Energy Conversion and Management 126 (2016) 971–981

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Experimental investigation on performance enhancement of forced draft

wet cooling towers with special emphasis on the role of stage numbers
Mehdi Rahmati a,b, Seyed Rashid Alavi a,b,⇑, Mohammad Reza Tavakoli a
a
Department of Mechanical Engineering, Isfahan University of Technology, Isfahan 84156-83111, Iran
b
Isfahan Mathematics House, Isfahan 81698-51177, Iran

a r t i c l e i n f o a b s t r a c t

Article history: Wet cooling towers play a prominent role in the transfer of waste heat from industrial thermodynamic
Received 1 June 2016 cycles. However it is a complicated task to dictate and optimize their operational conditions on account
Received in revised form 26 July 2016 of different mass and heat transfer mechanisms between water and air flows. In this research, air mass
Accepted 20 August 2016
flow rate, hot water temperature, water flow rate and stage numbers of packing are changed experimen-
tally in order to investigate their influences on the cooling tower performance. As a matter of fact, this
research attempts to focus on packing density effects with an emphasis on the role of stage numbers
Keywords:
of packing. Consequently, the obtained results indicate that coefficient of efficiency is in direct relation
Wet cooling tower
Packing stages
with the hot water temperature, stage numbers of packing and air mass flow rate, while it diminishes
Heat and mass transfer by increasing the water flow rate. On the other hand, the captured thermal images and pertinent
Thermal image temperature histograms are investigated so as to visualize and calculate the temperature distribution
Regression analysis within the studied cooling tower. Eventually, mathematical equations are derived from regression anal-
Optimum performance ysis of the measured data to recommend some practical procedures for the sake of reaching optimum
operating conditions of forced draft wet cooling towers.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Enhancing the dissipation process of unwanted thermal energy
is one of the most prominent tasks in the power plants. For this
sake, cooling towers (CTs) are utilized almost everywhere which
operate based on mass and thermal energy transfer from hot water
to coolant air. According to the heat transfer mechanism between
the hot water and coolant air flow, industrial cooling towers are
divided into two main types of dry and wet cooling towers (DCTs
and WCTs). In particular, WCTs operate based on direct interface
between the hot water and coolant air flow which is caused by
the water flow over the packing. Thus, WCTs utilize evaporation
as a latent heat transfer mechanism as well as the sensible heat
transfer mechanism (the only heat transfer mechanism in DCTs).
It is worth mentioning that WCTs are more efficient in hot weather
regions while DCTs are employed in low water areas that suffer
from lack of water. Operation of WCTs generally depends on many
parameters such as hot water temperature, air mass flow rate,
water flow rate, packing type, etc. Thus, engineers can enhance
performance of WCTs through optimizing the mentioned factors.
⇑ Corresponding author at: Department of Mechanical Engineering, Isfahan
University of Technology, Isfahan 84156-83111, Iran.
E-mail address: r.alavi@me.iut.ac.ir (S.R. Alavi).
In recent years, a large volume of studies have been carried out
on thermal performance of WCTs theoretically and experimentally.
Facao and Olliveira [1] determined thermal and mass energy
transfer coefficients experimentally. They presented the thermal
performance of WCTs for the purpose of chilling. Goshayshi and
Missendn [2] studied the mass transfer coefficient of various pack-
ings experimentally, and found out the dependency of packing
coefficient on distance and pitch of packings. Kloppers and Kroger
[3] proposed a model to relate thermal and mass energy transfer
processes to WCTs. In addition, influence of the Lewis factor on
thermal energy transfer performance of natural and forced draft
wet cooling towers was studied by Kloppers and Kroger [4]. Soon
after, Elsarrag [5] investigated the effects of ceramic tile packing
on the cooling tower performance through relevant experiments.
He also proposed a model to predict the outlet water and air tem-
perature conditions. Smrekar et al. [6] found out that efficiency of
the natural draft cooling tower may be enhanced by optimization
of thermal energy transfer across the cooling tower. In the sequel,
they suggested correlative relations to reach an optimum ratio of
water to air flow rates. Lemouari et al. [7] conducted some exper-
iments on heat transfer performance of a cooling tower utilizing
different parametric studies which yielded similar model to that
of Gharagheizi et al. [8]. Al-Waked and Behnia [9] investigated
the influence of windbreak walls on the thermal performance of

