hol29362_Ch10 11/3/2008 19:51

C H A P T E R 10 Heat Exchangers 531

Figure 10-4 Cross-flow heat exchanger, one fluid

mixed and one unmixed.
He
coo ating
lin or
gf
lui
d

Gas flow

Figure 10-5 Cross-flow heat exchanger, both fluids unmixed.

Gas flow

He
ati
ng
or
x coo
lin
gf
z lui
d

the heat exchanger flows should be considered as only “partially” mixed. Such cases are
discussed in Reference 11.
There are a number of other configurations called compact heat exchangers that are
primarily used in gas-flow systems where the overall heat-transfer coefficients are low and
it is desirable to achieve a large surface area in a small volume. These exchangers generally
have surface areas of greater than 650 m2 per cubic meter of volume and will be given a
fuller discussion in Section 10-7.

10-5 THE LOG MEAN TEMPERATURE

DIFFERENCE
Consider the double-pipe heat exchanger shown in Figure 10-2. The fluids may flow in either
parallel flow or counterflow, and the temperature profiles for these two cases are indicated
in Figure 10-7. We propose to calculate the heat transfer in this double-pipe arrangement

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.531 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

Figure 10-6 Typical temperature profile

for cross-flow heat exchanger
of Figure 10-5.
Tg

Gas
flow

Inlet gas
temperature

Figure 10-7 Temperature profiles for

(a) parallel flow and
(b) counterflow in double-pipe
heat exchanger.
T

Th1 dq = U dA (Th – Tc )
Ho
t fl u
id
Th
Th
2

dq Tc2
Tc
id
d fl u
Col
dA
Tc1

A
1 2
(a)

Th1
Hot f
lu id
Tc1
C ol Th 2
d fl
uid

Tc 2

A
1 2
(b)

532

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.532 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

C H A P T E R 10 Heat Exchangers 533

with
q = UA
Tm [10-5]
where
U = overall heat-transfer coefficient
A = surface area for heat transfer consistent with definition of U

Tm = suitable mean temperature difference across heat exchanger
An inspection of Figure 10-7 shows that the temperature difference between the hot and
cold fluids varies between inlet and outlet, and we must determine the average value for
use in Equation (10-5). For the parallel-flow heat exchanger shown in Figure 10-7, the heat
transferred through an element of area dA may be written
dq = −ṁh ch dTh = ṁc cc dTc [10-6]

where the subscripts h and c designate the hot and cold fluids, respectively. The heat transfer
could also be expressed
dq = U(Th − Tc )dA [10-7]
From Equation (10-6)
−dq
dTh =
ṁh ch
dq
dTc =
ṁc cc
where ṁ represents the mass-flow rate and c is the specific heat of the fluid. Thus


1 1
dTh − dTc = d(Th − Tc ) = −dq + [10-8]
ṁh ch ṁc cc
Solving for dq from Equation (10-7) and substituting into Equation (10-8) gives


d(Th − Tc ) 1 1
= −U + dA [10-9]
Th − Tc ṁh ch ṁc cc
This differential equation may now be integrated between conditions 1 and 2 as indicated
in Figure 10-7. The result is


Th2 − Tc2 1 1
ln = −UA + [10-10]
Th1 − Tc1 ṁh ch ṁc cc
Returning to Equation (10-6), the products ṁc cc and ṁh ch may be expressed in terms of the
total heat transfer q and the overall temperature differences of the hot and cold fluids. Thus
q
ṁh ch =
Th1 − Th2
q
ṁc cc =
Tc2 − Tc1
Substituting these relations into Equation (10-10) gives
(Th2 − Tc2 ) − (Th1 − Tc1 )
q = UA [10-11]
ln[(Th2 − Tc2 )/(Th1 − Tc1 )]
Comparing Equation (10-11) with Equation (10-5), we find that the mean temperature dif-
ference is the grouping of terms in the brackets. Thus
(Th2 − Tc2 ) − (Th1 − Tc1 )

Tm = [10-12]
ln[(Th2 − Tc2 )/(Th1 − Tc1 )]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.533 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

Figure 10-8 Correction-factor plot for exchanger with one shell pass and two, four, or any
multiple of tube passes.
T1

t2
t1

T2
1.0

0.9

R=4
Correction factor F

0 .4
0.8
3.0

0.6
2.0

1.5

1.0

0.2
.0
0.8

0.7

T1 – T2
0.6 R=
t2 – t1

0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
t –t
P= 2 1
T1 – t1

Figure 10-9 Correction-factor plot for exchanger with two shell passes and four, eight, or any
multiple of tube passes.
T1

t2

t1

T2
1.0

0.9
Correction factor F

R = 4.0

0.2
3.0

2.0

1.5

1.0

0.8

0.6

0.4

0.8

0.7

T1 – T2
0.6 R=
t2 – t1

0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
t –t
P= 2 1
T1 – t1

534

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.534 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

C H A P T E R 10 Heat Exchangers 535

Figure 10-10 Correction-factor plot for single-pass cross-flow exchanger, both fluids unmixed.
T1

t1 t2

T2

1.0

0.9
R=4
Correction factor F

3.0

0.2
2.0

0.8
1.5

0.4
1.

