Chapter 6

Design of Air-Conditioning
Apparatus

6.1 Introduction

Air-conditioning apparatuses have been generally classified into two major types.
viz., coil equipment and washer or spray equipment.

In the coil equipment. air comes in contact with a surface and not the conditioning
medium which may be either a refrigerant or water (chilled or heated). brine, steam,
etc.
In the washer equipment, air comes in direct contact with the conditioning water
(chilled, recirculated or heated) or hygroscopic fluids.

In either case the air comes in contact with a surface; which is wet in the case of
the washer or spray equipment, and dry in the case of the coil equipment. However,
in the case of the coil equipment also, if the temperature of the surface is below the
dew point temperature of air, condensation of moisture from the air on to the surface
will take place, and the surface eventually becomes wet.

6.2 Heat and Moisture Transfer in Air-Conditioning Apparatus

When heat is transferred between unsaturated air and a dry surface, the driving
force is the difference in the dry-bulb temperature of the air and the temperature of
the surface. This driving force is also termed temperature potential and the heat
transfer is termed sensible heat transfer.

But when heat is transferred between unsaturated air and a wetted surface, another
driving force other than temperature difference results. This driving force is the
difference in the vapour pressure in the unsaturated air and the saturation vapour
pressure at the temperature of the wetted surface. The force causes a transfer of
moisture resulting either in its condensation or evaporation. This driving force may
be termed vapour pressure potential and the accompanying heat transfer as latent heat
transfer.

287
 When both the driving forces are present, we find that both sensible and latent
heats are transferred resulting in total heat transfer, such that

dQT  dQ S  dQ L
The combined driving force in this case maybe termed enthalpy potential as
outlined below.

6.2.1 Enthalpy Potential

Consider an elementary wetted surface at temperature t s as shown in Fig. 6.1.

Above the wetted surface, there exists a film of air through which temperature and
vapour pressure gradients exist. In the immediate vicinity of the wetted surface, the
air is saturated at ts, WS and hS . Under equilibrium conditions, the rate of diffusion
of water vapour through the air film will equal the rate of condensation or
evaporation of water on the wetted surface.

W , h, t

Ws , hs , t s

Fig. 6.1 Condittons with Unsaturated Air Flowing over a Wetted Surface

Thus if fg is the film coefficient of heat transfer through the air film, we have for
the sensible heat transfer between the air and the wetted surface

dQ S  f g ( t  t S )dA (6.1)

288
 Also, if kW is the diffusion coefficient of water vapour through the air film, and if
hfgS is the latent heat a t s , we have for latent heat transfer between unsaturated air and
wetted surface
dQ L  h fgS dm v (6.2)
where
.
d m v  kW ( W  W S )dA (6.3)

so that
.
d Q L  kW ( W  W S )h fgS dA (6.4)

According to Lewis, we have

fg
kW 
Le C p
fg
 for Le  1
Cp

Combining Eqs (6.1) and (6.2) we have for total heat transfer
fg
dQ T  f g ( t  t S )dA  h fgS ( W  W S )dA
Le C p
fg h fgS
 [C p( t  tS )  ( W  W S )] dA (6.5)
Cp Le
Where
h fgS  hv  h fS
Rearranging
fg h fgS h fgS
dQ T  [( C P t  W )  ( C pt S  W S )] dA (6.6)
Cp Le Le

Taking average values of the humid specific heat CP and latent heat as h fgS we
may write
fg fg 1
dQ T  ( h  hS )dA  (  1 )h fgS ( W  W S )dA
Cp C p Le

289
 fg
 ( h  hS )dA , for Le  1 (6.7)
Cp
The quantity ( h  hS ) is the enthalpy potential. Thus, the driving force for the total
heat transfer is the difference in the enthalpy h of unsaturated air and the enthalpy hS
of saturated air at the temperature of the wetted surface. And the coefficient
determining this heat transfer is equal to f g /C p  kW .

6.2.2 Surface Temperature and Direction of Process

The following examples illustrate the usefulness of the concept of enthalpy

potential. Depending on the wetted surface temperature, and the state of the
unsaturated air in contact with the wetted surface, the direction of the process can be
explained. Let the air being conditioned be at state 1 as shown in Fig. 6.2, with the
dry bulb, wet bulb and dew point temperatures of t , t  and t d respectively. Then the
cases arise as follows:

Case I: Cooling and Dehumidification (Process 1 — S1)

The surface temperature is lower than dew point temperature of air.

tS1  t dQS is from air to surface
WS 1  W dQL is from air to surface
hS 1  h dQT is from air to surface

Fig. 6.2 Wetted Surface Temperatures and Directions of Processes

290
Case II: Simple Cooling (Process 1 — S2)

The surface temperature is equal to the dew point temperature.

t S 2  t d  t dQS is from air to surface
WS 2  W dQL is zero
hS 2  h dQT is from air to surface
dQT  dQS

Case III: Cooling and Humidification (Process 1 — S3)

The surface temperature is higher than the dew point temperature, but lower than
the wet bulb temperature of air.
td  tS 3  t dQS is from air to surface
WS 3  W dQL is from surface to air
hS 3  h dQT is from air to surface

The air is, on the whole, losing energy to the surface.

Case IV: Adiabatic Saturation (Process 1 — S4)

The surface temperature is equal to wet bulb temperature of air.

t  tS4  t dQS is from air to surface
W   WS 4  W dQL is from surface to air
h  hS 4  h dQT is zero
dQS  dQL

Case V: Cooling and Humidification (Process 1 — S5)

The surface temperature is higher than the wet bulb temperature, but lower than
the dry bulb temperature of air.
t  tS5  t dQS is from air to surface
WS 5  W dQL is from surface to air
hS 5  h dQT is from surface to air

291
The air is, on the whole ,gaining energy from the surface.

Case VI: Simple Humidification (Process 1 — S6)

The surface temperature is equal to the dry bulb temperature of air.

tS6  t dQS is zero
WS 6  W dQL is from surface to air
hS 6  h dQT  dQL

Case VII: Heating and Humidification (Process 1 — S7)

The surface temperature is higher than the dry bulb temperature of air.
t S7  t dQS is from surface to air
WS7  W dQL is from surface to air
hS7  h dQT is from surface to air

Note : It may be seen that in Cases III to V, there will be no latent heat transfer if the
surface is dry. On the other hand, in Case I there will be condensation of moisture
from the air on the surface, irrespective of the surface being either dry initially or wet.
Latent heat transfer is always present in sprayed coils.

6.2.3 Effective Surface Temperature

The surface temperature in air-conditioning equipment varies from one end of the
apparatus to the other end. Nevertheless, the effective surface temperature t s can be
considered as the uniform surface temperature that would produce the same leaving
air state as the varying surface temperature. In the case of a cooling and
dehumidifying equipment, it is equal to the apparatus dew point.

As an illustration Fig. 6.3 shows the entering and leaving states for air and chilled
water in a counter flow cooling and dehumidifying apparatus.