http://dx.doi.org/10.1016/j.enconman.2016.08.059
0196-8904/Ó 2016 Elsevier Ltd. All rights reserved.
972 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981
Nomenclature
C p;a specific heat capacity of air, kJ kg
1 K
1
ha1 enthalpy of inlet air, kJ kg
1
ha2 enthalpy of outlet air, kJ kg
1
m_a air mass flow rate, kg s
1
n size of the actual output
Q_ air cooling capacity, kW
Q_ K cooling capacity by convection, kW
Q_ V cooling capacity by evaporation, kW
qw water flow rate, m3 h
1
x2 outlet humidity ratio, kgw kg
1a
natural draft wet cooling towers (NDWCTs) under crosswind
through the three dimensional CFD modeling. They concluded that
the NDWCT performance could be enhanced by installing solid
walls at the NDWCT entry. Muangnoi et al. [10] investigated effects
of the inlet temperature and humidity on performance of a counter
flow WCT based on exergy analysis and the second law of thermo-
dynamics. Lemouari et al. [11] experimentally studied the mass
and heat transportation between air and water flow rates. They
suggested modification of the generalized mass and heat transfer
coefficients through involving the influence of water and air trans-
portation. In addition, Marques et al. [12] compared open/closed
loop efficiency of a counter flow WCT. Their results showed a
remarkable improvement in the performance of closed loop WCT
in comparison with open-loop operation. Heidarinejad et al. [13]
simulated a counter-flow wet cooling tower, and a good agreement
was reached between their results and experimental data in the lit-
erature. They concluded that it is important to consider the rain
zones and spray in analyses for enhancing the accuracy. In addi-
tion, Sarker et al. [14] investigated other properties of cooling tow-
ers including cooling capacity, pressure loss and efficiency. They
concluded that using the finned tube leads to higher thermal effi-
ciencies due to pressure loss enhancement in comparison with the
bare tube. Furthermore, the heat and fluid flow properties of
evaporation coolers were studied by Heyns and kröger [15].
Dependency of the thermal energy coefficient for water layer on
the spraying water temperature and mass flow rate of air and
water was demonstrated by their measured results. Zhang et al.
[16] provided a theoretical foundation for applied design by
proposing an analytical model for the coupled heat and mass trans-
fer procedures in reversibly used cooling towers, and a good agree-
ment was reached between their proposed model and the reported
experimental results. Moreover, effects of weather conditions on
the thermal properties of WCTs were studied by Papaefthimiou
et al. [17], and a thermodynamic model was proposed in order to
simulate the processes in WCT. It is notable that a good agreement
was obtained between their simulation results and the experimen-
tal results. Zheng et al. [18] found out utilization possibility of oval
tubes in the WCTs and suggested some correlations for optimum
operating conditions. Furthermore, a criterion of drift eliminators
design was developed by Lucas et al. [19] in order to enhance the
thermal performance of cooling towers. They utilized the chemical
balance method and carried out an experiment to study the drift
loss from a cooling tower without and with different drift elimina-
tors. Jiang at al. [20] carried out experimental and numerical anal-
yses of a cross-flow closed wet cooling tower based on finned tube
assembly. They also attempted to enhance the thermal perfor-
mance by comparing the bare-tube with the fin-tube patterns. In
addition, influences of rotational splash type packing on the oper-
ating conditions of forced draft counter flow WCTs were studied by
Lavasani et al. [21]. Their results demonstrated that higher veloci-
ties of rotational splash type packing lead to enhance the heat
r latent heat of evaporation, kJ kg
1
SN stage numbers of packing
T w1 hot water temperature, °C
T w2 outlet water temperature, °C
t db1 dry-bulb temperature of inlet air, °C
t db2 dry-bulb temperature of outlet air, °C
t wb1 wet-bulb temperature of inlet air, °C
t wb2 wet-bulb temperature of outlet air, °C
x1 inlet humidity ratio, kgw kg
1
a
transfer process. Ning et al. [22] investigated the WCT performance
through considering influences of the defects, including conditions
of nozzles drop, nozzles blockage and packing blockage. In partic-
ular, their results revealed more than 60% decreasing in the tower
characteristic ratio under nozzles drop conditions. Askari et al. [23]
investigated effects of multi-walled carbon nanotubes and nano-
porous graphene nanoparticles on the thermal performance of a
WCT through an experimental procedure. They studied thermal
conductivity of nanofluids and observed 20% and 16% improve-
ment compared with the base fluid in thermal conductivity of
nanofluids for multi-walled carbon nanotubes and nanoporous
grapheme nanofluids, respectively. In addition, Singh and Das
[24] analyzed performance of a forced draft WCT using trickle, film
and splash fills. They developed relevant correlations based on
experimental data for performance parameters by considering
mass flow rates of water and air as design variables. Moreover,
they employed Non-Dominated Sorting Genetic algorithm in order
to optimize performance of forced draft WCT. Shahali et al. [25]
experimentally investigated effects of different input parameters
such as water flow rate and air mass flow rate on thermal perfor-
mance of a WCT. They proposed a guideline to attain the optimum
operating conditions of WCT by the use of regression analysis.
