3/11/2023

CONDENSERS AND EVAPORATORS

CONDENSERS AND EVAPORATORS AS HEAT EXCHANGERS

One classification of condensers and Table 1. Some types of evaporators and condensers.
evaporators (Table 1) is according to
whether the refrigerant is on the
inside or outside of the tubes and
whether the fluid cooling the
condenser or being refrigerated is a
gas or a liquid. The gas referred to in
Table 1 is usually air, and the liquid
is usually water, but other substances
are used as well.

1
 3/11/2023

CONDENSERS AND EVAPORATORS AS HEAT EXCHANGERS

The most widely used types of condensers

and evaporators are shell-and-tube heat
exchangers (Figure 1) and finned-coil heat
exchangers (Figure 2). Table 1 indicates
that certain combinations are not
frequently used, particularly the
configuration where the gas is passed
through tubes. The reason is that volume
flow rates of gases are high relative to
Figure 1. Shell-and-tube water-cooled condenser. (ITT Bell & Gossett.)
those of liquids and would result in high
pressure drops if forced through the tubes.

Figure 2. An air-cooled condenser. (Kramer)

OVERALL HEAT-TRANSFER COEFFICIENT.

The overall heat-transfer coefficient for an
evaporator or condenser is the
proportionality constant, which, when
multiplied by heat-transfer area and the
mean temperature difference between the
fluids, yields the rate of heat transfer. If
heat flows across a tube, as in Figure 3,
between refrigerant on the outside and
water on the inside, for example, under
steady-state conditions the rate of heat
transfer q in watts is the same from the
refrigerant to the outside surface of the
tube to the water.
Figure 3. Shell flow of liquid across tube bundles.

2
 3/11/2023

OVERALL HEAT-TRANSFER COEFFICIENT

The expressions for q in each of these tos = temperature of outside surface of tube, C
transfers are, respectively, k = conductivity of tube metal, W/m-K
q = ho Ao (t o − t os ) (1) x = thickness of tube, m
tis = temperature of inside surface of tube, C
k
q= Am (t os − t is ) (2) Am = mean circumferential area of tube, m2
x
hi = heat-transfer coefficient on inside of tube,
q = hi Ai (t is − t i ) (3) W/m2-K
Ai = inside area of tube, m2
where ti = water temperature, C
q = rate of heat transfer, W
ho = heat-transfer coefficient on outside of
tube, W/m2-K
Ao = outside area of tube, m2
to = refrigerant temperature, C

OVERALL HEAT-TRANSFER COEFFICIENT

To express the overall heat-transfer To compute the U value from knowledge of
coefficient the area on which the coefficient the individual heat-transfer coefficients,
is based must be specified. Two acceptable first divide Equation 1 by hoAo, Equation 2
expressions for the overall heat-transfer by kAm/x, and Equation 3 by hiAi, leaving
coefficient are only the temperature differences on the
q = Uo Ao (t o − t i ) (4) right sides of the equations. Next add the
three equations, giving
and
q = Ui Ai (t o − t i ) (5)
q qx q
+ + = (t o − t os ) + (t os − t is ) + (t is − t i )
ho Ao kAm hi Ai
Where
Uo = overall heat-transfer coefficient based
on outside area, W/m2-K q qx q
+ + = to − t i (6)
Ui = overall heat-transfer coefficient based ho Ao kAm hi Ai
on inside area, W/m2-K

3
 3/11/2023

OVERALL HEAT-TRANSFER COEFFICIENT

Alternate expressions for to – ti are The physical interpretation of the terms in
available from Equation 4 and 5. Equation 8 is that 1/UoAo and 1/UiAi are the
total resistances to heat transfer between
q q the refrigerant and water. This total
to − ti = = (7)
Uo Ao Ui Ai resistance is the sum of the individual
resistances
Equating Equation 6 and 7 and canceling q a. From the refrigerant to the outside
provides an expression for computing the U surface of the tube 1/hoAo.
values
b. Through the tube x/(kAm)
1 1 1 x 1 c. From the inside surface of the tube to
= = + + (8) the water 1/hiAi
Uo Ao Ui Ai ho Ao kAm hi Ai

LIQUID IN TUBES; HEAT TRANSFER AND PRESSURE DROP.

