Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Energy Conversion and Management 51 (2010) 12981303

Contents lists available at ScienceDirect

Energy Conversion and Management


journal homepage: www.elsevier.com/locate/enconman

Thermal performance of cross ow cooling towers in variable wet bulb temperature


Ebrahim Hajidavalloo a,*, Reza Shakeri b, Mozaffar A. Mehrabian b
a
b

Mechanical Engineering Department, Shahid Chamran University, Ahvaz, Iran


Mechanical Engineering Department, Shahid Bahonar University of Kerman, Kerman, Iran

a r t i c l e

i n f o

Article history:
Received 3 October 2008
Received in revised form 11 June 2009
Accepted 9 January 2010
Available online 4 February 2010
Keywords:
Variable wet bulb
Cooling tower
Mathematical model
Thermal performance
Impact separator

a b s t r a c t
Cooling towers are widely used in most industrial units to reject waste heat to the atmosphere. Wet towers are usually designed to operate in hot and dry weather conditions with narrow range of wet bulb temperature, but many cooling towers are required to operate in weather condition with large variation of
wet bulb temperature which strongly affects the thermal performance of the towers. In this paper a conventional mathematical model is used to predict the thermal behavior of an existing cross ow tower
under variable wet bulb temperature and the results are compared with experimental data in various
operating conditions. Available ll characteristic curve of the tower is obtained to estimate its departure
from the design conditions. It is found that when the wet bulb temperature increases, the approach, range
and evaporation loss would increase considerably. Variation of evaporation loss versus wet bulb temperature was estimated. Finally the effect of placing an impact separator in front of air louvers on thermal
performance of the tower is investigated.
2010 Elsevier Ltd. All rights reserved.

1. Introduction
Cooling towers are the heat and mass transfer devices being in
widespread use. Due to their important role, different kinds of
cooling towers have been introduced to address the various demands of industries. Different mathematical models have been
developed to predict the thermal behavior of wet cooling towers.
The rst practical model to describe the heat and mass transfer
mechanisms in wet cooling towers was proposed by Merkel [1].
Using Merkels theory, most of the studies have paid more attention to analyze the counter ow towers compared to the cross ow
towers. The reasons for the lack of studies on the cross ow towers
are the widespread use of counter ow towers and also the difculty in the analysis of cross ow towers as compared to the counter ow towers. Snyder [2] applied the theory of heat exchanger
design to calculate the driving force of a cross ow tower in the
same way as was used to calculate the mean temperature difference in a cross ow heat exchanger and obtained the overall enthalpy transfer coefcient. He assumed a linear relationship
between the water temperature and enthalpy of saturated air. Zivi
and Brand [3] solved the differential equations numerically using a
non-linear relationship between the water temperature and enthalpy of saturated air. Schechter and Kang [4] applied the Zivi
and Brands method to more general operating conditions by representing an exponential function to express the equilibrium relation between the water temperature and enthalpy of saturated air
* Corresponding author. Tel.: +98 611 3738532; fax: +98 611 3369684.
E-mail address: hajidae_1999@yahoo.com (E. Hajidavalloo).
0196-8904/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2010.01.005

at a limited range. Baker and Shryock [5] proposed an integral solution based on Merkels theory.
Poppe and Rogener [6] developed a new model for cooling towers which did not use the simplifying assumptions made by Merkel. The critical differences between Merkel, Poppe and e-NTU
models were investigated by Kloppers and Kroger [7]. They concluded that when the water outlet temperature is the only important parameter to the tower designer, the less accurate Merkel and
e-NTU approaches can be used but when the heat transfer rates are
concerned; they give lower values than that predicted by Poppe
approach. Hayashi and Hirai [8] approximated the enthalpy of saturated air by a rst-order equation with respect to the water temperature, and applied the cross ow heat exchanger calculations to
obtain the overall enthalpy transfer coefcient by using a chart.
Inazumi and Kageyama [9] proposed a graphical method for calculation of the enthalpy driving force in a cross ow cooling tower.
Khan and Zubair [10,11] considered the effect of Lewis number
and heat transfer resistance in the airwater interface and developed a detailed model for counter ow wet cooling towers. Halasz
[12,13] developed a general mathematical model to describe the
thermal characteristics of all types of evaporative cooling devices.
The main feature of this model is its non-dimensionality which
efciently reduces the required parameters to analyze an evaporative device. He then applied his model to predict the thermal
behavior of wet cooling towers and compared the model results
with an accurate model. Kairouni et al. [14] applied the Halaszs
model to predict the thermal performance of cooling towers in
south Tunisia. Prasad [15] developed a numerical model for cross
ow wet cooling towers and applied the model to estimate the