Moreover, Zhao et al. [26] presented a three dimensional numerical
model in order to study characteristics of a high level NDWCT.
They obtained air mass flow rate, air pressure drop, air tempera-
ture and some other parameters for high level NDWCT and usual
NDWCTs under the same operational conditions. Rahmati et al.
[27] employed experimental data so as to study the heat transfer
performance of natural draft counter flow WCTs under cross-
wind and windless conditions. They investigated the effects of inlet
water temperature, water flow rate and cross-wind velocity on the
water temperature difference and cooling efficiency.
On the other hand, enhancing the overall performance of engi-
neering systems has been a major purpose of designers. Accord-
ingly, a large number of investigations have been conducted so
as to enhance technical performance of engineering systems such
as power systems, heat pumps, heat exchangers, and data storage
systems through experiments or mathematical modeling [28]. In
particular, a large number of studies have been carried out in order
to enhance the technical performance of cooling towers by
evaluating influences of different parameters and procedures like
rib numbers, alternative shell geometry, and radiator-type wind-
breakers [29].
To the best of our knowledge, no investigation has been
performed on the WCT performance under simultaneous effects
of hot water temperature, water flow rate, air mass flow rate and
stage numbers of packing, particularly by virtue of the captured
thermal images. Thus, effects of these parameters on the WCT per-
formance are experimentally investigated in this study. First of all,
the present study attempts to portray the measuring setup and test
conditions. Then, the measured results are illustrated in order to
 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981
consider influences of water flow rate, hot water temperature, air
mass flow rate and stage numbers of packing separately. After that,
the pertinent thermal images are captured in order to visualize the
temperature distribution within the studied cooling tower and
carry out further analyses. In the sequel, relevant temperature his-
tograms are plotted so as to elucidate the distribution percent of
each temperature inside the process region of the studied WCT.
Furthermore, the regression analysis is utilized to have a more
applicable investigation. Therefore, mathematical equations are
derived from the experimental results, and optimum operating
conditions of the employed WCT are attained. Finally, some practi-
cal procedures are proposed for engineers to improve operating
conditions of forced draft WCTs.
2. Experimental apparatus and test conditions
The experimental apparatus employed here is depicted in Fig. 1
which contains a single stage centrifugal air blower (1), air bleed
valve (2), air flow orifice (3), manometer (4) and overhead water
tank (5) with water filter (6). The experimental setup also contains
a water pump (7), inlet water thermometer (8), electric heater (9),
flow meter (10), flow control valve (11), discharge valve (12), lower
water tank (13), exhaust duct (14), wet and dry bulb thermometers
(15) and (18), water sprays (16), packing (17) and outlet water
thermometer (19). In addition, the utilized cooling tower had
dimensions of 30 cm  30 cm  125 cm (L  W  H), and was
filled with the VGA (vertical grid apparatus) packings. It should
be noted that three types of PVC packings were mounted
separately (each packing has the height of 95 cm). Furthermore,
Fig. 1. Schematic of experimental apparatus.
973
four electric elements; each has heating rate of 2.75 kJ s
1, are
placed above the flow meter to generate the required heating load.
In more detail, the experimental setup contains a water tempera-
ture controller responsible for adjusting and fixing the hot water
temperature precisely. Moreover, a feeder pump was utilized here
which is responsible for sending the outlet water to the overhead
water tank, and the single stage centrifugal air blower was
employed so as to produce the air mass flow rate.
On the other hand, copper-constantan thermocouple, psy-
chrometer and thermo hygrometer were employed for the sake
of measuring water temperature, dry and wet-bulb temperatures
of air, and air humidity, respectively so that the measured data
were attentively controlled through digital monitors. Table 1
presents further information about all measuring devices utilized
during the tests. During the tests, several adjusting valves and
buttons were responsible for fixing the water flow rate, hot water
temperature and air mass flow rate to their relevant set-points.
These values were controlled by the use of pertinent indicators
and digital monitors mounted on the setup.
This experimental work includes three major studies in which
the WCT performance was separately assessed based on variations
in the water flow rate, hot water temperature and stage numbers
of packing, respectively. For water flow rate, three rates of 0.06,
0.08 and 0.1 (m3 h
1) were investigated. In addition, three temper-
atures of 35, 40 and 45 (°C) were considered for hot water temper-
ature, and 5, 7 and 10 stages were used separately as illustrated in
Fig. 2. It should be noted that in all three studies the air mass flow
rate was varied from 0.02 to 0.07 kg s
1.
During the first study, the hot water temperature and stage
numbers of packing were fixed and adjusted to 35 (°C) and 7,
974 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981