The expression for the heat-transfer where
coefficient for fluids flowing inside tubes is h = convection coefficient, W/m2-K
of the form
D = ID of tube, m
Nu = C Ren Pr m
k = thermal conductivity of fluid, W/m-K
where n and m are exponents. The
constant C and exponents in the equation V = mean velocity of fluid, m/s
are ρ = density of fluid, kg/m3
0.8 0. 4 µ = viscosity of fluid, Pa.s
hD  VDρ   cp µ 
= 0.023    (9) cp = specific heat of fluid, J/kg.K
k  µ   k 

Equation 9 is applicable to turbulent flow,

which typically prevails with the velocities
and fluid properties experienced in most
commercial evaporators and condensers.

4
 3/11/2023

LIQUID IN TUBES; HEAT TRANSFER AND PRESSURE DROP

Example No. 1 Reynolds numbers:
Compute the heat-transfer coefficient for
water flow inside the tubes (8 mm ID) of VDρ (2 m s )(0.008 m )(1000 kg m3 ) = 12,214
an evaporator if the water temperature is Re = =
µ 0.00131 Pa ⋅ s
10 C and its velocity is 2 m/s.
Given: The flow is turbulent, so Equation 9
D = 8 mm = 0.008 m applies.
Water temperature = 10 C Prandtl number:
V = 2 m/s
Required: Pr =
(4190 J kg ⋅ K )(0.00131 Pa ⋅ s ) = 9.6
0.573 W m ⋅ K
h = heat-transfer coefficient
Equation 9:
Solution:
The properties of water at 10 C are Nu = 0.023(12,214 )0.8 (9.6)0.4 = 106
µ = 0.00131 Pa.s
The heat-transfer coefficient:
ρ = 1000 kg/m3
k = 0.573 W/m-K 0.573 W m ⋅ K
h= (106) = 7592 W m2 ⋅ K
cp = 4190 J/kg-K 0.008 m 9

LIQUID IN TUBES; HEAT TRANSFER AND PRESSURE DROP

As the fluid flow inside the tubes through a Since the pressure drop in the straight
condenser or evaporator, a pressure drop tubes in an evaporator or condenser may
occur both in the straight tubes in the U- represent only 50 to 80 percent of the total
bends or heads of the heat exchanger. Some pressure drop, experimental or catalog data
drop in pressure is also attributable to on the pressure drop as a function of flow
entrance and exit losses. The expression for rate are desirable. If the pressure drop at
pressure drop of fluid flowing in straight one flow rate is known, it is possible to
tubes is predict the pressure drop at other flow
rates. The expression applicable to straight
L V2
∆p = f ρ (10) tubes, Equation 10 indicates that the
D 2 pressure drop is proportional to the square
of the velocity and thus the square of the
where flow rate.
∆p = pressure drop, Pa
f = friction factor, dimensionless
L = length of tube, m

10

5
 3/11/2023

LIQUID IN TUBES; HEAT TRANSFER AND PRESSURE DROP

The other contributors to pressure drop
resulting from changes in flow area and
direction are also almost exactly
proportional to the square of the flow rate,
so if the pressure drop and flow rate ∆p1
and w1 are known, the pressure drop ∆p2 at
a different flow rate w2 can be predicted:
2
w 
∆p2 = ∆p1  2  (11)
 w1 

11

LIQUID IN SHELL; HEAT TRANSFER AND PRESSURE DROP

In shell-and-tube evaporators, where The analytical prediction of the heat-
refrigerant boils inside tubes, the liquid transfer coefficient of liquid flowing normal
being cooled flows in the shell across to a tube is complicated in itself, and the
bundles of tubes, as shown schematically complex flow pattern over a bundle of tubes
in Figure 3. The liquid is directed by makes the prediction even more difficult. In
baffles so that it flows across the tube order to proceed with the business of
bundle many times and does not short- designing heat exchangers, engineers
circuit from the inlet to the outlet. resort to correlations that relate the
Nusselt, Reynolds, and Prandtl numbers to
the geometric configuration of the tubes
and baffles.

Figure 3. Shell flow of liquid across tube bundles.

12

6
 3/11/2023

LIQUID IN SHELL; HEAT TRANSFER AND PRESSURE DROP

Such an equation by Emerson can be modified to the form

0.14
hD  µ 
= (terms controlled by geometry )Re0.6 Pr 0.3   (12)
k  µw 
Where
µ = viscosity of fluid at bulk temperature, Pa.s
µw = viscosity of fluid at tube-wall temperature, Pa.s

The Reynolds number in this equation is GD/µ, where G is

the mass velocity or mass rate of flow divided by a
characteristic flow area.