E. Hajidavalloo et al. / Energy Conversion and Management 51 (2010) 12981303

1299

Nomenclature
Av
cw
hm
H
L
_
m
t
x,y
V
FC

surface area of water droplets per unit volume of tower,


m2 m3
specic heat of water at constant pressure, J kg1 C1
mass transfer coefcient, kg m2 h1
enthalpy, J kg1
length of the tower, m
mass ow rate per square meter of the tower,
kg m2 h1
temperature, C
coordinates shown in Fig. 1
volume of the selected tower, m3
ll characteristic = hmm_Awv V

departure of available ll characteristics (FCav) of the packing of a


multi-cell cross ow cooling tower from their values at design
state.
In spite of vast application of cross ow cooling towers in industries, there are limited investigations to address the effect of large
variation of wet bulb temperature on the performance of this type
of cooling towers and most researches are devoted to the counter
ow cooling towers. Moreover, since cross ow cooling towers
have large inlet area for air as compared to counter ow cooling
towers, therefore, more pollution, and the way of cleaning the air
before tower is very important in this type of towers. This matter
has not been discussed yet.
In this study, the conventional Merkels model is used to analyze the thermal behavior of the tower at different wet bulb temperatures for an existing cooling tower working in south of Iran
and located in steel company in Ahvaz city. Ahvaz city has variable
wet bulb temperature due to its closeness to the Persian Gulf in the
Middle East. When the weather gets humid, the performance of
cooling towers deteriorates considerably. Experimental test was
carried out to validate the predicted results. Moreover, impact type
separator is introduced as an effective way to prevent polluted and
dusty air from entering the tower.

2. Mathematical modeling
Merkels model is used to investigate the behavior of the tower.
The basic assumptions of this model are:
1. The heat transfer resistance of the liquid lm is negligible.
2. The mass ow rate of water per unit cross sectional area of the
tower is constant (neglecting the mass of evaporated water).
3. The specic heat of moist air at constant pressure is the same as
that of dry air.
4. Lewis number for moist air is unity.
According to the Merkels theory, all the heat and mass transfer
occurring at each point of the cooling tower can be treated as a

Water

dx

Air
dy

Fig. 1. A differential element of a cross ow cooling tower.

_w
ow ratio, water to air = m
_a
m

FR

Subscripts
a
air
available
av
db
dry bulb
i
inlet
o
outlet
s
refers to saturated air
wb
wet bulb temperature
w
water

single transfer process with enthalpy difference as the driving


force.
Unlike the analysis of counter ow tower which is one dimensional, the cross ow tower must be treated as a two dimensional
system because there are variation of temperature and humidity
both in vertical and horizontal directions. Considering a differential
element of a cross ow cooling tower (Fig. 1), the energy balance
equation inside the tower is:

_ w cw dtw dx m
_ a dHa dy hm Av dxdyHs  Ha 
m

where, dxdy is the volume of the element, with its width assumed
unity.
Rearranging Eq. (1) results in the following set of PDEs for the
variations of water temperature and air enthalpy throughout the
tower:

_ w cw
m

_a
m



@t w
hm Av Hs  Ha
@y



@Ha
hm Av Hs  Ha
@x

The boundary conditions are:

tw x; 0 twi

Ha 0; y Hai

The relation between water temperature and enthalpy of saturated air [16] is:

Hs 4:7926 2:568t w  0:029834t2w 0:0016657t 3w

The governing equations (Eqs. (2) and (3)) in conjunction with


Eq. (6) are coupled and non-linear, which should be solved simultaneously. Finite difference technique is used to solve the set of
governing equations to nd air and water properties in each point
of the tower.
3. Tower specications and required characteristic curve
Fig. 2 shows a schematic diagram of the cooling tower, it is a
cross ow cooling tower with three cells (six cell halves). The lls
are splash type with rectangular cross section and made from redwood. The design conditions of the tower are listed in Table 1.
The required characteristic curve (FCr) of the tower for specied
conditions is shown in Fig. 3. In this gure two characteristic
curves are drawn, one, which is used frequently, is based on the
assumption that the inlet air is saturated at its wet bulb temperature (RH = 100%). The other one is based on real relative humidity
of inlet air (RH = 22%). As the two curves almost coincide, assuming
saturated inlet air is reasonable.

1300

E. Hajidavalloo et al. / Energy Conversion and Management 51 (2010) 12981303

Water inlet

material doesnt follow its design curve and it is required to be obtained experimentally. Obtaining this curve has two main
advantages:

Water inlet
Drift eliminators

Air inlet

_w
1. Predicting the thermal behavior of the tower with varying m
_ a , which helps the user to nd the optimum operating
and m
point of the tower at present conditions.
2. Estimating the departure of FCav from design conditions, this
feature helps the user to nd the percentage of degradation of
packing material, which can be used in maintenance program
of the tower.

Air inlet
Fill

Fill

Fig. 2. Schematic diagram of the selected tower.

Table 1
Design conditions of the selected tower.
Mass ow rate of water
Mass ow rate of air per fan
Inlet water temperature
Outlet water temperature
Wet bulb temperature of inlet air
Expected evaporation loss
Cell half dimensions
Length
Height
Width

3429734.6
1927932.5
58.0
30.0
24.0
5.0

kg/h
kg/h
C
C
C
%

5.5
9.6
12.0

m
m
m

RH=100%
RH=22%

2.5

Available ll characteristic curve for one cell-half of the tower


has been obtained using a numericalexperimental method
described by Prasad [15]. He used measured values of two,min and
two,max in order to predict FCav of the tower, since these values
are unique functions of FCav for a given set of wet bulb and inlet
water temperatures. For this purpose a set of experimental data
at four various operating conditions of the selected tower has been
obtained which is shown in Table 2.
In order to determine FCav after a period of service, Eqs. (2) and
(3) are used together with measured values of two,min and two,max. A
two-way iteration is performed, one for adjusting unknown FR and
the other for modifying unknown FCav, until the computed results
for two,min and two,max match with their respective measured values.
Each of the temperatures two,min and two,max represent the average
of a number of measurements recorded at equidistant points on
outer and inner ends of the ll bottom. Measurement at different
values of FR, which itself is an unknown, and computing the corresponding values using the developed computer program establishes current relationship between FR and FC. Available ll

FCr

Table 2
Measured data at four operating conditions of the tower.

twi (C)

two,min (C)

two,max (C)

twb(C)

tdb (C)

FR

FC

55.0
55.0
55.0
52.0

24.6
26.1
27.2
28.3

29.3
30.7
32.2
33.7

22.0
21.0
21.5
23.0

37.9
39.2
41.8
46.0

0.43
0.47
0.50
0.60

2.20
1.67
1.45
1.28

1.5

2.2
0.2

0.4

0.6

.
m
FR = .w
ma

0.8
2

Fig. 3. Required characteristic curves of the tower at two different conditions of


inlet air.

