Applied Thermal Engineering 40 (2012) 114e120

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

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

Performance enhancement of air-cooled chillers with water mist: Experimental

and analytical investigation
Jia Yang a, b, *, K.T. Chan a, Xiangsheng Wu b, Xiaofeng Yang b, Hongyu Zhang b
a
Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China
b
Department of Real Estate Management and Environmental Engineering, Logistical Engineering University, Chongqing, China

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

Article history: This paper investigates how water mist evaporative pre-cooling can be applied to air-cooled chillers to
Received 14 October 2011 improve the chiller efficiency by on-site experimental studies. There is a lack of detailed experimental
Accepted 1 February 2012 studies on the application of water mist system on air-cooled chillers. The experiment was conducted on
Available online 9 February 2012
a chiller plant with water mist system in a subtropical climate. The experimental results showed that the
dry bulb temperature (DBT) of entering condenser air with water mist pre-cooling could drop by up to
Keywords:
9.4 K from the ambient air temperature, and the approach could be as low as 0.5 K. A thermal effec-
Air-cooled chiller
tiveness of up to 0.91 was obtained in using the water mist system. The pre-cooled condenser air enabled
Water mist
Evaporative cooling
a drop of the condensing temperature by up to 7.2 K, and the chiller coefficient of performance (COP)
Energy saving could be improved in varying degrees by up to 18.6%. This study demonstrates that the water mist system
Experimental investigation coupled to air-cooled chillers is an energy-efficient and environment friendly technique, which has
significant potential to improve the efficiency of air-cooled chillers and reduce the electricity demand for
the commercial and industrial sector.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction a convention that the proper function of thermostatic expansion

valves needs a high pressure differential of at least 690 kPa to
Large commercial buildings in Hong Kong typically require control refrigerant flow properly. The use of electronic expansion
cooling year round, and centralized water chillers are commonly valves enables the required differential pressure to be as low as
used for providing space cooling. The operation of chillers accounts 290 kPa for ensuring compressor lubrication [3]. Under such
for about 40% of the annual total energy use in commercial build- control strategies, the condenser fan power can be minimized, but
ings in a subtropical climate [1,2], which contributes considerable the compression efficiency becomes lower and the compressor
greenhouse gas emissions. The chiller efficiency is a matter of power remains high, which causes a considerable decrease in COP
concern, which is subjected to influence by the heat rejection when the chillers operate under part load conditions. It is, there-
method, load ratio, external conditions and compressor efficiency. fore, desirable to increase the chiller COP through improving the
Air-cooled chiller systems are commonly used in commercial operational control of their components. Some researchers have
buildings due to its flexibility, the ease of installation, the simplicity stated the opportunity to lower the condensing temperature to
of operation and maintenance, especially for the cities with water improve the operating efficiency of air-cooled chillers [2,4,5].
shortage problem. Compared with water-cooled chillers, air-cooled As condenser fans force ambient air to condense and slightly sub-
chillers are generally energy inefficient. The deficient performance cool the refrigerant, the extent of the condensing temperature drop
of air-cooled chillers is mainly due to the traditional operation is constrained by the dry bulb temperature (DBT) of ambient air.
under head pressure control (HPC), whereby minimal condenser Evaporative cooling can pre-cool the ambient air before entering the
fans are staged to control the condensing temperature to float condenser, which is effective for improving the performance of air-
around a high set point of 50
C. The use of HPC is due to cooled chillers. This concept is enhanced at present in consideration
of energy saving and environmental protection.
Refrigeration systems with evaporative condensers have been
applied for years. For the evaporative condensers, ambient air is
* Corresponding author. Department of Building Services Engineering, The Hong
drawn or blown through a porous wetted surface with a film of cool
Kong Polytechnic University, Hong Kong, China. Tel.: þ852 2766 5837; fax: þ852
2765 7198.
water. The air stream is cooled by the evaporation of water, and its
E-mail address: [email protected] (J. Yang). DBT drops to approach its wet bulb temperature (WBT). With