13

LIQUID IN SHELL; HEAT TRANSFER AND PRESSURE DROP

One important but simple realization emerges from Equation 12:
for a given evaporator or condenser when water flows in the shell
outside the tubes

Water − side heat − transfer coefficien t = (const )( flow rate )0.6 (13)

The convection coefficient varies as the 0.6 power of the flow rate
compared with the 0.8 power for flow inside tubes, as indicated in
Equation 9.

14

7
 3/11/2023

LIQUID IN SHELL; HEAT TRANSFER AND PRESSURE DROP

The pressure drop of liquid flowing through the
shell across tube bundles is also difficult to
predict analytically, but when an experimental
point is available for one flow rate, predictions of
the pressure drop at other flow rates can be
made quite accurately. Figure 4 shows the water
pressure drop taken from catalog data of a
water-chilling evaporator. The applicable
exponent in the pressure-drop-flow-rate
relationship here is 1.9.

Figure 4. Pressure drop of water flowing in the

shell of an evaporator. (York Division of Borg
15
Warner.)

EXTENDED SURFACE; FINS.

When one of the fluids in a condenser or
evaporator is a gas (hereafter considered to
be air), the properties of the air compared
with those of the liquid, such as water,
result in heat-transfer coefficients of the
order of one-tenth to one-twentieth that of
the water. The air-side resistance in a
configuration such as shown in Figure 2
would provide the controlling resistance. In
order to decrease 1/hA, the area A is
usually increased by using fins.
Figure 5. Bar fin.
The bar fin, shown in Figure 5 is a
rudimentary fin whose performance can
be predicted analytically and will be
used to illustrate some important
characteristics.
16

8
 3/11/2023

EXTENDED SURFACE; FINS

The fins are of length L and thickness 2y
m. The conductivity of the metal is k
W/m.K, and the air-side coefficient is hf
W/m2.K. To solve for the temperature
distribution through the fin, a heat balance
can be written about an element of
thickness dx m. The heat balance states
that the rate of heat flow entering the
element at position 1 from the end of the
fin plus that transferred to the element
from the air equals the rate of heat
transferred out of the element at position 2
toward the base. Figure 5. Bar fin.

17

EXTENDED SURFACE; FINS

For one-half a fin width and a fin depth of Z For the differential length dx the change in
m, the heat balance in symbols is the temperature gradient is

 dt   dt  (14)  dt   dt  d  dt  d 2t
kyZ   + Zdxh f (t a − t ) = kyZ     −   =  dx = 2 dx (16)
 1
dx  dx 2  dx 1  dx 2 dx  dx  dx

where Substituting into Equation 15, we get

ta = temperature of air
d 2t h f (t a − t )
t = temperature of fin = (17)
dx 2 ky
Cancelling Z and factoring gives

 dt   dt  
ky   −    = dxh f (t a − t ) (15)
 dx 2  dx 1 

18

9
 3/11/2023

EXTENDED SURFACE; FINS

By solving the second-order differential When a finned coil cools air, points in the
equation 17 the temperature distribution fin farther away from the base are higher
throughout the fin can be shown to be in temperature than points close to the
base. The net result of the higher
temperature of most of the fin is that less
t − t b cosh M (L − x ) heat is transferred than if the entire fin
= (18)
t a −t b cosh ML were at temperature tb. The ratio of the
actual rate of heat transfer to that which
where would be transferred if the fin were at
tb = temperature of base of fin, C temperature tb is called the fin
effectiveness.
hf
M=
ky Fin effectiveness = η

actual q (19)
η=
q if fin were at base temperature

19

EXTENDED SURFACE; FINS

Harper and Brown found that the fin
effectiveness for the bar fin at Figure 5 can
be represented by

tanh ML
η=
ML

The bar fin is not a common shape but the

dominant type of finned surface is the
rectangular plate fin mounted on
cylindrical tubes. The net result is a Figure 5. Bar fin.
rectangular or square fin mounted on a
circular base, one section of which is shown
in Figure 6a.

20

Figure 6. Determining fin effectiveness of a rectangular plate fin (a) by

treating it as an (b) annular fin of the same area

10
 3/11/2023

EXTENDED SURFACE; FINS

The fin effectiveness of the rectangular
plate fins is often calculated by using
properties of the corresponding annular
(Figure 6b), for which a graph of the fin
effectiveness is available, as in Figure 7.
The corresponding annular fin has the
same area and thickness as the plate fin it
represents.

Figure 7. Fin effectiveness of an annular fin. The external

radius of the fin is re m and the internal radius is ri m.