0.

0.
.0

6
0.7

T1 – T2
0.6 R=
t2 – t1

0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
t –t
P= 2 1
T1 – t1

This temperature difference is called the log mean temperature difference (LMTD). Stated
verbally, it is the temperature difference at one end of the heat exchanger less the temperature
difference at the other end of the exchanger divided by the natural logarithm of the ratio of
these two temperature differences. It is left as an exercise for the reader to show that this
relation may also be used to calculate the LMTDs for counterflow conditions.
The above derivation for LMTD involves two important assumptions: (1) the fluid
specific heats do not vary with temperature, and (2) the convection heat-transfer coefficients
are constant throughout the heat exchanger. The second assumption is usually the more
serious one because of entrance effects, fluid viscosity, and thermal-conductivity changes,
etc. Numerical methods must normally be employed to correct for these effects. Section 10-8
describes one way of performing a variable-properties analysis.
If a heat exchanger other than the double-pipe type is used, the heat transfer is cal-
culated by using a correction factor applied to the LMTD for a counterflow double-pipe
arrangement with the same hot and cold fluid temperatures. The heat-transfer equation
then takes the form
q = UAF
Tm [10-13]

Values of the correction factor F according to Reference 4 are plotted in Figures 10-8 to
10-11 for several different types of heat exchangers. When a phase change is involved, as

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.535 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

536 10-5 The Log Mean Temperature Difference

Figure 10-11 Correction-factor plot for single-pass cross-flow exchanger, one fluid mixed,
the other unmixed.
T1

t1 t2

T2
1.0

0.9
Correction factor F

0.8
R=4.0

0.2
3.0

2.0

0.4
0.8
1.5

0.6
1.0
0.7
T –T
R =K1 2
t2 – t1
0.6

0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
t2 – t1
P =K
T1 – t1

in condensation or boiling (evaporation), the fluid normally remains at essentially constant

temperature and the relations are simplified. For this condition, P or R becomes zero and
we obtain
F = 1.0 for boiling or condensation

Examples 10-4 to 10-8 illustrate the use of the LMTD method for calculation of heat-
exchanger performance.
Calculation of Heat-Exchanger Size
EXAMPLE 10-4 from Known Temperatures
Water at the rate of 68 kg/min is heated from 35 to 75◦ C by an oil having a specific heat of
1.9 kJ/kg · ◦ C. The fluids are used in a counterflow double-pipe heat exchanger, and the oil enters
the exchanger at 110◦ C and leaves at 75◦ C. The overall heat-transfer coefficient is 320 W/m2 · ◦ C.
Calculate the heat-exchanger area.
Solution
The total heat transfer is determined from the energy absorbed by the water:
q = ṁw cw
Tw = (68)(4180)(75 − 35) = 11.37 MJ/min [a]
= 189.5 kW [6.47 × 105 Btu/h]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.536 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

C H A P T E R 10 Heat Exchangers 537

Since all the fluid temperatures are known, the LMTD can be calculated by using the temperature
scheme in Figure 10-7b:
(110 − 75) − (75 − 35)

Tm = = 37.44◦ C [b]
ln[(110 − 75)/(75 − 35)]

Then, since q = UA
Tm ,

1.895 × 105
A= = 15.82 m2 [170 ft 2 ]
(320)(37.44)

Shell-and-Tube Heat Exchanger EXAMPLE 10-5

Instead of the double-pipe heat exchanger of Example 10-4, it is desired to use a shell-and-tube
exchanger with the water making one shell pass and the oil making two tube passes. Calculate
the area required for this exchanger, assuming that the overall heat-transfer coefficient remains at
320 W/m2 · ◦ C.
Solution
To solve this problem, we determine a correction factor from Figure 10-8 to be used with the LMTD
calculated on the basis of a counterflow exchanger. The parameters according to the nomenclature
of Figure 10-8 are
T1 = 35◦ C T2 = 75◦ C t1 = 110◦ C t2 = 75◦ C

t −t 75 − 110
P= 2 1 = = 0.467
T1 − t1 35 − 110
T − T2 35 − 75
R= 1 = = 1.143
t2 − t1 75 − 110
so the correction factor is
F = 0.81

and the heat transfer is

q = UAF
Tm

so that
1.895 × 105
A= = 19.53 m2 [210 ft 2 ]
(320)(0.81)(37.44)