FCav

4. Available ll characteristic curve (FCav)


In order to simulate the existing cooling tower, it is required to
obtain the current ll characteristic curve of the tower. Fill characteristic is a non-dimensional parameter in each wet cooling tower,
which represents the overall potential of the ll to cool water and
is written as:

FCav

hm Av V
_w
m

1.8

FCav curves for different packing shapes and materials are obtained by ll suppliers using appropriate tests. After a length of
service, the FCav value of packing may diminish due to several reasons, like ll damage and water misdistribution. Therefore, the ll

1.6

1.4

1.2

0.45

0.5

0.55

m
FR = .w
ma
Fig. 4. Available ll characteristic curve for cell-half.

0.6

1301

E. Hajidavalloo et al. / Energy Conversion and Management 51 (2010) 12981303

5. Results and discussion


After nding the existing ll characteristic of the tower, the performance of the tower can be predicted at different conditions
using the mathematical model.
Variations of air enthalpy and water temperature through the
packing are shown in Fig. 5, which is in agreement with results reported in [5]. Hot water at temperature of 58 C enters from the top
and is cooled as it falls downward. The solid lines represent constant water temperatures. Air with 24 C wet bulb temperature enters from the left, across the OY axis, and is heated as it moves to
the right. The dotted lines show constant air enthalpies in the
tower. This gure shows that air and water properties vary in X
and Y directions, contrary to the counter ow tower where the
properties only depend on Y direction.
Table 3 presents the model predictions and experimental data
from cooling tower at eight various operating conditions. The relative error of the model predictions when compared with experimental results is less than 8%. Fig. 6 shows the model predictions
and experimental data for water outlet temperatures.
Having conrmed the accuracy of the model and cooling tower
characteristics, we can use the model to study the effect of other
parameters on the performance of the existing cooling tower.

Table 3
Comparing the model predictions of outlet water temperature with experimental
data at different wet bulb temperatures.
twb
(C)

tdb
(C)

twi
(C)

two,min
(C)

two,max
(C)

two,ave,exp
(C)

two,ave,Merkel
(C)

Error
(%)

20.5
21.0
21.6
22.0
22.0
23.0
24.0
26.0

44.0
36.0
44.0
42.0
42.0
40.0
38.0
38.0

48.0
41.0
48.0
48.0
49.0
38.0
50.0
42.0

24.2
24.5
25.0
24.7
24.1
25.2
26.4
29.1

27.4
28.3
26.8
28.3
28.0
28.5
30.0
33.5

25.8
26.4
25.9
26.5
26.0
26.9
28.2
31.3

27.1
26.3
27.7
27.9
28.0
26.8
29.3
29.4

5.05
0.49
6.76
5.29
7.67
0.13
3.82
6.22

35

Water outlet temperature (C)

characteristic values obtained from the illustrated numericalexperimental method are plotted against the respective FR for
the tower in Fig. 4.
To estimate the percentage of degradation of packing characteristic from the design condition, the current value should be compared with the design value obtained from Fig. 3. The design
values are: FRr = 0.587, FCr = 1.92. From Fig. 4, the available ll
characteristic at design ow ratio (FRr = 0.587) is FCav = 1.285. So
the degradation of the tower is about 33%.

30
25
20

10

41

42

48

48

48

49

50

38
37

Water outlet temperature (C)

38

Fig. 6. Comparing the model predictions of outlet water temperature with


experimental data at different wet bulb temperatures.

36

Merkel

Water inlet temperature (C)

Experiment

15

35

FR=0.6
FR=0.5
FR=0.4

34
33
32
31
30
29
28
27
26

25
20

25

30

35

Wet bulb temperature (C)


Fig. 7. Effect of ambient air wet bulb temperature on the water outlet temperature
of tower.

Y
0

10

Fig. 5. Variations of air enthalpy and water temperature through the packing.