Figure 6. Determining fin effectiveness of a rectangular plate fin (a) by

treating it as an (b) annular fin of the same area
21

EXTENDED SURFACE; FINS

Example No. 2
What is the fin effectiveness of a rectangular
plate fin made of aluminum 0.3 mm thick
mounted on a 16-mm-OD tube if the vertical
tube spacing is 50 mm and the horizontal
spacing is 40 mm? The air-side heat-transfer
coefficient is 65 W/m2-K, and the conductivity
of aluminum is 202 W/m-K.
Figure 8. Annular fin of same area as rectangular plate fin.
Given:
2y = 0.3 mm = 0.00030 m or y = 0.00015 m The half-thickness of the fin y = 0.15 mm =
0.00015 m.
16-mm-OD tube, vertical tube spacing is 50
mm, horizontal spacing is 40 mm. 65
M= = 46.3 m −1
hf = 65 W/m2.K, k = 202 W/m.K. 202(0.00015)
Required: (re − ri )M = (0.0252 − 0.008 )(46.3) = 0.8
Fin effectiveness, η
Solution: From Figure 7 for (re – r i)M = 0.8 and re/r i =
25.2/8 = 3.15 the fin effectiveness η is 0.72.
The annular fin having the same area as the
plate fin (Figure 8) has an external radius of 22
25.2 mm.

11
 3/11/2023

EXTENDED SURFACE; FINS

23

EXTENDED SURFACE; FINS

The air-side area of a finned condenser or evaporator is
composed of two portions, the prime area and the extended
area Ae is that of the fin. Since the prime area is at the base
temperature, it has a fin effectiveness of 1.0. It is to the
extended surface that the fin effectiveness less than 1.0
applies. Equation (8) for the overall heat-transfer coefficient
can be revised to read
1 1 1 x 1
= = + + (20)
Uo Ao Ui Ai h f (Ap + ηAe ) kAm hi Ai

24

12
 3/11/2023

GAS FLOWING OVER FINNED TUBES; HEAT

TRANSFER AND PRESSURE DROP
A precise prediction of the air-side heat- The drop in pressure of the air flowing
transfer coefficient when the air flows over through a finned coil is also dependent
finned tubes is complicated because the upon the geometry of the coil. Figure 9
value is a function of geometric factors, e.g., shows the pressure drop of a commercial
the fin spacing, the spacing and diameter of cooling coil when the finned surfaces are
the tubes, and the number of rows of tubes dry. As expected, the pressure drop is
deep. Usually the coefficient varies higher for coils with a large number of fins
approximately as the square root of the per meter of tube length. The ordinate is
face velocity of the air. A rough estimate of the pressure drop per number of rows of
the air-side coefficient hf can be computed tubes deep, so the values would be
from the equation derived from illustrative multiplied by 6 for a six-row coil, for
data in the ARI standard example.

h f = 38V 0.5 (21)

where V is the face velocity in meter per

second.
25

GAS FLOWING OVER FINNED TUBES; HEAT

TRANSFER AND PRESSURE DROP
For the coil series whose pressure drops are
shown in Figure 9 the pressure drop for a
given coil varies as the face velocity to the
1.56 power. The exponent is fairly typical of
commercial plate-fin coils.

Figure 9. Pressure drop of air flowing through a finned

coil (Bohn Heat Transfer Division of Gulf & Western 26
Manufacturing Company.)

13
 3/11/2023

CONDENSERS
For the condenser the fluid to which heat is
rejected is usually either air or water.
Another type of water-cooled condenser has
cleanable tubes (Figure 10). When the
condenser is water-cooled, the water is sent
to a cooling tower for ultimate rejection of
the heat to the atmosphere.

Some years ago air-cooled condensers were

used only in small refrigeration systems
(less than 100 kW refrigerating capacity), Figure 10. Water-cooled condenser with cleanable tubes. (Halstead
but now individual air-cooled condensers and Mitchell, a Division of Halstead Industries, Inc.)
are manufactured in sizes matching
refrigeration capacities of hundreds of
kilowatts. The water-cooled condenser is
favored over the air-cooled condenser
where there is a long distance between the
compressor and the point where heat is to
be rejected. 27

CONDENSERS
Most designers prefer to convey water
rather than refrigerant in long lines. In
centrifugal-compressor systems large pipes
are needed for the low-density refrigerants,
so that the compressor is close-coupled to
the condenser. Water-cooled condensers
therefore predominate in centrifugal-
compressor systems.

Figure 10. Water-cooled condenser with cleanable tubes. (Halstead

and Mitchell, a Division of Halstead Industries, Inc.)