Design of Shell-and-Tube Heat Exchanger EXAMPLE 10-6

Water at the rate of 30,000 lbm /h [3.783 kg/s] is heated from 100 to 130◦ F [37.78 to 54.44◦ C] in
a shell-and-tube heat exchanger. On the shell side one pass is used with water as the heating fluid,
15,000 lbm /h [1.892 kg/s], entering the exchanger at 200◦ F [93.33◦ C]. The overall heat-transfer
coefficient is 250 Btu/h · ft 2 · ◦ F [1419 W/m2 · ◦ C], and the average water velocity in the 34 -in
[1.905-cm] diameter tubes is 1.2 ft/s [0.366 m/s]. Because of space limitations, the tube length
must not be longer than 8 ft [2.438 m]. Calculate the number of tube passes, the number of tubes
per pass, and the length of the tubes, consistent with this restriction.
Solution
We first assume one tube pass and check to see if it satisfies the conditions of this problem. The
exit temperature of the hot water is calculated from
q = ṁc cc
Tc = ṁh ch
Th
(30,000)(1)(130 − 100)

Th = = 60◦ F = 33.33◦ C [a]
(15,000)(1)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.537 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

538 10-5 The Log Mean Temperature Difference

so
Th,exit = 93.33 − 33.33 = 60◦ C

The total required heat transfer is obtained from Equation (a) for the cold fluid:

q = (3.783)(4182)(54.44 − 37.78) = 263.6 kW [8.08 × 105 Btu/h]

For a counterflow exchanger, with the required temperature

(93.33 − 54.44) − (60 − 37.78)
LMTD =
Tm = = 29.78◦ C
ln[(93.33 − 54.44)/(60 − 37.78)]
q = UA
Tm
2.636 × 105
A= = 6.238 m2 [67.1 ft 2 ] [b]
(1419)(29.78)

Using the average water velocity in the tubes and the flow rate, we calculate the total flow area with
ṁc = ρAu
3.783
A= = 0.01034 m2 [c]
(1000)(0.366)
This area is the product of the number of tubes and the flow area per tube:

πd 2
0.01034 = n
4
(0.01034)(4)
n= = 36.3
π(0.01905)2

or n = 36 tubes. The surface area per tube per meter of length is

πd = π(0.01905) = 0.0598 m2 /tube · m

We recall that the total surface area required for a one-tube-pass exchanger was calculated in
Equation (b) as 6.238 m2 . We may thus compute the length of tube for this type of exchanger from
nπd L = 6.238
6.238
L= = 2.898 m
(36)(0.0598)

This length is greater than the allowable 2.438 m, so we must use more than one tube pass. When
we increase the number of passes, we correspondingly increase the total surface area required
because of the reduction in LMTD caused by the correction factor F. We next try two tube passes.
From Figure 10-8, F = 0.88, and thus

q 2.636 × 105
Atotal = = = 7.089 m2
UF
Tm (1419)(0.88)(29.78)

The number of tubes per pass is still 36 because of the velocity requirement. For the two-tube-pass
exchanger the total surface area is now related to the length by
Atotal = 2nπd L

so that
7.089
L= = 1.646 m [5.4 ft]
(2)(36)(0.0598)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.538 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

C H A P T E R 10 Heat Exchangers 539

This length is within the 2.438-m requirement, so the final design choice is

Number of tubes per pass = 36

Number of passes = 2
Length of tube per pass = 1.646 m [5.4 ft]

Cross-Flow Exchanger with One Fluid Mixed EXAMPLE 10-7

A heat exchanger like that shown in Figure 10-4 is used to heat an oil in the tubes
Figure Example 10-7
(c = 1.9 kJ/kg · ◦ C) from 15◦ C to 85◦ C. Blowing across the outside of the tubes is steam that
enters at 130◦ C and leaves at 110◦ C with a mass flow of 5.2 kg/sec. The overall heat-transfer T
coefficient is 275 W/m2 · ◦ C and c for steam is 1.86 kJ/kg · ◦ C. Calculate the surface area of the 130
Steam
heat exchanger. 110
Solution
The total heat transfer may be obtained from an energy balance on the steam Oil
85
q = ṁs cs
Ts = (5.2)(1.86)(130 − 110) = 193 kW