Table 1 identical between model and its prototype, are very small and
Further information on measuring devices utilized during the tests. can be neglected.
Sensor Parameter Accuracy
Psychrometer Dry/wet bulb temperature of air 0:1 ½ C 3. Experimental results and discussions
Copper-constantan Water temperature 0:1 ½ C
thermocouple
In this section, the WCT performance will be analyzed in more
Thermo hygrometer Air humidity 0:1 ½%
Rotameter Water flow rate 0:5 ½% detail, and thus several variables should be considered. The water
Manometer Air pressure 0:01 ½kPa temperature difference is indicated by DT, which refers to the dif-
ference between inlet and outlet water temperatures.

DT ¼ T w1
T w2 ð1Þ
respectively. In the second study, water flow rate and stage num-
In addition, the parameter (e) is utilized to represent the coeffi-
bers of packing were kept constant and their values were adjusted
cient of efficiency which states the cooling limit of WCT.
to 0.1 (m3 h
1) and 7, respectively. In addition, the water flow rate
and hot water temperature were adjusted to 0.1 (m3 h
1) and 35 T w1
T w2
(°C), respectively during the last study. It is worth mentioning that e¼ ð2Þ
T w1
twb1
two or three tests were conducted for each set-point of experiment
and the corresponding mean values were obtained for output In which T w1 and T w2 are inlet and outlet temperatures of water,
parameters. respectively. In addition, twb1 designates the wet-bulb temperature
On the other hand, it is noteworthy to mention that the exper- of air at entry.
iment process was conducted on the basis of model-prototype sim- Moreover, the cooling capacity of air ðQ_ Þ consists of two main
ilarity. Therefore, after this investigation, the results could be components, attributable to evaporation (Q_ V ) and convection
extended for industrial cooling towers. In more detail, based on (Q_ K ) mechanisms. The cooling capacity and its components could
the similarity theory, the parameters and operational conditions be stated as:
of an industrial cooling tower could be converted to laboratory-
scale ones, and vice versa. In general, the mechanical similarity Q_ ¼ Q_ V þ Q_ K ð3Þ
requires three criteria including geometric, kinematic and dynamic Q_ ¼ m_ a ðha2
ha1 Þ ð4Þ
similarity. Based on the geometric similarity, all of the dimension
_
QV ¼ m _ a rðx2
x1 Þ ð5Þ
ratios should be of equal value for the model and prototype of
tower and packing. As well, the number of experimental packing Q_ K ¼ m
_ a C p;a ðt db2
t db1 Þ ð6Þ
stages can be obtained based on the geometric similarity. Accord-
ing to the dynamic similarity theory, all relevant force ratios Here ha1 , ha2 , t db1 , t db2 , x1 , x2 and r denote the enthalpy of inlet and
(including Froude number, Reynolds number, etc.) should be iden- outlet air, dry-bulb temperature of inlet and outlet air, inlet and
tical between model and its prototype; however, in practice, it is outlet humidity ratio and latent heat of evaporation, respectively.
impossible to implement and satisfy similarity for all of the rele- In addition, the parameter m _ a refers to the air flow rate, and the
vant force ratios simultaneously. Therefore, a tradeoff is necessary specific heat capacity of air is indicated by C p;a .
to find out which ratio is more dominant. Froude number is useful
for phenomena where gravity and inertial forces are dominant and 3.1. Effects of water flow rate on the WCT performance
effect of remaining forces such as kinematic viscosity are small. In
addition, Froude number is considered in most of water phenom- As illustrated in Figs. 3 and 4, dependency of the water temper-
ena, in particular surface flows in cooling towers. Consequently, ature difference (DT) and coefficient of efficiency (e) upon the air
as a steady state thermal experiment, it is important to have same mass flow rate was experimentally investigated for different water
density Froude numbers (Fr) for the model and actual WCTs for the flow rates. In this section, the hot water temperature and stage
sake of reaching a more realistic experiment. It should be noted numbers of packing were adjusted to constant values of 35 (°C)
that scale effects, effects caused by force ratios which are not and 7, respectively.

Fig. 2. Different stage numbers of packing employed in this research (a) 5 stages, (b) 7 stages, and (c) 10 stages.
 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981 975

ture difference increases by 1.67% after enhancing the air mass

flow rate from 0.05 to 0.055 kg s
1.
Moreover, it is clearly observed from Fig. 4 that at a given value
of air flow rate, the coefficient of efficiency decreases through
enhancement of the water flow rate. On top of that, similar to
the water temperature difference behavior, the reduction in coeffi-
cient of efficiency is not proportional to the water flow rate
enhancement. For instance, at the air mass flow rate of
0.025 kg s
1, about 14.24% reduction in the coefficient of efficiency
is observed after increasing the water flow rate from 0.06 to
0.08 m3 h
1, while 14.57% reduction is found out after enhancing
the water flow rate from 0.08 to 0.1 m3 h
1. Interestingly, despite
the observed trend of water temperature difference (Fig. 3), coeffi-
cient of efficiency steeply increases through enhancement of the
air flow rate.

3.2. Effects of hot water temperature on the WCT performance

Dependency of the water temperature difference and coefficient

of efficiency upon the air mass flow rate was also studied experi-
Fig. 3. Dependency of the water temperature difference upon air mass flow rate at
mentally for different temperatures of hot water as displayed in
different water flow rates.
Figs. 5 and 6. Meanwhile, in this section, the water flow rate and
stage numbers of packing were fixed and adjusted to constant val-
ues of 0.1 (m3 h
1) and 7, respectively.
According to Fig. 5, at a given value of air mass flow rate, the
water temperature difference is in the direct relation with the
hot water temperature. It is noteworthy to mention that this rela-
tion is not proportional. For example, at the air mass flow rate of
m_ a ¼ 0:025 kg s
1 , about 16.42% increase in the water temperature
difference is observed after increasing the hot water temperature
from 35 to 40 °C, while about 24.36% increase is found out after
enhancing the hot water temperature from 40 to 45 °C.
As with Fig. 3, it is also perceived from Fig. 5 that steep slopes
gradually convert to the gentle slopes by increasing the air flow
rate. This trend is observed owing to the fact that driving force of
energy transfer gradually decreases by increasing the air mass flow
rate. In addition, after the air flow rate of 0.04 (kg s
1), the increase
percent of DT due to the hot water temperature enhancement is
approximately invariant at any value of air flow rate. In particular,
the invariant increase percent of DT is perceived as about 20.5% for
the hot water temperature enhancement of 35–40 °C.
Furthermore, it is observed from Fig. 6 that at any value of air
flow rate, the coefficient of efficiency increases through
Fig. 4. Dependency of the coefficient of efficiency upon air mass flow rate at
different water flow rates.