28

14
 3/11/2023

REQUIRED CONDENSING CAPACITY

The required rate of heat transfer in the condenser is
predominately a function of the refrigerating capacity
and the temperatures of evaporation and condensation.
The condenser must reject both the energy absorbed by
the evaporator and the heat of compression added by
the compressor. A term often used to relate the rate of
heat flow at the condenser to that of the evaporator is
the heat-rejection ratio

rate of heat rejected at condenser , kW

Heat − rejection ratio =
rate of heat absorbed at evaporator , kW

29

REQUIRED CONDENSING CAPACITY

Theoretical calculations of the condenser
heat rejection can be made from the
standard vapor-compression cycle, but they
do not take into consideration the
additional heat added by inefficiencies in
the compressor. A graph of typical values of
heat-rejection ratios is shown in Figure 11.

When the motor driving the compressor is

hermetically sealed, some of the heat
associated with inefficiencies of the electric
motor is added to the refrigerant stream
and must ultimately be removed at the
condenser. The heat-rejection ratios of the
hermetically sealed compressors are
usually slightly higher than those of the
open-type compressor.

30
Figure 11. Typical values of the ratio of the heat rejected at the
condenser to the refrigerating capacity for refrigerants 12 and 22.

15
 3/11/2023

CONDENSING COEFFICIENT.
The basic equation for calculating the local The equation for the local condensing
coefficient of heat transfer of vapor coefficient is
condensing on a vertical plate (Figure 12) 14
hcv x  gρ h fg x 
2 3
was developed by Nusselt by pure physical = (22)
analysis.  
where
k  4 µk∆t 
hcv = local condensing coefficient on vertical
plate, W/m2.K
x = vertical distance measured from top of
plate, m
g = acceleration due to gravity = 9.81 m/s2
ρ = density of condensate, kg/m3
hfg = latent heat of vaporization, J/kg
µ = viscosity of condensate, Pa.s
∆t = temperature difference between vapor
and the plate, K
31
Figure 12. Condensation on a vertical plate.

CONDENSING COEFFICIENT
The mean condensing coefficient over the Several investigators have found that
total height of the plate L is the constant 0.725 in Equation 24
L 14
agrees closely with experimental
hcv =
0 hcv dx = 0.943 gρ 2h fg k 3 
W m2 ⋅ K (23)
results. White found the constant to be
L  µ∆tL  0.63, and Goto et al. found it to be
 
The equation for the mean condensing approximately 0.65.
coefficient for vapor condensing on the
outside of horizontal tubes is
14
 gρ 2h fg k 3 
hct = 0.725  W m2 ⋅ K (24)
 µ∆tND 
 
Where
N = number of tubes in vertical row
D = OD of tube, m

32

16
 3/11/2023

FOULING FACTOR.
After a water-cooled condenser has been in Several different agencies have
service for some time its U value usually established standards for the fouling
degrades somewhat because of the increased factors to be used. One trade association
resistance to heat transfer on the water side specifies 0.000176 m2.K.W, which means
due to fouling by the impurities in the water that the condenser should leave the
from the cooling tower. The new condenser factory with a 1/Uo value 0.000176 Ao/Ai
must therefore have a higher U value in less than the minimum required to meet
anticipation of the reduction that will occur the quoted capacity of the condenser.
in service. The higher capacity with new
equipment is provided by specifying a
fouling factor 1/hff m2.K/W. This term
expands Equation 8 for the U value into

1 1 xA A A
= + o + o + o (25)
Uo ho kAm h ff Ai hi Ai

33

DESUPERHEATING.
Even when the refrigerant condenses at a
constant pressure, its temperature is
constant only in the condensing portion.
Because the vapor coming from the
compressor is usually superheated, the
distribution of temperature will be as
shown in Figure 13.

Figure 24. Centrifugal-compressor characteristics at various setting of

the prerotation vanes.

34

17
 3/11/2023

DESUPERHEATING
Because of the distortion in the It is common practice to use Equation 26
temperature profile caused by the anyway with the following justification.
desuperheating process, the temperature Although the temperature difference
difference between the refrigerant and the between the refrigerant and cooling fluid
cooling fluid is no longer correctly is higher in the desuperheating section
represented by the LMTD than calculated from Equation 26, the
convection coefficient in this section is
normally lower than the condensing
LMTD =
(tc − ti ) − (t c − t o ) coefficient. The two errors compensate
(26) somewhat for each other, and the
ln[(t c − t i ) (t c − t o )]
application of Equation 26 along with the
condensing coefficient over the entire
condenser area usually provides
reasonably accurate results.