The wet bulb temperature of air decreases from t 1 to t 2 , whereas the chilled water
temperature rises from t w 2 , to t w1 ,. The temperature of the surface changes from t S 1
to t S 2 from one end to the other end of the heat exchanger. The effective surface

292
temperature t S would lie between t S 1 and t S 2 , and would correspond to the same
entering and leaving states of both fluids.

The selection of the air-conditioning equipment is usually based on the concept of

the effective surface temperature and bypass factor. However, the actual design of
such an equipment would involve the basic principles of heat and mass transfer
between the air and the beating or cooling media through the air-conditioning
apparatus.

Note: In the case of a direct-expansion cooling coil, the chilled water is replaced by
the evaporating refrigerant so that tw1 = tw2 = tr In the case of a spray equipment with
chilled water which itself forms the surface, the temperature of water is the same as
the surface temperature, so that tw= tS, tw1= tS1, tw2=tS2, and so on.

6.2.4 Numerical Procedure for Heat and Mass Transfer Calculations

between Unsaturated Air and Wetted Surface
Consider a process of heat and mass transfer between unsaturated air and a wetted
surface of area dA at temperature tS. Then, referring to Fig. 6.1 we may write for the
moisture transfer dm v

v = m
dm  a dW = kW dA ( W S - W) (6.8)

where m a is the mass flow rate of dry air.

Equation (6.8) may be written as
dW k dA
 W (6.9)
WS  W ma
The integration of Eq. (6.9) between the two ends of the exchanger gives

2
dW kW A
 WS  W 
a
m
(6.10)
1

 a is called the number of transfer units (NTU).

The dimensionless term kW A/m

Again, writing for the total heat transfer between air and the wetted surface, we
have

293
 Fig. 6.3 Entering and Leaving States of Air and Chilled Water in a Cooling and
Dehumidifying Apparatus

fg
dQ T  m
 a dh  dA( hS  h )
Cp
 kW dA( hS  h ) (6.11)

The integration of Eq. (6.11) gives

2
dh kW A
 hS  h m a
 (6.12)
1

Combining Eqs (6.9) and (6.11), we obtain

dh h h dh dW k dA
 S or   W (6.13)
dW W S  W hS  h W S  W ma

294
 Thus, for an actual apparatus. the value of kW A/m  a may be experimentally
obtained by plotting either 1/(hS  h) against h or 1/(WS  W) against W along the
process. and subsequently integrating the curve by numerical integration, e.g., by
applying Simpson’s rule according to which

(b  a )
b

 y dx 
a
3N
[ y 0  y N  4( y 1  y 3  ...  y N  1 )
(6.14)
 2( y 2  y 4  ...  y N  2 )]

where a and b are the limits of integration.

The range (b - a) is divided in to N equal parts where N is an even number. The

ordinates of the curve are y0 , y1 , y 2 ,..., y N 2 , y N 1 , y N .Once the value of the number of
transfer units kW A/m  a is obtained, the heat and mass transfer process can be
established along the exchanger for any given operating conditions.

6.3 Coil Equipment-Design of Cooling and Dehumidification Coils

The design of coils for only sensible heating and cooling is simple and follows the
procedure prescribed for any heat exchanger. Coils for simultaneous cooling and
dehumidification involve sensible as well as latent heat transfer and since they have
major applications in air conditioning, their design procedure is outlined here.

A section of a cooling and dehumidifying coil with just one illustrative fin is
shown in Fig. 6.4. The temperatures of air, wetted surface, outside and inside metal
wall and refrigerant at any section are t. tS, tmo , tmi. and tr respectively. The thermal
resistances to heat transfer consist of those due to the air film, condensate water layer,
metal wall and refrigerant film. In addition, there is a mass-transfer resistance
between the air and the wetted surface.

The resistances can be divided into two sections5 which are given below.

Section I:

This section is that of the air film. In this section both sensible and latent heats are
transferred. The sensible heat QS is transferred from the air to the surface by virtue of
the temperature difference (t-tS). Also, the latent heat QL is transferred by virtue of
the partial pressure difference (pv - pvs) or the specific humidity difference ( W  WS ),

295
where pvs, is the water vapour pressure and WS is the specific humidity at
temperature ts of the wetted surface.

W ,t
W S ,t S

Fig. 6.4 Section of Finned Cooling and Dehumidifying Coil:

Thermal Resistances

Let us consider that the coil is a finned-type. The total heat transfer through
section 1 with the inside tube surface area dAi is
. . . fg
d QT  d Q S  d Q L  ( h  hS )  f R dAi (6.15)
Cp
where  f is the fin efficiency and
At
R = Ratio of the total fin-side surface area At to the inside surface
Ai
area Ai

Section II:

This section comprises the three thermal resistances due to the condensate film,
metal wall and refrigerant film. The heat is transferred as sensible heat only by virtue
of the temperature difference (tS - tr) Then, if Ui represents the overall heat transfer
coefficient for the three resistances based on the inside tube surface area, the heat
transfer through the section is given by
dQ II  U i ( t S  t r )dAi (6.16)

By energy balance
dQ I  dQ II

296
so that we have from Eqs (6.15) and (6.16)
fg
U i ( t S  t r ) dAi  ( h  hS )  f R dAi or
Cp
t S  t r  f f g R kW
  f R (6.17)
h  hS UiC p Ui

The right-hand side is constant in case fg and Ui are assumed to be constant.

Equation (6.17) can be used for design by dividing the coil into a number of segments
since ts is varying.

Equation (6.17) can be written in another simplified manner. Writing the total and
sensible heat transfers separately, we have for Section I
fg
dQ T  dQ I  ( h  hS )  f R dAi (6.17)
CP
dQ  f ( t  t )  R dA
S g S f i (6.18)

Taking the ratio of the total heat to the sensible heat transfer we have
TH dQ T h  hS 1
  
SH dQ S C p ( t  t S ) GSHF

where GSHF is the required grand sensible heat factor of the apparatus. Thus
1
( h  hS )  C p( t  tS ) (6.19)
GSHF

Substituting from Eq. (6.19) into Eq. (6.15), we obtain for the total heat transfer
fg
dQ T   ( t  t S )  f R dAi
GSHF
 f g ( t  t S )  f R dAi (6.20)
Where
fg
f g  (6.21)
GSHF

Thus f g can be taken as the total heat-transfer coefficient on the air side, when the
sensible heat-transfer coefficient is f g . Putting dQII in Eq. (6.16) equal to dQI in
Eq. (6.20), we obtain the energy balance relation in the form

297
 t S  tr f g
 f R (6.22)
t  tS Ui

The overall heat-transfer coefficient U i can be calculated employing the usual

relation for thermal resistances in series, viz.,
1 1 x 1
   (6.23)
U i Ai f r Ai kAm f c  f R Ai
where
fr = Refrigerant side heat-transfer coefficient
 x = Thickness of metal wall
k = Thermal conductivity of metal wall
fc = Conductance of condensate layer
Ai,Am, At= Inside, mean and outside tube surface areas.

Table 6.1 gives the representative values of the air-side film heat-transfer
coefficients f g for different coil-face velocities. The following simple
equation can be fitted into these values.

f g  38 ( F V )0.5

Table 6.2 similarly gives the approximate values of the refrigerant-side heat-
transfer coefficients f r for direct expansion R22 coils as a function of tube diameter
and refrigerant flow.