1359-4311/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2012.02.001
 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120 115

a reduction of entering condenser air temperature, the condensing 3.1. Description of the chiller and the water mist system
temperature is lowered, which results in a decrease of the
compressor power of air-cooled chillers. However, additional fan In this study, a chiller plant installed in an institutional complex
power is required for the evaporative condenser to draw or blow air comprising of three identical screw chillers connected in parallel
through the porous surface, which trades off the reduction of was investigated. Each chiller has two refrigeration circuits, namely
compressor power partly. Zhang et al. [6] indicated that the circuit A and circuit B, using refrigerant R134a. The chiller is
application of evaporative pre-coolers could improve the COP of equipped with one compressor for circuit A and two compressors
air-cooled chillers by 14.7% under the climatic conditions of Tianjin for circuit B. The nominal cooling capacity of the studied chiller is
in China. According to a simulation study [7], a 1.4e14.4% decrease 705 kW, rated under the operating conditions of entering
in chiller power was achieved when an air-cooled reciprocating condenser air temperature at 35
C and entering/leaving chilled
chiller with an evaporative pre-cooler operated under HPC. water temperatures at 12
C/7
C. The rated power of the studied
However, this technology has some adverse effects including the chiller is 242 kW. The condensers comprise of 10 identical
risk of mineral deposits and corrosion of the condenser coils, which condenser fans arranged with a total airflow rate of 53.45 m3/s,
will reduce the cooling efficiency with time. including four fans serving circuit A and six fans serving circuit B.
An alternative for evaporative condenser is to install a water The fan speed is 15.8 r/s, and each fan consumes a power of 2.4 kW.
mist system to pre-cool the air entering condensers. The water mist The air-cooled screw chiller plant has been installed to provide
pre-cooling system is not a new concept, and it has been applied space cooling for some years. In order to improve the chiller effi-
successfully in the industries [8e10]. However, the application of ciency, water mist systems were installed in July 2009. Each chiller
water mist pre-cooling associated with a chiller system is is served by a separate water mist system comprising of two water
uncommon. The water mist system has advantage over the evap- mist circuits, as shown in Fig. 1. The two water mist circuits contain
orative condenser because, firstly, it has no additional air pressure high pressure pumps at rating of 0.75 kW and 1.25 kW dedicated
loss through the wetted media and hence no additional fan power for refrigeration circuits A and B, respectively. The flow rate of the
will be incurred. Although the water mist pump will operate to high pressure pumps are 2 l/min and 4 l/min at 1500 rpm for circuit
deliver water at a high pressure of around 70 bars, the water flow A and circuit B, respectively, discharging through a total of 58
rates are very low and hence a small amount of power to drive the nozzles. According to the layout of the condenser coil, the nozzles
high pressure pumps is the only additional power to be considered. are evenly distributed in front of the entire condenser surface to
Secondly, it is simple and convenient to install a mist system, which ensure better evaporative effect. The DBT and relative humidity
is advantageous in retrofitting for the existing hundreds of thou- (RH) of entering condenser air were measured with a data logger
sands of air-cooled chillers. Due to the remarkable advantage of system, and the other variables related to chiller performance were
water mist system, it is expected to be widely applied. Yu and Chan recorded by the building management system (BMS). Fig. 2 shows
[11] studied the application of water mist system by simulation. the experimental setup of the air temperature and RH sensors at
There is a lack of field investigations on air-cooled chillers with the inlet to the condensers of the studied air-cooled screw chiller. A
water mist system. This paper aims to conduct an experimental
study on the energy performance of air-cooled chillers with water
mist pre-cooling system via on-site measurement under various Water supply
operating conditions. The enhancement of chiller COP with water

Chilled water
mist system is evaluated, which is necessary for developing and
6
designing more efficient systems.
air
2. Water mist system
2
Fig. 1 illustrates a schematic of a typical water mist system,
which comprises of a high pressure pump, a filter unit, atomization
5
nozzles, high pressure and low pressure tubing. The high pressure
pump can operate to deliver water at a high pressure of around 3
70 bars, and the water is forced through micro nozzles at very high 4
pressure to create a water mist of 10 micron sized droplets. When 1
the tiny water droplets are sprayed into the atmosphere, they 2
quickly absorb the heat in the environment and evaporate, and
then the air temperature decrease due to evaporative cooling effect.
When the water mist system is coupled to an air-cooled chiller, the 6
temperature of the entering condenser air will drop, which results
in lowering the condensing temperature and pressure. With the air
decrease of the condensing pressure, the compressor power is
reduced and the chiller performance is improved.