According to Fig. 3, at any value of air mass flow rate, the water
temperature difference diminishes through enhancement of the
water flow rate. However this reduction is not exactly proportional
to the water flow rate enhancement. For instance, at the air mass
flow rate of m_ a ¼ 0:025 kg s
1 , about 19.56% reduction in the water
temperature difference is observed after increasing the water flow
rate by 33% at qw = 0.06 m3 h
1, while 9.46% reduction is found out
after enhancing the water flow rate by 25% at qw = 0.08 m3 h
1.
Based on Fig. 3, it is also evident that the steep increasing trend
of water temperature difference gradually converts to the slight
increasing trend. This trend is ascribed to variations in driving
force of energy transfer (heat and mass transfer) [30]. As a matter
of fact, the driving force of energy transfer process between water
and unsaturated air (enthalpy potential) gradually decreases by
increasing the air flow rate. As a good case in point for the above
trend, at the water flow rate of 0.06 m3 h
1, the water temperature
difference increases by 15.22% through enhancing the air mass Fig. 5. Dependency of the water temperature difference versus air flow rate at
flow rate from 0.025 to 0.03 kg s
1, whereas, the water tempera- different hot water temperatures.
976 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981

enhancement of the hot water temperature. Additionally, similar

to the water temperature difference behavior, the increase in coef-
ficient of efficiency is not proportional to the hot water tempera-
ture enhancement. For instance, at the air mass flow rate of
0.025 kg s
1, about 6.77% increase in the coefficient of efficiency
is observed after increasing the hot water temperature from 35
to 40 °C, while 12.5% increase is found out after enhancing the
hot water temperature from 40 to 45 °C. It is also notable that
the coefficient of efficiency constantly increases through enhance-
ment of the air flow rate.

3.3. Effects of stage numbers of packing on the WCT performance

In continuation, Figs. 7 and 8 delineate dependency of the water

temperature difference and coefficient of efficiency upon the air
mass flow rate at different stage numbers of packing where the
water flow rate and hot water temperature were fixed and
adjusted to constant values of 0.1 (m3 h
1) and 35 (°C), respec-
tively. According to Fig. 7, at a given value of air mass flow rate,

Fig. 6. Dependency of the coefficient of efficiency versus air flow rate at different
hot water temperatures.

Fig. 9. Coefficient of efficiency versus water flow rate at different mass flow rates
of air.
Fig. 7. Dependency of the water temperature difference versus air flow rate at
different stage numbers of packing.

Fig. 8. Dependency of the coefficient of efficiency versus air flow rate at different Fig. 10. Coefficient of efficiency versus hot water temperature at different mass
stage numbers of packing. flow rates of air.
 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981
Fig. 11. Coefficient of efficiency versus stage numbers of packing at different mass
flow rates of air.
the water temperature difference is in a direct relation with the
stage numbers of packing. In more detail, at the air mass flow rate
of 0.025 kg s
1, about 20.5% increase in the water temperature dif-
ference is observed after increasing the stage numbers of packing
from 5 to 7, while about 8.96% increase is found out after increas-
ing the stage numbers of packing from 7 to 10. Therefore, it can be
clearly understood that this relation is not proportional exactly.
Furthermore, it is also recognized from Fig. 7 that steep increas-
ing behavior of water temperature difference converts to a slight
increasing one, when air flow rate increases. It is notable that after
the air flow rate of 0.045 (kg s
1), the increase percent of DT due to
enhancing the stage numbers of packing is approximately invari-
ant at any value of air flow rate. In particular, the invariant increase
percent of DT is perceived as about 16% and 12% after increasing
the stage numbers from 5 to 7 and 7 to 10, respectively.
Fig. 12. Thermal images captured for (a) 5, (b) 7, and (c) 10 stage numbers of packing.
977
Moreover, according to Fig. 8, it is apparent that the coefficient
of efficiency enhances through increasing the stage numbers at any
value of air flow rate. Therefore, enhancing the stage numbers of
packing is a key factor found out here that translates into the
higher coefficients of efficiency and consequently superior operat-
ing conditions. This enhancement arises from the better heat trans-
fer owing to greater interface area between the coolant air flow
and hot water. It is also worth mentioning that the coefficient of
efficiency constantly increases through enhancement of the air
flow rate.
3.4. Investigation of efficiency
For further details, the coefficient of efficiency is investigated
based on varying the water flow rate, hot water temperature and
stage numbers of packing at different air mass flow rates, and
the obtained results are depicted in Figs. 9–11, respectively. These
figures confirm the inverse relationship between the water flow
rate and coefficient of efficiency. As well, these figures illustrate
that the coefficient of efficiency is in the direct relationship with
the air mass flow rate, hot water temperature and stage numbers.
It should be noted that in these figures, two remained parameters
are fixed at constant values mentioned above.
3.5. Thermal images
After reporting the experimental results illustrated in Sections
3.1 to 3.4, the pertinent thermal images are captured by an infrared
thermal camera (Testo 882-Thermal Imager, Testo AG, Lenzkirch,
Germany) in order to visualize the temperature distribution within
the studied WCT. For this purpose, three thermal images are cap-
tured as representatives of the experimental results and illustrated
in Fig. 12. In this figure, the air mass flow rate, hot water temper-
ature and water flow rate are considered as 0.07 kg s
1, 45 °C and
0.1 m3 h
1, respectively, while three stage numbers of 5, 7 and
10 were used separately. In addition, Testo IRSoft (Testo AG,
Lenzkirch, Germany) is a suitable software employed in this study.
It should be noted that this software is specially designed for
978 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981

tion percent of each temperature. It should be noted that Fig. 13

contains the temperature histograms relevant to the whole region
of Fig. 12a, b and c separately. From Fig. 13, it can be clearly under-
stood that the general trend of histograms is similar to the Gaus-
sian distribution. Accordingly, it seems reasonable to consider
the Gaussian distribution as temperature distribution inside the
process region of forced draft WCTs. Consequently, high and low
temperatures have lower distribution percent compared to the
middle-range temperatures. Additionally, it can be observed from
Fig. 13 that the histogram profile begins to shrink by increasing
the stage numbers of packing.