Table 6.1 Air Side Film Conductance (With Half of Face Area as
Free Area and Staggard Tubes)

Coil Face Velocity, FV, m/s Conductance, f , W/m2 K-1

g
0.508 23.3
1.016 35.8
1.524 45.4
2.032 54.5
2.54 62.5
3.048 69.8

298
 Table 6.2 Boiling Refrigerant Conductance for R22 D-X Coil

Refrigerant Flow Tube OD, cm

TR/Circuit 1.27 1.5875 1.905
0.5 1420
0.7 2129
0.8 1420
1.2 2271 1420
1.5 2839
1.7 2129

For a precise estimation of fr, well-known boiling heat transfer correlations. For
chilled-water coils, the water-side coefficient can be found by using the Dittus-
Boelter equation.

Alternatively, we express the overall heat-transfer coefficient U t ,. for Sections

land II together for which the overall temperature difference is t - tr, and base it on the
total fin-side surface area At = RAi,
dQ T  U' t ( t  t r ) dAt
 f ' g ( t  t S )  f dAt

 U i ( t S  t r ) dAi (6.24)

Then the overall resistance based on the temperature potential (t - tr) is

1 1 1
 
U' t dAt f ' g  f dAt U i dAi

1 1 R
   (6.25)
U' t f'g  f Ui

Substituting the complete expression for Ui, we have

1 1 R x R Ai 1
    (6.26)
U' t f'g  f fr k Am fc  f

299
It may be noted that the design of a cooling coil involves the selection of number of
circuits, face velocity, refrigerant or chilled-water temperature, number of rows along
with parameters such as fin and tube spacing, arrangement of fins, etc.

Examples 6.1 and 6.2 illustrate the design procedure for a direct-expansion
cooling coil. Whereas in Example 6.1, a stepped approach is used and Simpson’s rule
is employed, in Example 6.2, the coil is not divided into sections, but an overall
design is given.

6.3.1 Air-Side Heat-Transfer Coefficient

Exact calculation of heat-transfer coefficient on the air side is based on the cor-
relations of Elmahdy and Biggs for flow of dry air over finned tubes as given below
J G Cp
fg  (6.27)
Pr 2 / 3

where J is the Colburn j-factor given by

J  C 1 Re C2 (6.28)
in which C1 and C2 are constants which are functions of geometric parameters of the
finned tubes. These can be expressed as
0.141 0.065
F   Dh 
C 1  0.159  T    (6.29)
 FH   FT 
0.049 0.077 0.549
F   FS   FD 
C 2  0.323  T      (6.30)
 FH   FT   S2 

wherein we have:
FT = Fin thickness
1  S1  S 2 
FH =Fin height = r2  r1  r2  r0    D0 
2 2 
Fs= Fin spacing
0.5
 S1 S 2  S1  S 2
FD=Fin diameter= 2 r2  2   
 4  2
 
S1=Tube spacing in a row
S2=Row spacing

300
 The hydraulic diameter Dh, is given by
A
Dh  4 S 2 C (6.31)
At
where AC is the minimum/clear flow area and At is the total air-side area of the coil.

The coefficient for wetted surface is then

fg
f g 
GSHF

The parameter for fin efficiency becomes

2 f g
m (6.32)
k fin FH

so that the fin efficiency can be evaluated from

tanh mF H
f  (6.33)
mF H

and the air-side thermal resistance from the expression

1
(6.34)
 f f g At

6.3.2 Diffusion Coefficient D and Mass-transfer Coefficient kW

For sensible heat transfer per unit area on the air side
Q S  t 
 f g ( t  t S )  k  
A  y  y  0, at wetted surface

The solution of the above equation has the form

 L u    C p  
b c
fg L
Nu   f (Re, Pr)  a  
 (6.35)
k     k 

For moisture transfer per unit area, we have in terms of diffusion coefficient D
mv  W 
  D  
A  y  y 0

301
 We have also defined mass-transfer coefficient k by
mv
 kW ( W S  W )
A

Thus, we have
 W 
kW ( W S  Wair )   D   
 y  y 0
The solution of the above equation has the form
c
L u    
b
kW L
 f (Re, Sc )  a     (6.36)
 D    D
 W   

where Sc  μ/ρ D is the Schmidt number. Comparing Eqs (6.35) and (6.36), we get
1 C
fg k  
Le    
C p kW  Cp D  D

where   k /  C p is the thermal diffusivity. The following correlations are

recommended for Lewis number.
For forced convection of air
 
2/3
Le    (6.37)
 D
For natural convection of air
 
0.48
Le    (6.38)
 D

For forced convection of air the values of  , D and Le have been compiled and are
given in Table 6.3 for various temperatures. The evaluation of properties has been
done at the wetted surface temperature:

Note .: It is seen that Le does not change very much. Average value for dry air is
0.897, and that far saturated air is 0.889. The average of bath is 0.893, which can be
accepted as its value in general.

302
Table 6.3: Thermal Diffusivity, Diffusion Coefficient and Lewis Number for Forced Convection
of Dry and Saturated Moist Air

Temp. Degree of  D Le
oC Saturation m 2 / s m2 / s
10 0 1.987 2.325 0.901
1 1.985 2.325 0.9
20 0 2.077 2.433 0.9
1 2.072 2.433 0.898
30 0 2.245 2.637 0.898
1 2.234 2.637 0.895
40 1 2.418 2.849 0.896
2.393 2.849 0.89
50 0 2.596 3.066 0.895
2.531 3.066 0.88
60 1 2.694 3.187 0.894
2.588 3.187 0.87

Example 6.1: Design of Chilled-Water Coil-Overall LMTD Method

Design a chilled-water cooling coil for each of the ground, first and third floors of
the studio building calculations of which are given in the load estimation sheet in
Table 19.6. The coil has to conform to some of the specifications given by the
manufacturer which are as follows:

5/8 in Cu tubes :OD = Do = 15.8 mm

ID=Di= 14.4 mm
No. of tubes/row :34
No. of circuits :34
Frontal pitch (Tube spacing) :S1 = 38.1 mm
Back pitch (Row spacing) :S2 = 43.2 mm
Al fins :
Fin thickness : FT = 0.23 3 mm
Fin density : 50/m
1
Fin spacing :Fs= =0.002 m=2 mm
5
Thermal conductivity of fin : kAl = 201 W / m. K
Fouling factor (inside) :0.00009 m 2 K/W
With these construction features, the area parameters are:

303
 Total outside finned surface area A
 t  10.567
Inside tube surface area Ai
Total outside finned surface area A
 t  15.55
Face area FA
Clear free flow area AC
  0.5
Face area FA
Estimate the size and rows of coil. Make suitable assumptions.