35

Given:
Refrigerant 22
80 kW refrigerating capacity
CONDENSER DESIGN. 5 C evaporating temperature
45 C condensing temperature
An example will illustrate how some of the
principles described in the previous 30 C to 35 C water
sections are combined in designing a Two pass condenser with 42 tubes as
condenser. shown.
Example No. 3 ID = 14 mm
The condensing area is to be specified for a OD = 16 mm
refrigerant 22 condenser of a refrigerating
system that provides a capacity of 80 kW
for air conditioning. The evaporating
temperature is 5 C and the condensing
temperature is 45 C at design conditions.
Water from a cooling tower enters the
condenser at 30 C and leaves at 35 C.
A two-pass condenser with 42 tubes,
arranged as shown in Figure 154 will be
used, and the length of tubes is to be
specified to provide the necessary area. The
tubes are copper and are 14 mm ID and 16
36
mm OD.
Figure 14. Tube arrangement of condenser

18
 3/11/2023

CONDENSER DESIGN.
Required: Condensing coefficient:
Area and tube length From Equation 24.
14
Solution:  gρ 2 h fg k 3 
hcond = 0.725 
Calculate the required rate of heat  µ∆tND 
 
transfer, the individual heat-transfer The density ρ and latent heat of
coefficients, and the overall heat-transfer vaporization hfg at 45 C are available from
coefficient; then compute the required area Tables.
and tube length.
Rate of heat transfer: 1
ρ= = 1.109 kg L = 1109 kg m3
From Figure 11, assuming the compressor 0.90203 L kg
to be a hermetic one, the heat-rejection h fg = 160,900 J kg
ratio at a condensing temperature of 45 C
and an evaporating temperature of 5 C is The conductivity k and viscosity µ of the
1.27. The rate of heat rejected at the liquid refrigerant at 45 C are available
condenser q is from Tables.
q = (80 kW )(1.27) = 101.6 kW k = 0.0779 W m ⋅ K
µ = 0.000180 Pa ⋅ s 37

CONDENSER DESIGN.

38

19
 3/11/2023

CONDENSER DESIGN.
The average number of tubes in a vertical Resistance of metal:
row N is The conductivity of copper is 390 W/m.K,
2 +3+ 4 + 3+ 4 + 3+ 4 +3 + 4 + 3+ 4 +3+ 2 and the resistance of the tube is
N= = 3.23
13
xAo (0.016 − 0.014 ) 2 16
=
The temperature difference between the kAm 390 (14 + 16 ) 2
vapor and the tube is unknown at this point; xAo
therefore ∆t will be assumed to be 5 K and = 0.000002735 m2 ⋅ K W
the value adjusted later if necessary. kAm
a value that will prove to be negligible in
3 14
comparison to the other resistances.
 9.81(1109 ) (160,900 )(0.0779 ) 
2
hcond = 0.725  
 0.000180(5)(3.23)(0.016 ) 
Fouling factor:
2 From Sec. 10
hcond = 1528 W m ⋅ K
1
= 0.000176 m 2 ⋅ K W
h ff
39

CONDENSER DESIGN.
Water-side coefficient: Equation 9 can be used to calculate the
The flow rate of water needed to carry the water-side heat-transfer coefficient hw
heat away from the condenser with a using the water properties at 32 C
temperature rise from 30 to 35 C is
ρ = 995 kg m 3
101.6 kW
= 4.85 kg s
(4.19 kJ kg ⋅ K )(35.0 − 30.0 K ) c p = 4190 J kg ⋅ K
and the volume flow rate is µ = 0.000773 Pa⋅ s
4.85 kg s k = 0.617 W m ⋅ K
= 0.00485 m3 s
1000 kg m3
The water velocity through the tubes V is 0.8 0. 4
hD  VDρ   cp µ 
= 0.023   
0.00485 m s 3  k 
V= = 1.5 m s
k  µ   
(21 tubes pass)(π 4)(0.014 m )2 0.8
0.617(0.023) 1.5(0.014)(995)   4190(0.000773) 
0. 4
hw =  0.000773   
0.014 0.617

hw = 1.014(27,030 )0.8 (5.25)0.4

40
hw = 6910 W m2 ⋅ K

20
 3/11/2023

CONDENSER DESIGN.
Required area:
1 1 xA A A
= + o + o + o
Uo ho kAm h ff Ai hi Ai
1 1 0.016
= + 0.000002735 + (0.000176) + 0.016 1
Uo 1528 0.014 0.014 6910
1
= 0.001023
Uo