4. Regression analysis

Although the WCT performance was studied experimentally to

consider influences of various packing stages, water flow rate, hot
water temperature and air mass flow rate, this investigation may
be more practical by using mathematical equations derived from
regression analysis as a guide for improving WCT operating condi-
tions. Therefore, in what follows, the regression analysis is utilized
in order to provide functional relationships for correlating the
experimental data with known variables. It should be noted that
this technique is a powerful statistical method to analyze multi
variable data. In addition, the root mean square error (RMSE) and
correlation coefficient are two widely-used criteria utilized in this
research. These parameters are responsible for accuracy examina-
tion of regression outputs. The correlation coefficient demonstrates
quality of the data fit and linear relationship between regressed
and experimental outputs so that larger amounts of correlation
coefficient result in closer agreements between regressed and
experimental outputs, and thus stronger accuracies of regression.
This parameter could be given by [31]:

Cov ðY; YÞ ^
^ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
CorðY; YÞ ffi ð7Þ
Cov ðY; YÞ:Cov ðY; ^ YÞ ^

In which Cov ðY; YÞ

^ refers to covariance between the experimental
^ which is expressed as:
results ðYÞ and regression outputs ðYÞ,
X
n
Þðy
ðyi
y ^i
y
^Þ
Cov ðY; YÞ
^ ¼ i¼1
ð8Þ
n
1
where n indicates the number of data, y  is the mean value of
experimental results and y^ denotes the mean value of regression
outputs. In addition, Cov ðY; YÞ and Cov ðY;
^ YÞ
^ designate the auto
covariance of experimental results and regression outputs, respec-
tively and are expressed as:
Pn 2
^
y^Þi¼1 ðyi
Cov ðY;
^ YÞ
^ ¼ ð9Þ
n
1
Fig. 13. Temperature histograms relevant to the whole region of (a) Fig. 12a, (b) Pn
Þ2
ðy
y
Fig. 12b, and (c) Fig. 12c.
Cov ðY; YÞ ¼ i¼1 i ð10Þ
n
1
further analyses of thermal images captured by the Testo infrared As mentioned above, the parameter RMSE is another criterion
cameras. which could be utilized in accuracy examination of regression. This
In continuation, the captured thermal images are utilized so as parameter is defined as [32]:
to investigate the WCT performance. Histogram is a kind of graph rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 Xn
which has wide applications in statistics. In fact, a histogram RMSE ¼ ðy
y ^ i Þ2 ð11Þ
n i¼1 i
provides visual interpretation of numerical data and depicts the
frequency of data items in successive numerical intervals of equal
size. In other words, a histogram indicates the number of data 4.1. Correlative equation of the water temperature difference
points that lie within a range of values, called a class or a bin.
The frequency of the data that falls in each class is shown by the In what follows, according to the experimental results illus-
use of a bar. In this section, the temperature histogram of each trated in the previous sections, the regression analysis is carried
thermal image is plotted in Fig. 13 which elucidates the distribu- out in order to derive correlative equation of the water
 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981 979

Table 2 ferent input parameters. It should be noted that the stage numbers
Statistical parameters of Eqs. (12) and (13). of packing are considered as 10 in Fig. 14, and the water flow rate is
Evaluating parameters Evaluating parameters considered as 0.06 (m3 h
1) in Fig. 15.
of Eq. (12) of Eq. (13)
RMSE 0.1262 °C 0.5944% 4.2. Correlative equation of the coefficient of efficiency
Correlation coefficient 0.9891 0.9936

In continuation, the experimental results are also utilized to

reach correlative equation of the coefficient of efficiency versus
water and air flow rates, hot water temperature and stage numbers
temperature difference in terms of water and air flow rates, hot of packing on the basis of regression analysis. This correlative
water temperature and stage numbers of packing as follows: equation is obtained as:

_ a
0:1229 T w1
133:4304 qw
DT ¼ 4:8019 þ 115:0289 m e ¼
6:5974 þ 861:6723 m_ a þ 0:7269 T w1
78:5205 qw
_ 2a þ 1:5332 m
þ 0:5934ðSNÞ
1207:7182 m _ a T w1 _ 2a þ 2:7612 m

1:9207ðSNÞ
3715:2033 m _ a T w1
_ a qw þ 2:3003 m

386:3957 m _ a ðSNÞ þ 0:0033 T 2w1 _ a qw þ 16:9942 m

4151:5638 m _ a ðSNÞ
0:0049 T 2w1
2
þ 672:7273 q2w
0:0322ðSNÞ ð12Þ þ 151:1364 q2w þ 0:1786ðSNÞ2 ð13Þ
In which m _ a ; T w1 ; qw and SN refer to the mass flow rate of air, hot Similar to Eq. (12), the correlation coefficient and RMSE are also
water temperature, water flow rate and stage numbers of packing, obtained for Eq. (13) and presented in Table 2. Furthermore, the
respectively. In the sequel, the correlation coefficient and RMSE are coefficient of efficiency can be investigated through numerical
determined for the above equation to examine the regression accu- changes of the input parameters in Eq. (13), as illustrated in Figs. 16
racy. Table 2 contains the value of these parameters. In addition, and 17. The stage numbers of packing are considered as 10 in
according to the derived correlative equation (Eq. (12)), the water Fig. 16, and the water flow rate is considered as 0.06 (m3 h
1) in
temperature difference is illustrated in Figs. 14 and 15 versus dif- Fig. 17.