One of the most important parameters that should be considered in the design and operating of wet cooling towers in mixed
weather conditions, like Ahvaz climate, is the effect of wet bulb
temperature on tower performance. In Ahvaz climate, the maximum dry bulb temperature in summer approaches 52 C, while
the wet bulb temperature is moderately low, around 24 C. But
sometimes this situation is changed and weather becomes humid

1302

E. Hajidavalloo et al. / Energy Conversion and Management 51 (2010) 12981303

with 100% relative humidity and wet bulb temperature reaches


around 35 C. This large change in the wet bulb temperature has
an important effect on the tower performance and the tower does
not work efciently.
Fig. 7 shows the effect of wet bulb temperature on water outlet
temperature at different FRs. The gure shows that increasing the
wet bulb temperature will increase the water outlet temperature.
The rate of increase is higher as the wet bulb temperature increases. This means that the tower approach decreases as the

5.5

t50
db=50C
t46
db=46C
t42
db=42C

+
+

+
+

4.5

80

70
+

FR=0.9
FR=0.7
FR=0.5

60
+

50

3.5

20

25

30

35

Wet bulb temperature (C)


Fig. 8. Effect of ambient air wet bulb temperature on the evaporation loss of the
tower.

Air enthalpy (kJ/kg)

Evaporation loss (% of circulating water)

wet bulb temperature increases. This is in agreement with the


same result reported for counter ow cooling tower in [11].
Fig. 8 shows the effect of wet bulb variations on the evaporation
loss of water at different dry bulb temperatures. This gure shows
that increasing the wet bulb temperature, decreases the evaporation rate of water considerably. It also shows that increasing the
dry bulb temperature at constant wet bulbs, increases the evaporation rate. The rate of increase in the evaporation rate at different
dry bulb temperatures is almost constant as the wet bulb temperature increases. Comparing these results with the evaporation loss
data at design conditions (Table 1), shows that the evaporation loss
at design conditions is only consistent at dry bulb temperature of
42 C, which is not a good estimate for the tower, since the ambient
air temperature reaches to a maximum of 52 C with summer average temperature of 46 C.
The effect of wet bulb temperature on water temperature distribution along the tower has been investigated at four hot summer
days in Ahvaz and shown in Table 4 and Fig. 9. The gure shows

40
30
20
10
00
90

Table 4
Effect of wet bulb on outlet water temperature at four hot summer days in Ahvaz.
Date

Time

tdb (C)

twb (C)

two,ave (C)

2007/6/25
2007/7/31
2007/7/26
2007/8/28

5:30
6:30
7:30
6:30

27.0
31.0
31.0
30.0

18.6
19.8
27.2
28.4

27.4
28.0
31.9
32.7

X (m)
Fig. 10. Effect of FR on air enthalpy variations across the cell-half.

twb=28.4
28.44
27.19
twb=27.2
19.84
twb=19.8
18.59

50

twb=18.6

45

40

35

Water outlet temperature (C)

55

55

Water outlet temperature (C)

80

FR=0.6
FR=0.5
FR=0.4

50

45

40

35

30
30

25

Y (m)
Fig. 9. Effect of wet bulb on water temperature distribution along the tower.

Y (m)
Fig. 11. Effect of FR on temperature distribution along the cell-half.

E. Hajidavalloo et al. / Energy Conversion and Management 51 (2010) 12981303

1303

suspended in its immediate surrounding. Dust deposit on the inlet


louvers and packings of cooling tower creates a thick layer of scale
around these parts after a while as shown in Fig. 12. In order to reduce entering dust and suspended solids in the tower, an impact
separator was proposed to place in front of air louvers of the tower.
High collection efciency [17], simple construction and low cost,
stable operation, low pressure drop, and easy scale up are the major attractions of impact separators for applications in cooling towers. The effect of putting a U-shaped impact separator with 5%
reduction in air ow rate, on the outlet water temperature of the
tower has been shown in Fig. 13. This gure shows that the impact
separator has no important effect on increasing the outlet water
temperature and may be used easily in front of the towers to lter
dusty air.
7. Conclusions
Fig. 12. Layer of dust scale on the body and packing of the tower.