Uo = 977 W m 2 ⋅ K

41

CONDENSER DESIGN.
The LMTD is A recheck should now be made of the
assumption of the 5-K temperature
LMTD =
(t c − t i ) − (t c − t o ) difference used in calculating the
ln[(t c − t i ) (t c − t o )] condensing coefficient. A recalculation may
(45 − 30) − (45 − 35) = 12.33 C be necessary.
LMTD =
ln
(45 − 30) The designer would also check the water-
(45 − 35) pressure drop in the condenser to see that
it does not exceed a reasonable value
q 101,600 W (perhaps of the order of 70 kPa).
Ao = = = 8.43 m 2
( )
Uo LMTD 977 W m 2 ⋅ K (12.33 K )
Length of tubes:
8.43 m 2
L= = 4.0 m
(42 tubes)(0.016π )

42

21
 3/11/2023

WILSON PLOTS.
Constructing a Wilson plot is a technique of
processing heat-transfer data to determine
the individual heat-transfer coefficients in
a heat exchanger. The concept was
introduced by Wilson and is often applied
to condensers and evaporators to determine
the condensing or evaporating heat-
transfer coefficient along with the air- or
water-side coefficient.
If it is a water-cooled condenser that is
being analyzed, for example, a series of
heat-transfer tests is run and the U value
determined for various flow rates of cooling
water. If the condenser tubes are clean, Figure 15. A Wilson plot to determine the individual heat-transfer
Equation 8 applies and ho is the coefficients of a condenser
condensing-side coefficient and hi the
water-side coefficient.
1 1 xA A
= + o + o 43
Uo ho kAm hi Ai

WILSON PLOTS.
The properties of the cooling water are
primarily a function of temperature, and if
the temperature range throughout the tests
is not large, the properties may be assumed
constant. Equation 9 can then be simplified
to

( )
hi = (const ) V 0.8 (28)

The Wilson plot for this heat exchanger is a

graph of 1/Uo versus 1/V0.8, as shown in
Figure 15.
Figure 15. A Wilson plot to determine the individual heat-transfer
coefficients of a condenser

44

22
 3/11/2023

AIR AND NONCONDENSABLES.

If air or other non-condensable gases enter Noncondensables can be removed from the
the refrigeration system, they will condenser by purging. The purging
ultimately collect in the condenser, where operation consists of drawing a mixture of
foreign gases reduce the efficiency of the refrigerant vapor and noncondensables
system for two reasons: (1) The total from the condenser, separating the
pressure in the condenser is elevated, refrigerant, and discharging the non-
which requires more power for the condensables. Ammonia systems are
compressor per unit refrigeration capacity commonly equipped with purgers, but the
The condenser pressure is raised over the only other systems equipped with purgers
saturation pressure of the refrigerant by are the centrifugal-compressor systems,
the amount of the partial pressure of the which use low-pressure refrigerants such
non-condensable gas. (2) Instead of as refrigerants 11 and 113.
diffusing throughout the condenser, the
non-condensables cling to the condenser
tubes. Thus the condensing-surface area is
reduced, which also tends to raise the
condensing pressure.

45

EVAPORATORS.
In most refrigerating evaporators the
refrigerant boils in the tubes and cools the
fluid that passes over the outside of the
tubes. Evaporators that boils refrigerant in
the tubes are often called direct-expansion
evaporators, and Figure 16 shows an air-
cooling evaporator and Figure 18 a liquid
cooler. The tubes in the liquid chiller in
Figure 16. Air-cooling evaporator. The device on the left end is a
Figure 17 have fins inside the tubes in refrigerant distributor to feed the several circuits uniformly. (Super
order to increase the conductance on the Radiator Coils.)
refrigerant side.

46
Figure 17. A liquid chilling evaporator in which refrigerant boils
inside finned tubes. (Dunham-Bush, Inc.)

23
 3/11/2023

EVAPORATORS
Direct-expansion evaporators used for air-
conditioning applications are usually fed by
an expansion valve that regulates the flow
of liquid so that the refrigerant vapor
leaves the evaporator with some superheat,
as shown in Figure 18a. Another concept is
the liquid-recirculation or liquid-overfeed
evaporator in Figure 18b, in which excess
liquid at low pressure and temperature is
pumped to the evaporator.

Figure 18. (a) Air-conditioning evaporator with refrigerant

leaving in a superheated state, (b) liquid-recirculation
evaporator with liquid refrigerant carried out of the evaporator.

47

BOILING IN THE SHELL.