Fig. 14. Variation of the water temperature difference versus hot water temperature and air mass flow rate for different water flow rates, (SN = 10).

Fig. 15. Variation of the water temperature difference versus hot water temperature and air mass flow rate for different stage numbers of packing, (qw = 0.06 m3 h
1).
980 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981

Fig. 16. Variation of the coefficient of efficiency versus hot water temperature and air mass flow rate for different water flow rates, (SN = 10).

Fig. 17. Variation of the coefficient of efficiency versus hot water temperature and air mass flow rate for different stage numbers of packing, (qw = 0.06 m3 h
1).

From Eqs. (12) and (13), and Figs. 14–17, a perceptible point is
that appropriate optimum intervals exist and can be discerned for
the present study. In other words, optimum ranges of the air mass
flow rate, hot water temperature, water flow rate and stage num-
bers of packing are determined so as to improve the water temper-
ature difference and coefficient of efficiency. These optimum
ranges vary from 0.06 to 0.07 (kg s
1), 42–45 °C, 0.06 to 0.075
(m3 h
1), and 8–10 for the air mass flow rate, hot water tempera-
ture, water flow rate and stage numbers of packing, respectively.
Ultimately, some practical procedures are suggested for engi-
neers to improve operating conditions of forced draft WCTs. How-
ever spatial and economic restrictions should be considered in
addition to the following procedures. Enhancing the stage numbers
of packing is a key factor found out here that leads to the better
operating conditions. This enhancement arises from the better heat
transfer owing to greater interface area between the coolant air
flow and hot water. Moreover, increasing the air mass flow rate
translates into the higher evaporation and convection heat losses.
Thus, another procedure is to utilize a more powerful fan, as far
as possible. More to the point, decreasing the water flow rate is
another approach to upgrade the WCT performance. For the sake
of diminishing the water flow rate at the WCT entry, a helpful pro-
cedure could be increasing the number of WCTs in the plant.
5. Conclusions
In the foregoing research, the WCT performance was studied
experimentally to consider influences of water flow rate, hot water
temperature, air mass flow rate and stage numbers of packing,
with an emphasis on the role of stage numbers of packing. In addi-
tion, the pertinent thermal images were captured in order to visu-
alize the temperature distribution within the studied cooling
tower and carry out further analyses. Moreover, the regression
analysis was employed to derive mathematical equations from
the experimental results, and optimum operating conditions of
the employed WCT were attained. The main findings of this
research indicate that the coefficient of efficiency is in a direct rela-
tionship with the air mass flow rate, stage numbers of packing and
hot water temperature. On top of that, the less water flow rate is,
the more water temperature difference and coefficient of efficiency
are. As a consequence, the WCT performance could be improved by
larger amounts of the coefficient of efficiency which are reached at
lower water flow rates and higher values of hot water temperature,
stage numbers of packing and air mass flow rate. Furthermore, the
general trend of temperature histograms was similar to the
Gaussian distribution. Accordingly, it seems reasonable to consider
the Gaussian distribution as temperature distribution inside the
 M. Rahmati et al. / Energy Conversion and Management 126 (2016) 971–981
process region of forced draft WCTs. Additionally, it was apparent
that the temperature histogram profile begins to shrink by increas-
ing the stage numbers of packing. Ultimately, some practical pro-
cedures were suggested to improve the WCT performance.
Acknowledgement
The authors would like to express their sincere appreciation to
Prof. Mohsen Nasr Esfahany and Mr. Hamid Niknafs for their signif-
icant guides during the research and experiments conducted in
heat and mass transfer lab of Isfahan University of Technology.
References
[1] Facão J, Oliveira AC. Thermal behaviour of closed wet cooling towers for use
with chilled ceilings. Appl Therm Eng 2000;20(13):1225–36.
[2] Goshayshi HR, Missenden JF. The investigation of cooling tower packing in
various arrangements. Appl Therm Eng 2000;20:69–80.
[3] Kloppers JC, Kröger DG. A critical investigation into the heat and mass transfer
analysis of counter flow wet-cooling towers. Int J Heat Mass Transf
2005;48:765–77.
[4] Kloppers JC, Kroger DG. The Lewis factor and its influence on the performance
prediction of wet-cooling towers. Int J Therm Sci 2005;44(9):879–84.
[5] Elsarrag E. Experimental study and predictions of an induced draft ceramic tile