A mathematical model is used to simulate the effect of any


change in operating conditions of cooling tower, especially the
wet bulb temperature, on the thermal performance of a cross ow
tower. Available characteristic curve for packing material has been
obtained using a numericalexperimental method which provides
an insight on the current performance of the tower. It is found that
increasing the wet bulb temperature, at constant dry bulb, will decrease the approach, range and evaporation loss in the tower considerably. The evaporation rate is increased as the dry bulb
temperature increases and the rate of increase is almost constant
at different wet bulb temperatures. An impact separator could be
used as a reasonable solution to reduce the amount of suspended
solids in the air without any considerable loss in the tower
performance.

30

Water outlet temperature (C)

29

28

27

26

24
38

References

without impact separator


with impact separator

25

40

42

44

46

48

50

Water inlet temperature (C)


Fig. 13. Effect of impact separator on water outlet temperature of cooling tower.

that increasing the wet bulb temperature is more effective in the


bottom section of the tower.
Fig. 10 shows the effect of various FR on the air enthalpy across
the tower, at design conditions. The gure shows that increasing FR
will increase the air enthalpy at any position in the cooling tower.
The effect of FR on water temperature along the tower has been
shown in Fig. 11. The gure shows that the temperature of water is
increased when FR is increased. This can be explained from the fact
that an increase in FR, means that more water should be cooled for
a given tower volume. Therefore, one would expect that the surface
area required both for convection and evaporation will be reduced,
resulting in higher water outlet temperatures.
6. Effect of impact separator on tower performance
Many cooling towers working at polluted areas, suffer from
scale forming as a result of entering considerable amount of dust

[1] Merkel F. Evporative cooling. Z Verein Deutsch Ingen (VDI) 1925;70:1238.


[2] Snyder NW. CEP Sympos Ser 1956:6179.
[3] Zivi SM, Brand BB. An analysis of the cross ow cooling tower. Refrig Eng
1956;64:314.
[4] Schecheter RS, Kang TL. Ind Eng Chem 1959;51:137384.
[5] Baker DR, Shryock HA. A comprehensive approach to the analysis of cooling
tower performance. J Heat Transf 1961;83:33949.
[6] Popp M, Rogener H. Calculation of cooling process. VDI-Warmeatlas; 1991. p.
Mi 1Mi 15.
[7] Kloppers JC, Kroger DG. Cooling tower performance evaluation: Merkel, Poppe,
and e-NTU methods of analysis. J Eng Gas Turbines Power 2005;127:17.
[8] Hayashi Y, Hirai E. Analysis of a multi-unit co current cross ow cooling tower.
J Heat Transf 1974;3:6774.
[9] Inazumi H, Kageyama S. A successive graphical method of design of a cross
ow cooling tower. Chem Eng Sci 1974;30:71721.
[10] Khan JR, Zubair SM. Performance characteristics of counter ow wet cooling
towers. Energy Convers Manage 2002;44:207391.
[11] Khan JR, Zubair SM. An improved design and rating analyses of counter ow
wet cooling towers. J Heat Transf Trans ASME 2001;123:7708.
[12] Halasz B. A general mathematical model of evaporative cooling devices. Int J
Therm Sci 1998;37:24555.
[13] Halasz B. Application of a general non-dimensional mathematical model to
cooling towers. Int J Therm Sci 1999;38:7588.
[14] Kairouani L, Hassairi M, Tarek Z. Performance of cooling tower in south of
Tunisia. Build Environ 2003;39:3515.
[15] Prasad M. Economic upgradation and optimal use of multi-cell cross ow
evaporative water cooling tower through modular performance appraisal.
Appl Therm Eng 2003;24:57993.
[16] Stoecker WP, Jones JW. Refrigeration and air conditioning. New York: McGraw
Hill; 1983.
[17] De S, Lal AK, Nag PK. An experimental investigation on pressure drop and
collection efciency of simple plate-type impact separator. Powder Technol
1999;106:1928.

You might also like