3. Experiment setup 5

The main objective of this study is to evaluate experimentally 3

4
the performance of air-cooled screw chillers with or without water
mist system. This involved measurements of the temperature, flow
rate, pressure, power and other parameters related to the chiller 1. Evaporator 2. Compressor 3. Condenser 4. Expansion valve
performance. With the aid of a data logging system, data of related 5. High pressure pump 6. Nozzles
parameters can be continuous recorded and compiled for further
analysis. Fig. 1. Schematic of the water mist system.
116 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120

and 3% for relative humidity. One data logger was installed at the
discharge of the condenser to measure the DBT and RH of the air
leaving the condenser, whose uncertainty was 0.3
C for air
temperature and 2.5% for relative humidity.
When the water mist system was operated, the evaporative
cooling effect reduced the temperature of the entering condenser
air. Fig. 3 shows the temperature change of the entering condenser
air against the wet bulb depression (WBD) which is the difference
between DBT and WBT of the ambient air. The scattered points are
grouped with different levels of approach to WBT (at 1 K interval).
Approach is the temperature difference between the dry bulb
temperature of entering condenser air (Tecd) and the wet bulb
temperature of ambient air (Twb), which can be as low as 0.5 K. It
illustrates that the air temperature may decrease by varying
degrees up to 9.4 K. It was observed that the WBD had a great
Fig. 2. Photograph of the water mist system with data loggers. impact on the temperature drop. The greater the WBD of the
ambient air was, the greater the temperature drop could be ach-
ieved when water mist was operated. When the DBT was high and
distance of 750 mm was maintained between the nozzles and the RH was low, the temperature could drop more, as the evaporation
condenser surface to ensure better cooling effect by full evapori- process occurred intensively under such conditions. Therefore,
zation of water mist. The distance between the nozzles and the water mist system is especially suitable for dry and hot area, and
condenser face was determined based on the configuration of the can be sized smaller for a given heat load under such weather
condenser coil and preliminary on-site tests, which depended on conditions.
the ambient air conditions, heat rejection airflow, water mist As the chiller COP was not directly measured but calculated
generation rate, layout of nozzles, etc. The numerical study by from the measured variables, an uncertainty analysis on the chiller
Tissot et al. [12] confirmed that almost all the sprayed water mist COP was carried out. The chiller load (Q) could be calculated by Eq.
fully evaporate after a distance of 0.7 m from the injection point. (1). The chiller COP was calculated by the chiller load divided by the
The detailed experiment on the distance between the nozzles and total power input, including the chiller power (Ech) and high pres-
the condenser face will be a future research work. With the water sure pump power (Ehp) for generating water mist, and it was
mist data and the chiller data, the chiller performance could be described as Eq. (2). Chiller power included the compressor power
analyzed. (Ecc), condenser fan power (Ecf) and lubrication pump power (Elp).
As the lubrication pump power was very small, it was included in
3.2. Experimental method the measured compressor power.

The studied air-cooled screw chiller was operated under head Q ¼ mw $Cp ðTchwr  Tchws Þ (1)
pressure control. Experiments were conducted to investigate the
performance of the air-cooled chiller with or without water mist



system over a representative range of ambient temperature (Tdb:
COP ¼ Q = Ech þ Ehp ¼ mw $Cp ðTchwr  Tchws Þ= Ecc þ Ecf þ Ehp
23.8e33.5
C), relative humidity (RH: 31.3e90.1%) and part load
ratios (PLRs: 0.1e1.2) in the local climate, over a period of 4 months (2)
from July to November 2009.
where, Cp is the specific heat capacity of water.
To investigate the performance of the air-cooled chiller with or
ASHRAE Guideline 2 [13] provides a guideline for reporting
without water mist system, there were many parameters needed to
uncertainty in results of experimental data as applied to HVAC
be measured, including evaporating temperature (To) and
equipment. Due to the accuracy of measured variables, the
condensing temperature (Tk) of each refrigeration circuit, power of
each compressor (Ecc), condenser fan power of each circuit (Ecf),
chilled water flow rate (mw), temperatures of supply and return
chilled water (Tchws and Tchwr), water mist generation rate (mwm),
0
Temperature change of the entering condenser air (K)

condenser air temperature and relative humidity at 3 locations:

before the water mist (ambient air), at the inlet of the condenser -1

coil and leaving the condenser coil. The operating data were -2 Approach
to WBT
monitored and logged at 10-min or 15-min intervals over the -3
0~1
experimental period. Resistance type temperature sensors were
-4 1~2
used to measure the temperature of chilled water with an uncer- 2~3
tainty of 0.1
C. The chilled water flow rate was measured by an -5
3~4
ultrasonic flow meter with an uncertainty of 0.5% of the measured -6 4~5
5~6
value. Compressor power (Ecc), condenser fan power (Ecf) and high -7 6~7
pressure pump power (Ehp) were metered by power analyzers with 7~8
-8
an uncertainty of 0.1 kW. The DBT and RH of ambient air were
monitored by a transmitter with an uncertainty of 0.1
C for air -9