packing cooling tower. Energy Convers Manage 2006;47(15):2034–43.
[6] Smrekar J, Oman J, et al. Improving the efficiency of natural draft cooling
towers. Energy Convers Manage 2006;47:1086–100.
[7] Lemouari M, Boumaza M, Mujtaba IM. Thermal performances investigation of
a wet cooling tower. Appl Therm Eng 2007;27:902–9.
[8] Gharagheizi F, Hayati R, Fatemi S. Experimental study on the performance of
mechanical cooling tower with two types of film packing. Energy Convers
Manage 2007;48:277–80.
[9] Al-Waked R, Behnia M. Enhancing performance of wet cooling towers. Energy
Convers Manage 2007;48(10):2638–48.
[10] Muangnoi T, Asvapoositkul W, Wongwises S. Effects of inlet relative humidity
and inlet temperature on the performance of counterflow wet cooling tower
based on exergy analysis. Energy Convers Manage 2008;49(10):2795–800.
[11] Lemouari M, Boumaza M, Kaabi A. Experimental analysis of heat and mass
transfer phenomena in a direct contact evaporative cooling tower. Energy
Convers Manage 2009;50:1610–7.
[12] Marques CAX, Fontes CH, Embiruçu M, Kalid RA. Efficiency control in a
commercial counter flow wet cooling tower. Energy Convers Manage 2009;50
(11):2843–55.
[13] Heidarinejad G, Karami M, Delfani S. Numerical simulation of counter-flow
wet-cooling towers. Int J Refrig 2009;32(5):996–1002.
[14] Sarker MMA, Shim GJ, Lee HS, Moon CG, Yoon JI. Enhancement of cooling
capacity in a hybrid closed circuit cooling tower. Appl Therm Eng
2009;29:3328–33.
[15] Heyns JA, Kröger DG. Experimental investigation into the thermal-flow
performance characteristics of an evaporative cooler. Appl Therm Eng
2010;30:492–8.
981
[16] Zhang Q, Wu J, Zhang G, Zhou J, Guo Y, Shen W. Calculations on performance
characteristics of counterflow reversibly used cooling towers. Int J Refrig
2012;35(2):424–33.
[17] Papaefthimiou VD, Rogdakis ED, Koronaki IP, Zannis TC. Thermodynamic study
of the effects of ambient conditions on the thermal performance
characteristics of a closed wet cooling tower. Appl Therm Eng 2012;33–
34:199–207.
[18] Zheng WY, Zhu DS, Song J, Zeng LD, Zhou HJ. Experimental and computational
analysis of thermal performance of the oval tube closed wet cooling tower.
Appl Therm Eng 2012;35:233–9.
[19] Lucas M, Martínez PJ, Viedma A. Experimental determination of drift loss from
a cooling tower with different drift eliminators using the chemical balance
method. Int J Refrig 2012;35(6):1779–88.
[20] Jiang JJ, Liu XH, Jiang Y. Experimental and numerical analysis of a cross-flow
closed wet cooling tower. Appl Therm Eng 2013;61(2):678–89.
[21] Lavasani AM, Baboli ZN, Zamanizadeh M, Zareh M. Experimental study on the
thermal performance of mechanical cooling tower with rotational splash type
packing. Energy Convers Manage 2014;87:530–8.
[22] Ning T, Chong D, Jia M, Wang J, Yan J. Experimental investigation on the
performance of wet cooling towers with defects in power plants. Appl Therm
Eng 2015;78:228–35.
[23] Askari S, Lotfi R, Seifkordi A, Rashidi AM, Koolivand H. A novel approach for
energy and water conservation in wet cooling towers by using MWNTs and
nanoporous graphene nanofluids. Energy Convers Manage 2016;109:10–8.
[24] Singh Kuljeet, Das Ranjan. An experimental and multi-objective optimization
study of a forced draft cooling tower with different fills. Energy Convers
Manage 2016;111:417–30.
[25] Shahali P, Rahmati M, Alavi SR, Sedaghat A. Experimental study on improving
operating conditions of wet cooling towers using various rib numbers of
packing. Int J Refrig 2016;65:80–91.
[26] Zhao Yuanbin, Sun Fengzhong, Long Guoqing, Huang Xiaofeng, Huang
Wenqiang, Lyv Dongqiang. Comparative study on the cooling characteristics
of high level water collecting natural draft wet cooling tower and the usual
cooling tower. Energy Convers Manage 2016;116:150–64.
[27] Rahmati Mehdi, Alavi Seyed Rashid, Sedaghat Ahmad. Thermal performance of
natural draft wet cooling towers under cross-wind conditions based on
experimental data and regression analysis. In: 2016 6th conference on thermal
power plants (CTPP). IEEE; 2016. p. 1–5.
[28] Alavi SR, Rahmati M, Ziaei-Rad S. A new approach to design safe-supported
HDD against random excitation by using optimization of rubbers spatial
parameters. Microsyst Technol 2016. http://dx.doi.org/10.1007/s00542-016-
2944-x.
[29] Goodarzi M, Keimanesh R. Heat rejection enhancement in natural draft cooling
tower using radiator-type windbreakers. Energy Convers Manage
2013;71:120–5.
[30] Treybal Robert E. Mass-transfer operations, vol. 3. New York: McGraw-Hill;
1968.
[31] Chatterjee S, Hadi AS. Regression analysis by example. In: Wiley Series in
Probability and Statistics. Wiley; 2006.
[32] Alavi SR, Rahmati M. Experimental investigation on thermal performance of
natural draft wet cooling towers employing an innovative wind-creator setup.
Energy Convers Manage 2016 Aug;15(122):504–14.