We can solve for the area from Equation (10-13). The value of
Tm is calculated as if the exchanger
were counterflow double pipe (i.e., as shown in Figure Example 10-7). Thus, x
(130 − 85) − (110 − 15)

Tm =
 = 66.9◦ C
130 − 85
ln
110 − 15

Now, from Figure 10-11, t1 and t2 will represent the unmixed fluid (the oil) and T1 and T2 will
represent the mixed fluid (the steam) so that
T1 = 130 T2 = 110 t1 = 15 t2 = 85◦ C

and we calculate
130 − 110 85 − 15
R= = 0.286 P= = 0.609
85 − 15 130 − 15
Consulting Figure 10-11 we find
F = 0.97

so the area is calculated from

q 193,000
A= = = 10.82 m2
UF
Tm (275)(0.97)(66.9)

Effects of Off-Design Flow Rates

for Exchanger in Example 10-7 EXAMPLE 10-8
Investigate the heat-transfer performance of the exchanger in Example 10-7 if the oil flow
rate is reduced in half while the steam flow remains the same. Assume U remains constant at
275 W/m2 · ◦ C.
Solution
We did not calculate the oil flow in Example 10-7 but can do so now from
q = ṁo co
To
193
ṁo = = 1.45 kg/s
(1.9)(85 − 15)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.539 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch10 11/3/2008 19:51

540 10-6 Effectiveness-NTU Method

The new flow rate will be half this value or 0.725 kg/s. We are assuming the inlet temperatures
remain the same at 130◦ C for the steam and 15◦ C for the oil. The new relation for the heat transfer is
q = ṁo co (Te,o − 15) = ṁs cps (130 − Te,s ) [a]

but the exit temperatures, Te,o and Te,s are unknown. Furthermore,
Tm is unknown without
these temperatures, as are the values of R and P from Figure 10-11. This means we must use an
iterative procedure to solve for the exit temperatures using Equation (a) and
q = UAF
Tm [b]

The general procedure is to assume values of the exit temperatures until the q’s agree between
Equations (a) and (b).
The objective of this example is to show that an iterative procedure is required when the inlet
and outlet temperatures are not known or easily calculated. There is no need to go through this
iteration because it can be avoided by using the techniques described in Section 10-6.

10-6 EFFECTIVENESS-NTU METHOD

The LMTD approach to heat-exchanger analysis is useful when the inlet and outlet temper-
atures are known or are easily determined. The LMTD is then easily calculated, and the heat
flow, surface area, or overall heat-transfer coefficient may be determined. When the inlet
or exit temperatures are to be evaluated for a given heat exchanger, the analysis frequently
involves an iterative procedure because of the logarithmic function in the LMTD. In these
cases the analysis is performed more easily by utilizing a method based on the effectiveness
of the heat exchanger in transferring a given amount of heat. The effectiveness method also
offers many advantages for analysis of problems in which a comparison between various
types of heat exchangers must be made for purposes of selecting the type best suited to
accomplish a particular heat-transfer objective.
We define the heat-exchanger effectiveness as
actual heat transfer
Effectiveness =  =
maximum possible heat transfer
The actual heat transfer may be computed by calculating either the energy lost by the hot
fluid or the energy gained by the cold fluid. Consider the parallel-flow and counterflow heat
exchangers shown in Figure 10-7. For the parallel-flow exchanger
q = ṁh ch (Th1 − Th2 ) = ṁc cc (Tc2 − Tc1 ) [10-14]

and for the counterflow exchanger

q = ṁh ch (Th1 − Th2 ) = ṁc cc (Tc1 − Tc2 ) [10-15]

To determine the maximum possible heat transfer for the exchanger, we first recognize that
this maximum value could be attained if one of the fluids were to undergo a temperature
change equal to the maximum temperature difference present in the exchanger, which is
the difference in the entering temperatures for the hot and cold fluids. The fluid that might
undergo this maximum temperature difference is the one having the minimum value of
ṁc because the energy balance requires that the energy received by one fluid be equal to
that given up by the other fluid; if we let the fluid with the larger value of ṁc go through
the maximum temperature difference, this would require that the other fluid undergo a
temperature difference greater than the maximum, and this is impossible. So, maximum
possible heat transfer is expressed as
qmax = (ṁc)min (Thinlet − Tcinlet ) [10-16]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.540 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services