temperature and 2% for relative humidity, which were used to -10
0 2 4 6 8 10 12 14
control the operation of the water mist systems. To measure the
Wet bulb depression (K)
DBT and RH of entering condenser air, data loggers were installed at
the inlet of the condenser as shown in Fig. 2, which were weath- Fig. 3. Impact of ambient air WBD on temperature change of entering condenser air
erproof sensors with an uncertainty of 0.2
C for air temperature due to water mist pre-cooling.
 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120 117

uncertainty associated with COP was determined by the single

a0
sample analysis [13], as shown in Eq. (3).
24 25 26 27 28 29 30 31 32 33
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u n -1
uX
¼ t
2
dCOPðrmsÞ ½dxi $ðvCOP=vxi Þ (3) -2
RH of
i¼1
ambient air
-3
40-50%
where, xi is the ith independent variable, dxi is the uncertainty of
-4 50-60%
the variable xi, and n is the number of measurements related to COP. 60-70%
Using Eqs. (1)e(3), the root sum square error of chiller COP -5 70-80%
(dCOP(rms)) due to all the uncertainties of the individual variables 80-90%
-6
was evaluated to be 0.099 in a COP value of 2.9 at the design
condition, and the uncertainty of COP was 3.4%. This was mainly -7
due to the uncertainties in chilled water supply and return
temperature. -8
Ambient dry bulb temperature ( C)

4. Results and discussion b 4

3
A detailed analysis is reported in the following section on how
2
water mist pre-cooling improve the performace of air-cooled
1 RH of
chillers and what parameters should be considered to optimize ambientair
the operation of the mist system coupled to the chiller. 0
40-50%
24 25 26 27 28 29 30 31 32 33
-1 50-60%
60-70%
4.1. Condensing temperature -2
70-80%
-3 80-90%
For any given cooling capacity, the condensing temperature
-4
depends on the chiller load, heat rejection airflow and ambient air
-5
temperature. When the chiller operated without water mist, the
condensing temperature was above 50
C for more than half of the -6

operating conditions during the experimental period. When water Ambient dry bulb temperature ( C)

mist system was operated, the condensing temperature was less Fig. 5. Change in refrigerant condensing temperature (DT) due to water mist
than 55
C for all operating conditions and was below 45
C for pre-cooling at part load ratios of 0.9e1.0 (a) Circuit A (b) Circuit B.
more than half of the time. The lowest condensing temperature was
37.8
C during the experimental period. Figs. 4 and 5 show the
change of refrigerant condensing temperature with the ambient air
temperature under different RH ranges at part load ratios of
0.4e0.5 and 0.9e1.0 for circuits A and B respectively, which was due or drop by up to 5.6 K during the experimental period. The variation
to the effect of water mist pre-cooling. The positive numbers of the condensing temperature changes was caused by the HPC
represent a condensing temperature increase while the negative with constant speed condenser fans and the specific configuration
numbers represent a condensing temperature drop. According to of the studied chiller having two refrigeration circuits. Circuit A was
the control strategy of the studied chiller, only circuit B was oper- simpler and equipped with one compressor only. It was more
ated at part load ratios of 0.4e0.5. For refrigeration circuit A, the complicated to control circuit B which was equipped with two
condensing temperature drop varied regularly by up to 7.2 K. When compressors and six condenser fans. Under HPC with constant
the RH was lower, the condensing temperature drop could be speed condenser fans, the condensing temperature could fluctuate
larger. However, for refrigeration circuit B, the condensing considerably because the heat rejection airflow varied in steps by
temperature varied irregularly, which could increase by up to 3.3 K staging pairs of condenser fans.
When the condensing temperature exceeded the high
condensing temperature setting, one more group of condenser fans
2 would be switched on to increase heat rejection airflow, which
enabled the condensing temperature to fall below the high setting.
1
Because water mist pre-cooling would reduce the temperature of
0 entering condenser air, the condensing temperature could fall
RH of
24 25 26 27 28 29 30 31 32 33
ambient air
below the high setting even if the number of staged condenser fans
-1 remained unchanged. For certain conditions, when the condensing
-2
temperature exceeded the high setting, both staging on more
condenser fans and operating water mist could enable the
-3 condensing temperature to drop below the high setting, but the
extent to which the condensing temperature could drop depended
-4
on the working conditions. As shown in Fig. 4, when the ambient
-5 temperature varied from 30 to 33
C at part load ratios of 0.4e0.5,
there was no obvious condensing temperature difference by
-6
o
staging on one more group of condenser fans and operating water
Ambient dry bulb temperature ( C)
mist, although the condensing temperature of chiller with mist pre-
Fig. 4. Change in refrigerant condensing temperature (DT) due to water mist cooling could be slightly higher than that of chiller without water
pre-cooling at part load ratios of 0.4e0.5 (Circuit B). mist when ambient DBT was 32
C and RH was above 70%.
118 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120