Solution

Coil Face Area

Assume coil face velocity FV = 54 m/s

Coil face area
cmmS 351
FA  
FV 2.54  6

Air-Side Heat- Transfer Coefficient

Air enters at 28.6ºC, and leaves at 15.8ºC. The properties of air at the mean
temperature of 22ºC are:   1.205 kg / m 3 ,   1.82  10 5 N .s / m 2
C p =1.0057 kJ/kg.K Pr = 0.71

Hydraulic diameter
S2  0.5 
Dh  4 AC  4  (43.2mm)=5.567 mm
At  15.55 

Fin diameter and fin height

S  S 2 38.1  43.2
FD  1   40.65mm
2 2
1  S  S2  1
FH   1  D0   (40.65-15.8)=12.425mm
2 2  2
Constants of Eq. (6.28)
0.141 0.065
F   Dh 
C1  0.159 T   
 FH   FT 

304
 0.141 0.065
 0.233   5.564 
 0.159     0.11155
 12.425   0.233 
0.049 0.077 0.549
F   FS   FD 
C 2  0.323 T     
 FH   FT   S2 
0.049 0.077 0.549
 0.233   2   40.65 
 0.323     
 12.425   0.233   43.2 
=-0.3034
Mass velocity based on clear minimum flow area AC = 0.5 (FA)

cmmS  1
Gmax  
60 0.5( FA )
351(1.205)
  6.096kg / m 2 .s
60(0.5)(2.303)

Reynolds number
Gmax D0 6.096( 0.0158 )
Re    5292
 1.82  10 5

Colburn j-factor
J  C1 Re C2
 0.11155 ( 5292 )0.3034  0.008274

Dry air-side heat-transfer coefficient

f g  J Gmax C p Pr 2 / 3
=0.008274 (6.096) (1005.7) (0.71) 2 / 3
 63.74 W / m 2 .K

Wet air-side heat-transfer coefficient

fg 63.74
f g    67.3 W / m 2 .K
GSHF 0.9469
Fin parameter
2 f g 67.31  2
m   53.3
k Al FH 201( 0.233  10 3 )

305
Fin efficiency
tanh m FH tanh[( 53.3 )( 12.425  10 3 )]
f    0.92
m FH 53.3( 12.425  10 3 )

Water-Side Heat-Transfer Coefficient

Assume entering water temperature, EWT = 7.8ºC, and leaving water temperature,
LWT = 13.3ºC. Thus,  Tw = LWT — EWT = 5.5ºC. Properties of water at the mean
temperature of 10.6ºC are:
2
k w =0.588 W/m.K,  w = 1.31  10-3 N .s / m , Prw = 9.12

Mass flow rate of chilled water and water velocity

Q 93.67
m w    4.068 kg / s
C p 4.1868( 5.5 )
m w 4.068
uw    0.735 m/s
Di2 
 N 1000( 34 ) ( 0.0144 ) 2
4 4

Note : This water velocity is low. It will result in low heat-transfer coefficient, and
hence more rows and high capitol cost. A very high water velocity will, however,
cause erosion of tubes.

Reynolds number
Di u w  w ( 0.0144 )( 0.735 )( 1000 )
Re w    6814
w 1.31  10 3

Water-side heat-transfer coefficient

k 
f i   w 0.023 Re 0.8 Pr 0.4
 Di 
 0.588 
 0.023 ( 6814 )0.8 ( 9.12 )0.4
 0.0144 
 2651 W / m 2 .K
Overall heat-transfer coefficient (neglecting Cu-tube resistance)
1 1 1 At 1 At
  
U t  f f g f i Ai f fouling Ai

306
 1 1
  (10.567)  0.00009(10.567 )
0.92(67.31) 2651

 U t  47.4 W / m 2 .K

Log-mean temperature difference

( 28.6  13.3 )  ( 15.8  7.8 )
Tm  LMTD   11 .26  C
28.6  13.3
ln
15.8  7.8
Total fin-side surface area
Q 98.67  10 3
At    175.5m 2
U t Tm 47.4( 11 .26 )

Number of rows
At 175.5
n   4.92(  5 rows)
 At  ( 15.5 )( 2.303
  FA
 FA 

Notes:

This is a high sensible heat factor application, (GSHF = 0.946). BPF assumed is 0.15. For
such on application number of rows calculated, 5, is too high. Area calculated, 175.5 m 2 , is
very large. The coil will hove high capital cost. The manufacturer’s specification, and
assumed design values need modifications. The following points emerge.

(i) The manufacturer is making coils with a fin density of only 50 fins per metre. This
comes to 0.5 fin/cm. It has resulted in a large coil. Hence, the fin density must be
increased to about 5-6 fins/cm.
(ii)The present face velocity assumed is 2.54 m/s. It is giving low air-side, and hence low
overall heat-transfer coefficient. The air velocity could be increased to 3 to 3.5 m/s.
(iii) Water velocity assumed/calculated is 0.735 m/s. It can be raised to at least 1 m/s.
This will increase water-side heat-transfer coefficient.
(iv) The air-side coefficient is, however, the controlling coefficient. The overall
coefficient U is lower than air-side coefficient. Hence, increase in water-side coefficient
will not have much effect. Nevertheless, it will affect LWT.
(v) Further, the effect of changing EWT, and hence the refrigerant evaporation temperature
on total cost should also be studied. All this is best done on a digital computer.

Romachandran and Arora have developed a detailed computer programs for the
simulation, design and optimization of both direct-expansion and chilled water cooling coils.

307
Example 6.2: Design of Direct-Expansion Coil: Heat-Transfer
Coefficient Given

4.717 m 3 / s of air at 32.2ºC DBT and 50 percent RH enters a finned-coil direct-

expansion dehumidifier with the coil surface area per row of 42.02 m 2 . The
refrigerant temperature is 4.4ºC. The air-side heat-transfer coefficient may be taken as
83 W / m 2 .K . The overall heat-transfer coefficient from the air-water interface to the
refrigerant bulk based on the fin-side surface area is 275 W / m 2 .K . Calculate the
required number of coil rows for the leaving air state at 9.7 o C WBT.

Solution

Refer to Fig. 6.5.

Fig. 6.5 Figure for Example 6.2

308
Air inlet conditions
t1=32.2ºC; RH1=50%; h1=71.94kJ/kg;
v1  0.901 m 3 / kg
Mass flow of air of air
4.717
m a   5.235 kg/s
0.901

At 4.4ºC and 100 per cent RH

hA = 17.69 kJ/kg

Mass-transfer coefficient
kW  f g / C p Le
83
=  0.09 kg / s.m 2
( 1021.6 )( 0.893 )

Enthalpy of air at the exit at 9.7ºC WBT

h2= 28.54 kJ/kg

Enthalpy difference between inlet and outlet air conditions

h1-h2= 71.94 - 28.54= 43.4 kJ/kg

Let the difference be divided into six equal increments of 7.23 kJ/kg each.
Now
Ut At = Ui Ai
so that/
Ui
 U t  275 W / m 2 K ( given )
R
Hence, taking  f  1
h  hS Ui U 275/1000
  t   3.l kJ/kg.K
t S  t r kW R K W 0.09

The above equation can also be solved graphically for each increment as shown in
Fig. 6.5. The 4.4ºC DBT line cuts the h1 line at X. A line of slope

(h  hS )
h / t  = 3.1
(t S  t r )

309
may now be drawn from point X to cut the saturation line at S1. Point S1 on the
saturation line satisfies this condition of slope of 3.1. The interface temperature at this
section tS1 is then found to be 14.4ºC. Join point 1 to point S1 and where it cuts the ha
= h1 -7.23=64.71 kJ/kg line, gives point a on the process line.