The classic prediction for the heat-transfer
coefficient for pool boiling of water at
atmospheric pressure is shown in Figure
19. The tests were conducted by immersing
a heated wire in a container of water. The
tests were conducted by immersing a
heated wire in a container of water.
In the boiling regime AB the boiling is
called nucleate boiling, where bubbles form
on the surface and rise through the pool.
The equation of the curve is approximately
q
= C∆t 3 to 4
where A Figure 19. Heat-transfer coefficient for pool boiling of water. (From
q = rate of heat transfer, W W.H. McAdams, “Heat Transmission,” 2d ed., p. 370, McGraw-Hill,
New York, 1954.)
A = heat-transfer area, m2
C = constant
∆t = difference in temperature between 48
metal surface and boiling fluid, K

24
 3/11/2023

BOILING IN THE SHELL.

To write the equation in another form
divide both sides by ∆t,

q
= hr = C∆t 2 to 3
A∆t
where hr is the boiling coefficient,
W/(m2.K). The value of hr increases as the
temperature difference increases, which
physically is due to the greater agitation.
The disturbance frees the bubbles of vapor
from the metal surface sooner and allows
the liquid to come into contact with the
metal.

Figure 19. Heat-transfer coefficient for pool boiling of water. (From

W.H. McAdams, “Heat Transmission,” 2d ed., p. 370, McGraw-Hill,
New York, 1954.)

49

BOILING IN THE SHELL

The rate of evaporation can increase to a
peak, point B, where so much vapor covers
the metal surface that the liquid can no
longer intimately contact the metal. A
further increase in the temperature
difference decreases the rate of heat
transfer.
The graph in Figure 19 is useful in
predicting the trends for heat-transfer
coefficients for boiling outside tube
bundles. Hoffmann summarized the work
of several investigators to provide the band
shown in Figure 20.

Figure 20. Heat-transfer coefficients for refrigerants 12

and 22 boiling outside of tube bundles.
50

25
 3/11/2023

BOILING INSIDE TUBES.

When refrigerant boils inside the tubes, the
heat-transfer coefficient changes
progressively as the refrigerant flows
through the tube. The refrigerant enters
the evaporator tube with a low fraction of
vapor. As the refrigerant proceeds through
the tube, the fraction of vapor increases,
intensifying the agitation and increasing
the heat-transfer coefficient. When the
refrigerant is nearly all vaporized, the
coefficient drops off to the magnitude
applicable to vapor transferring heat by
forced convection.

Figure 21. Heat-transfer coefficients of refrigerant 22

51
boiling inside tubes. Curve 1 at 10oC, curve 2 at 3oC,
and curve 3 at 2.8oC temperatures of evaporation.

BOILING INSIDE TUBES.

Figure 21 shows local coefficients
throughout a tube for three different
levels of temperature. The heat-transfer
coefficient is highest for the high
evaporating temperature, probable
because at high evaporating
temperatures and pressures the vapor
density is high, permitting a greater
fraction of the metal to be wetted with
liquid.

Figure 21. Heat-transfer coefficients of refrigerant 22

52
boiling inside tubes. Curve 1 at 10oC, curve 2 at 3oC,
and curve 3 at 2.8oC temperatures of evaporation.

26
 3/11/2023

PRESSURE DROP IN TUBES.

The pressure of the refrigerant drops as it flows
through tube-type evaporators. The effect of pressure
drop on system performance is that the compressor
must pump from a lower suction pressure, which
increases the power requirement. On the other hand
a high refrigerant velocity can be achieved if more
pressure drop is permitted, and this high velocity
improves the heat-transfer coefficient. Typical
pressure drops for air-conditioning evaporators are
15 to 30 kPa.

53

FROST.
When the surface temperatures of an air- Numerous methods of defrosting are
cooling evaporator fall below 0 C frost will available, and probably the most popular
form. Frost is detrimental to the operation ones are hot-gas defrost and water defrost.
of the refrigeration system for two reasons: In hot-gas defrost, discharge gas from the
(1) thick layers of frost act as insulation, compressor is sent directly to the
and (2) in forced-convection coils the frost evaporator and the evaporator performs
reduces the airflow rate. With a reduced temporarily as a condenser. The heat of
airflow rate the U value of the coil drops, condensation melts off the frost, which
and the mean temperature difference drains away. In water defrost, a stream of
between the air and refrigerant must water is directed over the coil until all the
increase in order to transfer the same rate frost is melted.
of heat flow. Both these factors penalize the
system by requiring a lower evaporating
temperature.

54

27
 3/11/2023

END

55

28