In the case without water mist, when one more group of range temperatures at various operating conditions in this experi-
condenser fans were cycled on under some conditions, the heat mental investigation. The thermal effectiveness could be as high as
rejection airflow provided by the constant speed fans could exceed 0.91 when the approach temperature was 0.5 K. For evaporative
the airflow required to maintain the set point of condensing cooling, the ability to generate cooling medium approaching the
temperature, and the reduction of the condensing temperature was WBT of ambient air is crucial. A well designed cooling tower can
more significant than that due to the evaporative cooling effect of give an approach to WBT of 3 K. Costelloe et al. [14] reported that
water mist when the number of staged condenser fans remained low primary approach temperature varying from 0.5 to 2.0 K could
unchanged. Hence the condensing temperature of the chiller with be achieved. The approach to WBT for both water mist system and
water mist could be higher than that of the chiller without water cooling towers can be as low as 0.5 K, but the water mist system will
mist under HPC for certain conditions, as indicated by positive give larger approach temperature in many conditions due to the
values of change of refrigerant condensing temperature in Fig. 5 designed water mist generation rate.
when ambient temperature varied from 27 to 30
C and chiller In this experiment, the water mist cooling could provide an
load ratios were close to one. approach to WBT as low as 0.5 K, and 52.3% of the approach
It should be noted that the condensing temperature dropped temperatures were less than 2 K as shown in Fig. 3. The thermal
significantly for the chiller with water mist from the case without effectiveness of the water mist was higher than 0.6 for more than
water mist, when the number of staged condenser fans were the half of the working conditions during the experimental period. It
same in these two cases. These figures reveal that the condensing was found that the thermal effectiveness could be lower than 0.4
temperature of an air-cooled chiller with constant speed condenser due to the fact that the designed mist generation rate was much
fans may decrease or increase due to water mist pre-cooling under smaller than the required peak mist generation rate. If the water
HPC, and the RH has a greater impact on the reduction of mist system was designed with larger water mist generation rate,
condensing temperature than the DBT of the ambient air. there would be more chances to achieve higher thermal
Compressor power can be saved by decreasing the condensing effectiveness.
temperature, and the chiller performance is improved. The thermal effectiveness of a direct evaporative pre-cooler
filled with corrugated holed aluminum foil under the climatic
conditions of Tianjin varied from 0.65 to 0.85 when the water
4.2. Thermal effectiveness
sprinkling density increased from 0.2 to 1.2 kg m2s1 [6]. The
variation of the thermal effectiveness of water mist and an evap-
A suitable means of assessing the thermal performance of the
orative pre-cooler indicates that the effectiveness of evaporative
water mist is the thermal effectiveness (ε), which is defined as
cooling by both methods is greatly affected by the ambient air
a fraction of the maximum possible cooling in the ambient condi-
conditions, heat rejection airflow, design of the water mist system,
tions, and is given by the following equation:
water sprinkling intensity and water mist generation rate.
Tdb  Tecd It should be noted that thermal effectiveness of the water mist
ε ¼ (4) system coupling to the air-cooled chiller could be different with the
Tdb  Twb
same approach temperature, which was greatly affected by the
where, Tdb is the DBT of the ambient air, and Tecd is the DBT of range temperature as shown in Fig. 6. With certain approach
entering condenser air, and Twb is the WBT of the ambient air. temperature, the thermal effectiveness would be higher with larger
The thermal effectiveness involves both the approach and the range temperature. To achieve certain thermal effectiveness with
range condition, as they are the two key determinants of energy a low approach temperature, a relative lower range temperature
performance. Range is the temperature difference between the was required.
ambient air and the entering condenser air. The thermal effec-
tiveness can be expressed as Eq. (5). 4.3. Chiller COP