Now, the enthalpy line ha cuts the 4.4ºC vertical at Y. Again, draw a line of slope
3.1 from Y, to cut the saturation line at Sa. Join a to Sa. This cuts the hb = 64.71 -7.23
= 57.48 kJ/kg line at b. In a similar manner, the remaining points c, d, e and 2 may be
located and the process line completed.

Table 6.4 may thus be drawn up. The same may be established numerically as
well. Using the above equation for calculation of hs by trial and error for finding
suitable values of tS, we obtain Table 6.4.

Table 6.4 Stepped Calculations for Example 6.2

h, kJ/kg h1=71.94 64.17 57.48 50.25 43.02 35.79 h2=28.54

hs1=39.58 36.32 33.52 30.0 27.70 24.68 hs2=22.12
hs, kJ/kg
0.0309 0.0352 0.0417 0.0506 0.0653 0.0900 0.158
1/(h-hs)

Then from Simpson’s rule

71.94  28.54
71..94
dh
 h  hS

36
=[0.0309 + 0.1558 + 4(0.0352+0.0506 + 0.09)
28..54
+ 2(0.0417 + 0.0653)]
kW A
= 2.66 =
m a
2.66(5.235)
 A  154.7m 2
0.09

Required number of rows

154.7
n  3.7  4( Select )
42.02

310
Example 6.3: D - X Coil Design: Heat Transfer Coefficients Appropriately
Selected

0.472 m 3 / s of air are cooled from 26.7ºC DBT and 21.1ºC WBT to 12,8ºC DBT
and 12.2ºC WBT. Design a suitable direct-expansion cooling coil.

Select 1.5875 cm OD tubes at 3.81 cm centres staggered and 0.71 mm wall

thickness. The total fin-side surface area per row is 22 m 2 for each m 2 of face area.
The inside surface area per row is 1.2 m 2 per m 2 of face area. Assume the resistance
of the metal wall as 0.0044 W 1 m 2 K .

Solution

Extended surface to inside tube surface area ratio

Finned surface/r ow/m 2 FA 22
R   18.3
2
Inside tube surface /row/m FA 1.2

Figure 6.6 shows the path of the conditioning process. If the entering and leaving air
states are joined, and the line extended to intersect the saturation curve, we obtain the
apparatus dew point of the coil
t ADP  10.4 oC

This represents the average temperature t S of the coil. But the surface temperature t S
will be higher on the air-entering side and lower on the air-leaving side. The coil
bypass factor can also be found.

t 2  t ADP 12.8  10.4

X    0.147
t1  t ADP 26.7  10.4
The total and sensible heat-transfer rates are as follows:
TH  Q T  Q v  ( h1  h2 )
=(0.472) (1.2) (61.1 -34.5) 15.0 kW

SH  Q S  Q v  C p ( t1  t 2 )
= (0.472) (1.2) (1.0216) (26.7 - 12.8) = 8.04 kW
Q 1 15
Ratio = T    1.866
Q S SHF 8.04

311
 W

Fig. 6.6 Figure for Example 6.3

Since the surface temperature is not the same throughout the coil, this ratio will
actually not be the same. However, the assumption of a constant ratio leads to only a
small error and results in a slight over sizing of the coil (see the actual path in Fig.
6.6).

Take a face velocity of 2.54 m/s.

Face area
Q 0.472
FA  v   0.186 m 2
FV 2.54

Air-side film conductance (from Table 6.1)

f g  62.5 W m 2 K 1

Total heat-transfer coefficient

TH
f g  fg
SH
=1.866 (62.5) = 116.53 W m 2 K 1

Assume a coil loading of 1 TR per circuit.

Boiling heat-transfer coefficient (from Table 6.2)
f r  1704 W m 2 K 1

Assuming the fin efficiency to be equal to unity and neglecting the resistance of the
thin condensate layer, we have for the overall heat-transfer coefficient

312
 1 1 At x At 1
  
U' t f r Ai k Am f ' g
18.3 1
  0.0044 
1407 116 .63
U t  42.17 W m K 1
2 

First approximation:
Assume a refrigerant temperature of t r  1o C . Then the temperature differentials at
the two ends of the cooling coil are
t1  t1  t r  26.7  ( 1 )  27.7 oC
t2  t2  tr  12.8  ( 1 )  13.8oC
Log mean temperature difference
27.7  13.8
t m   19.95º C
27.7
ln
13.8

Total fin side surface area

Q
At 
U t t m
15  10 3
  17.83m 2
( 42.17 )( 19.95 )

If n is the number of rows, then

At = n (FA) (surface area/row/ m 2 FA )
17.83=n (0.186) (22)

whence
n = 4.4 rows
Surface temperatures, from Eq. (6.24)
U
t S1  t1  t ( t1  t r )
f g
42.17
 26.7  (26.7  1)  16.7º C
116.63
U'
tS 2  t2  t ( t2  tr )
f g

313
 42.17
 12.8  (12.8  1)  7.8º C
116.63

Note: As the number of rows is 4.4, it is not a whole number. Hence, further approximation
is necessary.

6.4 Optimal Design of Cooling and Dehumidifying Coils

It will be seen from Example 6.3 that for the given construction of the cooling
coil, two choices are available, viz., to select a 6-row coil or a 4-row coil in lieu of the
calculated result of 4.4 rows. Let us see the effect of selecting 6 or 4 rows on other
parameters.

6.4.1 Six-Row Coil

It can be seen that if a 6-row coil is selected, it will result in the following
changes:

(i) Higher refrigerant temperature tr can be used and hence a higher surface
temperature tS is maintained.

(ii) Leaving conditions at a lower temperature as a result of lower BPF:

Back calculations show that the leaving air conditions are:

DBT = t2 = 12.2ºC
WBT = t 2 =1l.7ºC
h2 = 33.5 kJ/kg

(iii) Increased total heat removal

Q T  Q v  h = (0.472) (1.2) (61.1 - 33.5) = 15.63kW

Required mean temperature difference

15.63
t m   15.3 o C
( 42.17 )( 6 )( 0.186 )( 22 )

Refrigerant temperature required

tr = 2.8ºC (As against - 1ºC selected in Example 6.3)

314
6.4.2 Four-Row Coil

A smaller number of rows will result in the following:

(i) Lower refrigerant temperature

tr=-2.8ºC
and a lower mean surface temperature.

(ii) Leaving conditions at a higher temperature as a result of higher BPF.

