “Standards of the Tubular
cooler in a refinery and petrochemi- Exchanger Manufacturers Associa-
shilpi@giasdla.vsnl.net.in), where he has
cal complex. The process parameters been employed since 1971. He has over 26 tion,” 7th ed., TEMA, New York
are listed in Table 9, and the con- years of experience in the design, revamping, (1988).
struction parameters in Table 10. and troubleshooting of air-cooled and shell- 2. Mukherjee, R., “Don’t Let Baffling
The existing design was undersur- and-tube heat exchangers (especially for oil Baffle You,” Chem. Eng. Progress, 92
refineries, gas processing plants, and
faced by 21%, mainly because the (4), pp. 72–79 (Apr. 1996).
petrochemical plants), and also has
temperature profile distortion factor 3. Mukherjee, R., “Use Double-Segmen-
considerable experience in heat-exchanger-
was 0.6, which is lower than the network synthesis and optimization. He has tal Baffles in Shell-and-Tube Heat Ex-
written several articles in technical journals changers,” Chem. Eng. Progress, 88
minimum recommended value of (11), pp. 47–52 (Nov. 1992).
and has presented two papers in the Industrial
0.75. The existing design had a baf- Session of the 10th International Heat Transfer 4. Tinker, T., “Shellside Characteristics of
fle spacing of 140 mm and a baffle Conference at Brighton in August 1994. Shell-and-tube Heat Exchangers: A
cut of 21% (of the diameter). The He has served as faculty for several courses in Simplified Rating System for Commer-
shell-to-baffle leakage stream frac- heat exchanger design, energy conservation, cial Heat Exchangers,” Trans. ASME,
and heat exchanger network optimization.
tion was 0.24. 80, pp. 36–52 (1958).
He is an honors graduate in chemical
To improve the design, the baffle engineering from Jadavpur Univ., Calcutta,
spacing was progressively increased. and is a member of the Indian Institute of
The undersurfacing decreased with Chemical Engineers and the Indian Society Further Reading
for Heat and Mass Transfer.
increasing baffle spacing, up to a Kakac, S., et al., “Heat Exchangers: Ther-
spacing of 190 mm; thereafter, per- mal-Hydraulic Fundamentals and De-
Acknowledgment sign,” Hemisphere Publishing Corp.,
formance again started to deteriorate.
The author is grateful to the management of New York (1981).
Thus, 190 mm is the optimum baffle
Engineers India, Ltd., for permission to publish Schlunder, E.V., et al., eds., “Heat Ex-
spacing. this article and acknowledges the use of Heat changer Design Handbook,” Hemi-
The detailed results of the vari- Transfer Research, Inc.’s software for the sphere Publishing Corp., New York
ous iterations are compared in worked-out examples and their design (1983).
Table 11. CEP methodology.

FEBRUARY 1998 • CHEMICAL ENGINEERING PROGRESS