Tdb  Tecd Range The COP of the chiller operating with water mist pre-cooling
ε ¼ ¼ (5)
ðTdb  Tecd Þ þ ðTecd  Twb Þ Range þ Approach was compared with the COP of the chiller without water mist
pre-cooling under similar working conditions, and the chiller COP
The ‘Approach’ is a better indicator of water mist performance,
improvement resulting from the lower entering condenser air
which is an important parameter in determining the capacity and
temperature and hence lower condensing temperature could be
optimizing the operation of the water mist system. Fig. 6 shows the
identified. Fig. 7 illustrates the percentage change of the chiller COP
impact of thermal effectiveness on the approach to WBT for typical
under different RH ranges of ambient air at part load ratios of
0.4e0.5 and 0.9e1.0 from the baseline (without water mist pre-
8 cooling). A positive number means a chiller COP increase while
7
a negative number means a chiller COP reduction. It was observed
Range that the chiller COP with water mist was improved noticeably from
6 T emp.
the baseline for most operating conditions, and could be improved
Approach temperature (K)

2~3
5 3~4
in varying degrees by up to 18.6% when the reduction in
4~5 compressor power exceeded the additional power of high pressure
4
5~6 pump generating water mist. The chiller COP could be improved
6~7
3
7~8
more when RH was lower, as shown in Fig. 7, because evaporation
2 was dependent on the RH of ambient air. With lower RH, evapo-
ration process occurred more intensively, and a greater tempera-
1
ture drop could be achieved. Therefore, the improvement of chiller
0 COP could be more significant for chillers operating in hot and arid
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
regions.
Thermal effectiveness
However, the chiller COP could decrease by 3.8% for certain
Fig. 6. Variation of approach temperature with thermal effectiveness. condition, which indicated that water mist pre-cooling had both
 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120 119

a 20 condensing temperature control and variable speed condenser fans

are applied to air-cooled screw chillers, the COP can increase by
12.1e110.4% [15]. When the water mist system is coupled to the air-
15
cooled chillers with variable speed condenser fans and CTC, the
RH of
ambient air
chiller COP will be improved further.
Zhang et al. [6] investigated the evaporative pre-cooler filled
COP change (%)

40-50%
10
50-60%
with corrugated holed aluminum foil and reported that COP of the
60-70% chiller could be improved by 14.7% under the climatic conditions of
70-80% Tianjin. Hajidavalloo [16] studied the effect of incorporating evap-
5
80-90% orative cooler in the window-air-conditioner by injecting water on
the media pad installed in both sides of the air conditioner in the
0 regions with very hot weather conditions (about 50
C), and
24 25 26 27 28 29 30 31 32 33
reported that the power consumption decreased by about 16% and
the COP increased by about 55%. According to a reported simulation
-5
o study [11], the use of water mist pre-cooling enabled an increase of
Ambient dry bulb temperature ( C)
chiller COP in various degrees up to 7.7%, but it could drop from the
b 20 baseline by up to 1.3% for an air-cooled screw chiller operating
under HPC in subtropical regions. Compared with these studies on
evaporative cooling, both the evaporative cooler device and the
15
RH of water mist system could improve the COP of air-cooled chillers, but
ambient air the improvement of the chiller efficiency depended on the ambient
COP change (%)