(iii) Decreased heat removal

Q T = 14.95 kW

A comparison will thus show that a 6-row coil will cost more. However, for the
same condensing temperature, say tk = 37ºC although a 4-row coil would cost less but
other costs will increase as follows:

(i) It will require 14 per cent more compressor displacement. As a result of lower
saturated suction temperature.
(ii) It will require 10 per cent more power for a capacity of 14.95 kW at tr =-2.8ºC,
as against a capacity of 15.63 kW at tr = + 2.8ºC for a 6-row coil. Obviously, for
an optimal design, an analysis of costs is required. It may be pointed out here
that other choices are also available. These include:

(i) Lower face velocity: This will result in a lower surface temperature and bring
the outlet condition nearer to 2. Again, the face area and hence the cost of
cooling coil will increase.

(ii) Higher tons refrigeration per circuit: This can be adopted only if it is practical
without excessive pressure drop. This will give a higher conductance f r on the
refrigerant side, which will lower the surface temperature t S and allow the use
of a 4-row coil at the same refrigerant temperature of t r  1o C .

(iii) Another type coil: A coil of different construction may be used, i.e., a coil
with a lower metal resistance and a lower or higher area ratio R.
All the alternatives can be compared for the optimal design of the cooling coil.

Note : Usually, a refrigerant temperature 3 to 5ºC below the coil ADP provides on
economical selection.

315
6.5 Spray Equipment-Design of Air Washers and Cooling Towers

In a spray chamber, air is brought in direct contact with a dense spray of water.
The schematic diagram of an air-spray equipment is shown in Fig. 6.7.

Fig 6. 7 Spray Equipment

The air washer and cooling tower are two common types of spray equipment used
in air conditioning. In spray equipment, there is direct contact between air and the
sprayed water. Consider such a heat and mass exchanger in which the mass flow rates
of the dry air and water are ma and mw respectively. The processes undergone by air
and water are similar to those shown in Fig. 6.3. Equations derived in Sec. 6.2 can be
applied to the process. However, in the case of spray equipment, the wetted-surface
area is not clearly known as it is formed by the surfaces of individual droplets in the
total volume of the equipment. We, therefore, denote the interfacial contact area of
the surface of water droplets per unit volume of the equipment by a, so that the
contact area in a differential volume dV is dA = a dV, and the total area A = a V.
Also, the temperature of water t w , in the equations replaces the temperature t S of
the surface.

The energy balance equation for the process can be written using Eq. (6.7) as
dQ   m
 W CW dtW  m  a dh  kW ( a dV )( hS  h ) (6.39)

where CW , is the specific heat of water. Equation (6.39) on integration gives

2
CW dtW kW a V
  (6.40)
1
hS  h mW

316
Equation (6.40) can be used in conjunction with Eq. (6.12). Comparing the two, we
have
dh m
 W (6.41)
CW dtW ma

Hence, if the properties of moist air are taken in the form of an enthalpy-temperature
chart, which is another form of the psychrometric chart shown in Fig. 6.8, a line AB
of constant slope given by CW m W / m a can be drawn for any process provided the
initial state of water tW 1 and that of air at 1, and their mass flow rates are known.
Such a line is called the energy balance line. The line AB is drawn from A, the point
where the horizontal line corresponding to h1 cuts the vertical line corresponding to
tW 1 upto the point B which is on the vertical corresponding to the water inlet
temperature tW 2 .Thus the enthalpy h of air varies along the line AB at any section
according to the water temperature tW . The vertical distance between the saturation
line and the line AB represents the enthalpy potential (hS  h) at any section at the
water temperature tW . See how the driving potential changes from (hS1  h1 ) at one
end to (hS 2  h2 ) at the other end.

.
C W mW
Slope=
.
ma

Fig.6.8 Process Line for Air Washer on h-t Chart

317
 To determine the value of the dimensionless performance coefficient termed NTU,
viz., kW a V / m W , for any equipment from the performance curve as in Fig. 6.3,
one can also plot a graph between 1/(hS  h) against tW and evaluate the area of the
curve for substitution in Eq. (6.40).

In the case of pumped recirculation or evaporative cooling, WS is constant, and

hence Eq. (6.9) can be integrated directly, so that
W  W2 aV
 ln S   kW
W S  W1 ma
W S  W2
 e Z (6.42)
W S  W1
where
kW a V
Z
a
m

The process is shown in Fig. 6.9. It will be seen that the air washer humidifying
efficiency and bypass factor are given in terms of Z by the following expression

W 2  W1
H  ( 1  X )   1  e Z
W S  W1

Fig.6.9 Process Line for Air Washer on Psychrometric Chart

318
 Example 6.4 illustrates the design procedure for a spray dehumidifier. The same
procedure is adopted in respect of air washers in general (humidifier and evaporative
cooler) and cooling towers.

Example 6.4 Sprayed Coil

Air at 32.2°C DBT and 50 per cent RH, enters a spray-type dehumidifier at the rate
of 4.717 m 3 / s . Chilled water enters at 4.4°C and leaves at 11.2°C. The ratio of
water to air mass flow rate is 1.2. The face velocity of air is 2.032 m/s. The value of
the product kW may be taken as 1.334 kg / s m 3 . Calculate the length of the
dehumidifier and the state of air at the exit assuming parallel flow.

Solution

Refer to Figs 6.10 and 6.11 Air inlet conditions

3
t 1 =32.2°C, RH1 = 50%, h1 = 71.94 kJ/kg of dry air; v1= 0.901 m /kg of dry air

Now
dh m
  W  1.2
CW dtW m a
Note: The negative sign is taken far parallel flaw.

Also, we have
2
dtW kW a V
 CW  
h h
1 S
m W

319
Fig 6.10 Air and Water Temperature for Example 6.4

Fig. 6.11 State Changes of Air and Water for Example 6.4

320
or
11 .2
dtW kW a V
CW  h  hS 
m W
4.4

Divide 6.8°C temperature interval into four equal parts with ordinates at 4.4, 6.1, 7.8,
9.5 and 11.2°C. Table 6.5 can be prepared by calculating h, by energy balance for
each increment, i.e.,
h  1.2 CW tW  1.2( 4.187 )( 1.7 )  8.54kJ / kg of dry air

Table 6.5 Stepped Calculations for Example 6.4

Section 1 a b c
tw o C 4.4 6.1 7.8 11.2
h kJ / kg 71.94 63.40 54.86 37.78
hS kJ / kg 17.72 21.07 24.58 32.24
h  hS 54.22 42.33 30.28 5.54
1 / h  hS 0.0184 0.0236 0.033 0.1805

Using Simpson’s rule

( 11 .2  4.4 )
11 .2
dt w
 h  hS 
34
[ 0.0184  0.1805  4( 0.0236  0.555 )  2( 0.033 )]
4.4

=0.3294
Hence
kW aV
 0.3294( 4.187 )  1.3792
m W

Mass flow rate of dry air

Q 4.717
m a  v   5.235 kg/s
v 0.901

Mass flow rate of water

 W  5.235  1.2  6.282 kg/s
m

Volume of dehumidifier

321
 1.3792 m W
V
k a
6.282( 1.3792 )
  6.495 m 3
1.334
Face area
4.717
FA   2.322 m 2
2.032

Length of dehumidifier
V 6.495
L   2.8 m
FA 2.322

In order to find the state of air at exit, the process can be plotted on the psychrometric
chart as shown in Fig. 6.11. Joining 1 and S1, we get a, joining a and Sa, we get b,
and on. Finally joining c and Sc we get the leaving air state 2, for which the dry bulb
and wet bulb temperatures are
t 2  15 o C
and
t 2  13.6 oC

 W in Cooling
6.5.1 Significance of Performance Coefficient kW a V / m
Tower Selection

 W the integral on the left hand side of Eq. (6.40) must

To obtain he value of kW a V/m
be evaluated.

The value of the integral cannot be obtained directly. For the purpose, the cooling
tower is subdivided into “N” sections, and the integral is replaced by a summation of
values prevailing in each section.

Thus, Eq. (6.40) takes the form

kW a V N
( tW )i
 CW  (6.43)
mW i  1 ( hS  h )i

This gives the value of the coefficient as a function of m W /m

 a . If the summation is
done for specified values of

322
 (i) WB To fair t 1 at inlet
(ii) Water inlet temperature t w1
(iii) Water outlet temperature t w2

then the summation gives required value of the coefficient. A plot of the same as a
function of m W / m a represents the thermal demand as shown by curve A in Fig.
6.12. The cooling tower must fulfil this condition. According to the curve, the larger
. .
the value of m w / ma , the larger is the coefficient and hence the demand on the
cooling tower in terms of size/volume „V‟ and the wetted surface area ratio ‘a’ which
means smaller droplet size/atomization, number of stacks, packing, etc.

kW a V m W
Fig. 6.12 as a Function of Experimental and Thermal
m W m a
Demand Curves

On the other hand, if the summation is done, by actual measurements in an

experiment on the cooling tower, we get available value of the coefficient for that
cooling tower. It is found that entering WBT of air t1 and inlet water temperature t w1
do not affect the coefficient. However, the values of m W and m a do influence water

323
cooling, and hence affect the coefficient. The variation of the coefficient for a cooling
 W / m a can be expressed by
tower as a function of m
y
kW aV m
 
 x W  (6.44a)
mW ma 

where x and y are determined experimentally. Taking logarithms, this relation takes
the form of a straight line
 kW a V  m 
ln   ln x - y ln W 
 (6.44b)
 m W  
m a 

A typical relation for a cooling tower

0.72
kW aV m  
 0.29  W  (6.44c)
mW ma 

is shown by line B in Fig. 6.12.

Note : kW  f g /Le C p depends mainly on f g and hence on m a . (a) depends on the

construction features of the tower, and an m W / m a . Hence, the coefficient naturally
depends on m W / m a . The larger the value of m W per unit mass of m a , the lower is the
coefficient signifying less contact surface available for heat and mass transfer to
result in cooling.

The intersection of curve A and line B, viz., the operating point C, represents a
complete match between the capacity of the cooling tower and the required cooling of
water at the rate ( m W / m a )C for given t1 , tW 1 and tW 2 At a lower value of m W / m a
the cooling capacity of the tower is greater than required, while at a higher value of
m W / m a the capacity is less than required.

Example 6.5: Experimental Determination of Performance Coefficient

A test is performed on an induced draft counter flow cooling tower. The following
observations are made:

Water flow rate : 12.62 kg/s

Airflow rate : 11.9 kg/s
Entering water temperature : 36.3ºC

324
 Leaving water temperature : 32.1ºC
Ambient air conditions : 43.3ºC DBT, 25.6ºC WBT

 W If the dimensions of
Determine the value of the performance coefficient kW a V/m
the tower are length L = 3.9624 m, width W = 2.616 m and height H = 2.438 m, what
is the value of kW a ?

Solution

Let the cooling tower be considered as divided into 6 sections as shown in Fig.
6.13 with the water temperature dropping by 0.7ºC in each section. The values of t w ,
and hS  f ( t S  tW ) are given in Table 6.6 for each section. The values of enthalpy
of air for each section are calculated from the change in enthalpy corresponding to
m W / m a = 12.62/11.9 = 1.06 given by
m
h  W C pw tW
m a
= (1.06) (4.1868) (0.1) 3.9 kJ/kg
The values are also given in Table 6.6.

Table 6.6 Values of Properties at Sections for Example 6.5

Section 0 1 2 3 4 5 6
32.1 32.8 33.5 34.2 34.9 35.6 36.3
ts= tw,ºC
111.2 115.1 119.5 123.6 128.4 133.0 137.8
hs, kJ/kg
h, kJ/kg 79.0 82.9 86.8 90.7 49.6 98.5 102.4
1 .03106 .03106 .03058 .0304 .02959 .02899 .02825
y0 y1 y2 y3 y4 y5 y6
hS  h

Using Simpson’s rule

dtW tW
 hS  h  3N [ y0  y6  4( y1  y3  y5 )  2( y2  y4 )]
4.2
 [.03106 + .02825 + 4 (.03106 + .0304 + .02899)
3 6
+ 2 (.03058 + .02959)]
=0.12634

325
Performance coefficient
kW a V dt
 CW  W  4.1868 (0.12634) = 0.529
m W hS  h

Cross-sectional area of flow of air and volume of cooling tower

A = WL = 2.616  3.9624 = 10.366 m 2

V = AH = 10.366  2.438 = 25.27 m 3

Value of coefficient kW a

m W ( 12.62 )
kW a  0.529  0.529
V 25.27
= 0.2642 kg / s.m 3

Fig. 6.13 Induced Draft Cooling Tower: Figure Showing Sections for Example 6.5

326
Notes :

1. This value of performance coefficient is much below the value given by line B in
Fig. 6.12.

2. More the number of sections token, better is the accuracy.

3. Calculations con also be done by using Eq. (6.43) for summation, instead of
Simpson’s rule, taking mean values for sections 0- 1, 1-2, etc.

6.5.2 Crossflow Cooling Tower

Cooling towers without fans are known as atmospheric cooling towers in which
the air velocity depends on wind velocity. The flow of air and water in them is cross
flow. Some forced or induced draft cooling lowers are also designed as cross flow
towers. This reduces the height of the tower. Such towers are convenient for
installation on the roofs of high-rise buildings.

Calculations in crossflow towers can be done by dividing the tower in N number

of horizontal and vertical sections. The volume of each section is thus reduced to V/N.
Similarly, the area A = aV as well as the performance coefficient are reduced
accordingly. Example 6.6 illustrates the procedure by dividing an atmospheric
cooling tower into 4 horizontal sections.

Example 6.6: Atmospheric Cooling Tower

Water enters an atmospheric cooling tower at 7.57 kg/s at 36.4ºC. The dimensions
of the tower are W = 1.853 m, L = 2.134 m, H = 3.886 km. The wind velocity can be
taken as minimum/normal equal to 3 miles per hour. The performance coefficient for
the cooling tower as a whole can be taken as kW a V / m  W = 0.8. Find the
temperature of water leaving cooling tower, and the heat removed, if the WBT of
ambient air is 28.3 ºC.