10
40-50% air conditions, size of the evaporative pre-cooler, design of the
50-60%
water mist system, layout of the mist nozzles, chiller load ratios and
60-70%
70-80%
control strategies of the chillers.
5 The strategy to operate the water mist system is vital. It was
80-90%
suggested to operate the water mist system only when DBT is
0 greater than 28
C and RH is less than 75%. Under this criteria, the
24 25 26 27 28 29 30 31 32 33 chiller efficiency could be improved, but many operating hours will
be screened out. As shown in Fig. 7, the water mist system worked
-5 effectively even when the chiller operated beyond the above
o
Ambient dry bulb temperature ( C) criteria so as to maximize its energy saving potential. At RH of
80e90% and part load ratios of 0.9e1.0, the chiller COP could be
Fig. 7. Chiller COP percentage change at various conditions due to water mist
pre-cooling (a) at part load ratios of 0.4e0.5 (b) at part load ratios of 0.9e1.0. improved by 6.6% and 9.8% when the chiller operated at ambient
temperatures of 24
C and 32
C, respectively. The criteria for
operating air-cooled chillers with water mist system for optimal
efficiency have to be investigated further with more experiments
positive and negative effects on chillers operating under HPC. The conducted at various combinations of PLR, weather parameters,
casual drop of the chiller COP was due to HPC and the extra pump water mist generation rates and droplet sizes.
power generating water mist. For the air-cooled chillers with
constant speed condenser fans, the condenser fans would be cycled 4.4. Water consumption
on and off with reference to a high and a low condensing
temperature settings under HPC. For certain operating conditions, The designed flow rate of each nozzle was 1.25  103 l/s when
due to the pre-cooling effect, more groups of condenser fans would the high pressure pumps operated to deliver water through the
be staged without water mist to let the condensing temperature tubing at a high pressure of 70 bars. The water mist system was
drop below the high setting than that operating with water mist. designed to provide a total water mist generation rate of 0.073 l/s
With more condenser fans, the reduction of the condensing for the studied air-cooled screw chiller. For water-cooled chillers,
temperature could be more significant than that due to the evap- the total water losses from cooling tower are the sum of drift losses,
orative cooling effect of water mist when the number of staged evaporation losses and blow down losses. According to Standard
condenser fans remained unchanged. Under such conditions, the 550/590 [17], the total water losses could be calculated based on an
compressor power without mist pre-cooling was less than that assumption that the water losses accounts for 1.5% of the cooling
with water mist pre-cooling, and hence the chiller COP without water flow rate, which is designed at 0.054 l/s per kW cooling
water mist could be higher than that with water mist pre-cooling, capacity. For the water-cooled chiller with same cooling capacity as
as shown in Fig. 7 when ambient temperature varied from 27 to the studied air-cooled chiller, whose capacity is 708 kW, the
30
C at part load ratios of 0.9e1.0. To make maximum advantage of designed cooling water flow rate is 38.2 l/s and the expected water
the water mist pre-cooling, it is undesirable to operate the air- loss is 0.57 l/s. Comparatively, the water consumption rate of the
cooled chiller under HPC, and HPC should be replaced by variable water mist system is much smaller, and is about 9.5% of the water
condensing temperature control (CTC), whereby the condensing losses for an open-loop cooling tower serving a water-cooled chiller
temperature can approach its lower boundary via staging all with same cooling capacity.
condenser fans in most operating conditions. Under CTC, the sum of The advantages of applying water mist pre-cooling have been
compressor power and condenser fan power can be minimized by investigated for air-cooled chillers operating in a subtropical
staging condenser fans with optimal number and speed for all climate. As WBD is associated with the maximum possible lowering
operating conditions. Variable speed control for condenser fans is of air temperature, it is expected that energy savings from this
another complement to water mist pre-cooling, which is superior technology are more significant when WBD of ambient air is higher
to cycling constant speed condenser fans in steps, with regard to during the cooling season. As the water mist system consumes less
the controllability of condensing temperature. When the variable than 10% of the water losses required by an open-loop cooling
120 J. Yang et al. / Applied Thermal Engineering 40 (2012) 114e120

tower system, this technology of water mist pre-cooling is benefi- simple setup and its independence with the arrangement of
cial for wide application, especially for cities challenged by shortage multiple chillers. This technique has a significant potential to
of water. It occupies an intermediate position between air-cooled improve the efficiency of air-cooled chillers and reduce the elec-
chillers and water-cooled chillers, and is particularly applicable in tricity demand for the commercial and industrial sector.
a hot and arid environment where water is scarce and the WBD is
significant.
In the condenser air stream, the water mist generated may not Acknowledgements
fully evaporate before entering the condenser coil, and the unva-
porized or excessive mist may cause damping of the condenser fins The work described in this paper was supported by a grant from
and coil. On the other hand, any presence of water mist in the air the central research grant of The Hong Kong Polytechnic University,
stream passing through the condenser coil may provide further Project Code: RP1R.
evaporation and reject more heat from the refrigerant. This is an
unknown process which deserves further research. References