Solution

Volume of the tower

V= WLH = 1.853 (2.134) (3.886) = 15.366 m 3

Face area of the tower

FA = WH = 1.853 (3.886) = 7.2 m 2

327
Wind velocity
3  1760  3  0.3048
C  1.341 m/s
3600

Divide the tower into 4 sections horizontally as shown in Fig. 6.14. Flow rate of air
through the whole tower
m a  ( FA )( C )
= (7.2)(l.341)(1.2)= 11.59kg/s

Flow rate of air through each section

11 .59
m a / 4   2.9 kg/s
4

Coefficient kW a for each section

 kW a V  m W
kW a   
 m  V
 W 
0.8( 7.57 )
  0.394 kg/s.m3
15.366

Fig. 6.14 Atmospheric Cooling Tower Divided into 4 Section for Example 6.6

328
Let the temperature of water leaving sections be tW 2 ,tW 3 ,tW 4 t, and tW 5 . The enthalpy
of air entering is ha = 92 kJ/kg of dry air. at the given WBT of 28.3ºC. Let the
2 3 4
enthalpy of air leaving sections be hb1, hb , hb and hb . Volume of each section is
15.366/4 = 3.8415 m 3 . Calculations for sections are then done as follows:

Section 1:
tWin  tW 1  t S 1  36.4C
hS  hS 1  116 .5 kJ / kg of dry air
ha  92 kJ/kg of dry air
Q  ( k a )( V )( h  h )
W S
 ( 0.394 )( 3.8415 )( 116 .5  92 )  37.08 kW
Q 1 37.08
h    12.8 kJ/kg of dry air
m a1 2.9
Q 1 37.08
tW    1.17C
m W CW 7.57( 4.1868 )
hb1  62  12.8  104.8 kJ/kg of dyr air
tWout  tW 2  36.4  1.17  35.23C
Values similarly calculated for all section are given in Table 6.7.

Table 6.7 Section-Wise Calculations for Crossflow Atmospheric Cooling Tower in Example
6.6, Fig. 6.14.

Section t win hS ha . h Δt whb t wout

ΔQ kJ/kg
oC kJ/kg kJ/kg o C kJ/kg oC
kW
1-2 36.4 116.5 92 37.08 12.8 1.17 104.8 35.23
2-3 35.23 115.0 92 34.8 13.54 1.09 105.5 34.15
3-4 34.15 114.0 92 33.3 12.96 1.05 105.0 33.1
4-5 33.1 113.0 92 31.17 12.36 1.00 104.4 32.09

Note that the wetted surface temperature for each section has- been taken as-equa1 to
the temperature of water at inlet to the section. For example, for section 1 - 2,
t S  tW 1 . Results can be improved if we take

tW 1  tW 2 36.4  35.23
tS    35.82C
2 2

329
for section 1-2, and similar average values for all the sections. Doing accordingly, in
the second approximation, we find the heat-transfer rates, and the temperatures of
water leaving sections are as follows:

Δ Q, kW 36.1 34.2 32.8 31.3

tw, ºC 35.26 34.18 33.14 32.2

Thus, the temperature drop of water is 36.4-32.2 = 4.2 ºC. Wet bulb approach is 32.2 —
28.3 = 3.9 ºC.

Note: The efficiency of the tower is found as follows:

Temperature drop of water
 tower 
Maximum possible temperatur e drop
36.4  32.2 4.2
   0.52( 52%)
36.4  28.3 8.1

Total heat-transfer capacity of the tower is 35.26 + 34. 18 + 33. 14 + 32.2 = 134.8
kW
 equal to the heat rejected in condenser.
=Q k

Notes:

1. if the COP of the refrigerating machine is taken as 3, this will correspond to a

refrigerating capacity of [3/(1+3)] 134.8 = 101 kW (28.8 TR).
2. Heat transfer can also be found as Q k  m
 W CW tW .

330
6.6 Problems

6.6.1:
(a) 30 cmm of air is to be cooled from 25ºC DBT and 50 per cent RH to 12ºC DBT
and 11ºC WBT using a direct-expansion coil with 1.27cm O.D. tubes with 0.71 mm
wall thickness and staggered at 3.81 cm centres. Determine the total fin-side surface
area per row per m 2 of face area, and also per m 2 of the inside-tube surface area.
The coil has 5 fins/cm.

(b) Assume the air-side heat-transfer coefficient as 62.5 W / m 2 K and the

refrigerant-side coefficient as 1500 W / m 2 K . The thermal conductivity of copper is
385 W/m.K. Assuming the fin efficiency to be 95 per cent, and neglecting the thermal
resistance of the condensate layer, find the value of the coefficient for total heat
transfer based on the fin-side surface area.

(c) If the coil contains four rows. determine the evaporating temperature of the
refrigerant.

6.6.2. Water at the rate of 1 kg/s is to be cooled in a cooling tower from 32ºC to 28ºC.
The ambient dry bulb and wet bulb temperatures are 40ºC and 26ºC respectively.
Taking the ratio of rates of the mass flow of water to mass flow of air as 0.75,
calculate:

(a) Required tower volume

(b) Condition of air at exit
(c) Height of tower if the air velocity at the exit is 1.5 m/s.

The mean value of the product of the coefficient of diffusion of water vapour into air
based on the specific humidity and surface area of water droplets per unit volume of
equipment, viz., kW a , may be taken as 0.5 kg / s.m 3 .

6.6.3. Air enters an evaporative cooler at the rate of 4.5 m 3 / s at 40ºC DBT and 26ºC
WBT. The required relative humidity at exit is limited to 60 per cent.
Calculate:

(a) The dry bulb temperature of air at exit.

(b) The length of air washer if the face velocity at the inlet is 2 m/s.
(c) The required humidifying efficiency of the air washer.

Take the mean value of the product kW a for the apparatus as 1.3 kg / s.m 3 .

331
Chapter 7

Air Refrigeration Cycles

7.1 Introduction

In an air refrigeration cycle, the air is used as a refrigerant. In olden days, air
was widely used in commercial applications because of its availability at free of
cost. Since air does not change its phase i.e. remains gaseous throughout the
cycle, therefore the heat carrying capacity per kg of air is very small as compared
to vapour absorbing systems. The air-cycle refrigeration systems, as originally
designed and installed, are now practically obsolete because of their low
coefficient of performance and high power requirements. However, this system
continues to be favoured for air refrigeration because of the low weight and
volume of the equipment.. The basic elements of an air cycle refrigeration system
are the compressor, the cooler or heat exchanger, the expander and the
refrigerator. Before discussing the air refrigeration cycles, we should first know
about the unit of refrigeration, coefficient of performance of a refrigerator and the
difference between the heat engine, a refrigerator and a heat pump.

7.2 Units of Refrigeration

The practical unit of refrigeration is expressed in terms of `tonne of

refrigeration' (briefly written as TR). A tonne of refrigeration is defined as the
amount of refrigeration effect produced by the uniform melting of one tonne
(1000 kg) of ice from and at 0°C in 24 hours.

Since the latent heat of ice is 335 kJ/kg, therefore one tonne of refrigeration,

1TR = 1000  335 kJ in 24 hours

1000  335
= = 232.6 kJ/min
24  60

In actual practice, one tonne of refrigeration is taken as equivalent to 210 kJ/min or

3.5 kW (i.e. 3.5 kJ/s).

332