5. Conclusions [1] J.C. Lam, Energy analysis of commercial buildings in subtropical climates,
Build. Environ. 35 (2000) 19e26.
[2] K.T. Chan, F.W. Yu, Applying condensing-temperature control in air-cooled
The effect of operating water mist system to enhance the reciprocating water chillers for energy efficiency, Appl. Energ. 72 (2002)
performance of air-cooled chillers under various operating condi- 565e581.
tions in a subtropical climate was investigated by field measure- [3] F.W. Yu, K.T. Chan, H.Y. Chu, Constraints of using thermostatic expansion
valves to operate air-cooled chillers at lower condensing temperatures, Appl.
ments. As the evaporation of water mist pre-cooled the ambient air Therm. Eng. 26 (2006) 2470e2478.
entering the condenser from its DBT to approach the WBT, the [4] K.A. Manske, D.T. Reindl, S.A. Klein, Evaporative condenser control in indus-
condensing temperature dropped, and the compressor power was trial refrigeration systems, Int. J. Refrig. 24 (2001) 676e691.
[5] F.W. Yu, K.T. Chan, Advanced control of condensing temperature for
reduced at the expense of insignificant power of the high pressure enhancing the operating efficiency of air-cooled chillers, Build. Environ. 40
pump for generating water mist. The entering condenser air (2005) 727e737.
temperature could drop by up to 9.4 K, and the condensing [6] H. Zhang, S.J. You, H.X. Yang, J.L. Niu, Enhanced performance of air-cooled
chillers using evaporative cooling, Build. Serv. Eng. Res. T. 21 (2000) 213e217.
temperature could drop by up to 7.2 K. Water mist system could
[7] F.W. Yu, K.T. Chan, Application of direct evaporative coolers for improving the
give an approach as low as 0.5 K, and the thermal effectiveness energy efficiency of air-cooled chillers, J. Sol. Energ. Eng. 127 (2005) 430e433.
could be up to 0.91, which would enable the air-cooled chiller [8] N. Cheung, C.A. Santos, J.A. Spim, et al., Application of a heuristic search
plants to operate at higher efficiency. When an air-cooled chiller technique for the improvement of spray zones cooling conditions in contin-
uously cast steel billets, Appl. Math. Model. 30 (2006) 104e115.
was controlled using the conventional HPC, the chiller COP with [9] C.C. Hsieh, S.C. Yao, Evaporative heat transfer characteristics of a water spray
water mist could be higher or lower than that without water mist on micro-structured silicon surfaces, Int. J. Heat Mass Trans. 49 (2006)
pre-cooling, depending on the working conditions. In general, 962e974.
[10] Y.F. Zhang, Y.D. Fang, X. Huang, et al., Experimental study of the interation
applying water mist pre-cooling could improve the chiller COP in between the water mists and PVC fire, Process Saf. Environ. 85 (2007) 39e44.
various degrees up to 18.6%. Yet a 3.8% reduction of the COP could [11] F.W. Yu, K.T. Chan, Modelling of improved energy performance of air-cooled
be resulted due to undesirable staging of constant speed fans under chillers with mist pre-cooling, Int. J. Therm. Sci. 48 (2009) 825e836.
[12] J. Tissot, P. Boulet, F. Trinquet, et al., Air cooling by evaporating droplets in the
HPC. Variable speed fan control and variable condensing temper- upward flow of a condenser, Int. J. Therm. Sci. 50 (2011) 2122e2131.
ature control (CTC) are favorable complements to water mist pre- [13] ASHRAE Guideline 2-2010, Engineering Analysis of Experimental Data, ASH-
cooling. When the water mist system is coupled to the air-cooled RAE, Atlanta, GA, 2010.
[14] B. Costelloe, D. Finn, Thermal effectiveness characteristics of low approach
chillers with variable speed condenser fans and CTC, the chiller
indirect evaporative cooling systems in buildings, Energ. Buildings 39 (2007)
COP can be improved further. 1235e1243.
It is noted that the water consumption rate of the water mist [15] F.W. Yu, K.T. Chan, Tune up of the set point of condensing temperature for
more energy efficient air cooled chillers, Energ. Convers. Manage. 47 (2006)
system is small, and is only about 9.5% of the water losses for
2499e2514.
a cooling tower serving a water-cooled chiller with same cooling [16] E. Hajidavalloo, H. Eghtedari, Performance improvement of air-cooled
capacity as the studied air-cooled chiller. This study demonstrates refrigeration system by using evaporatively cooled air condenser, Int. J.
that the water mist pre-cooling coupled to air-cooled chillers is an Refrig. 33 (2010) 982e988.
[17] ARI Standard 550/590, Performance Rating of Water Chilling Packages Using
energy-efficient and environment friendly technique, and can be the Vapor Compression Cycle, Air-Conditioning, Heating and Refrigeration
widely applied to new or existing chiller plants because of its Institute, Arlington, VA, 2003.