Energy and Buildings 42 (2010) 203209

Contents lists available at ScienceDirect

Energy and Buildings

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

Economic benets of optimal control for water-cooled chiller systems serving

hotels in a subtropical climate
F.W. Yu a,*, K.T. Chan b
a
b

Hong Kong Community College, The Hong Kong Polytechnic University, Hong Kong, China
Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 27 July 2009
Received in revised form 25 August 2009
Accepted 25 August 2009

A water-cooled chiller system in an air-conditioned hotel can take up about one-quarter of the total
electricity consumption and considerable amounts of water in the heat rejection process. This paper
evaluates operating cost savings of a chiller system integrated with optimal control of cooling towers and
condenser water pumps. A sophisticated chiller system model was used to ascertain how different
control methods inuence the annual electricity and water consumption of chillers operating for the
cooling load prole of a reference hotel. It is estimated that applying load-based speed control to the
cooling tower fans and condenser water pumps could reduce the annual system electricity use by 8.6%
and operating cost by 9.9% relative to the equivalent system using constant speed fans and pumps with a
xed set point of 29.4 8C for cooling water temperature control. The ways to implement this advanced
control for system optimization are discussed.
2009 Elsevier B.V. All rights reserved.

Keywords:
Water-cooled chiller
Cooling tower
Economic analysis
Optimum control

1. Introduction
Hotels play an important role in sustaining the economic
growth of Hong Kong. Yet they constitute one of the major energy
end users in the commercial sector because of their round-theclock operation. A previous survey [1] indicated that the overall
electricity consumption in the hotel segment increased from 443
to 921 GWh (a 108.2% increase) from 1988 to 2000. With the
increasing supply of local hotels, it was estimated that their
electricity demand could grow by 36.8 GWh per annum over the
period of 20012005 [2]. One possible way to moderate the
increasing electricity demand is to improve the energy performance of air-conditioning systems which account for around half
of the total electricity consumption. According to studies on the
electricity consumption of 16 local hotels [3,4], their energy use
intensity (EUIthe annual electricity consumption in kWh per unit
oor area of a building in m2) varied widely from 198 to 926 kWh/
m2 with an average of 406 kWh/m2. For hotels with central airconditioning plants, water-cooled chiller systems with cooling
towers are commonly used to provide cooling energy in the form of
chilled water to maintain the thermal conditions required for
indoor areas. The operation of chillers and cooling towers leads to
the peak electricity demand and accounts for about half of
electricity consumption for air conditioning. Moreover, cooling

* Corresponding author. Tel.: +852 37460416; fax: +852 23647375.

E-mail address: ccyufw@hkcc-polyu.edu.hk (F.W. Yu).
0378-7788/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.enbuild.2009.08.016

towers rely on evaporation of water in the heat rejection process,

leading to considerable water consumption when the chillers are
operating.
Variable speed technology has long been considered a standard
energy-efcient feature to enhance the energy performance of
chiller systems. Hartman [5] launched the equal marginal
performance principle (EMPP) to assist in optimizing the
performance of chiller systems in which variable speed drives
(VSDs) are applied to all the chillers, condenser water pumps and
cooling tower fans. Yet the all-variable speed arrangement is
seldom considered in system design and VSDs are applied solely to
the secondary loop pumps and cooling tower fans in most systems.
Variable ow of chilled water is increasingly used to reduce energy
use, given that pump energy can be saved when the primary
pumps deliver less ow to their dedicated chillers at part load
conditions. The successful application of variable-primary ow
depends on how the ow and chiller capacity can be adjusted to
match changing load conditions [6]. This application is subject to
the design of the chilled water distribution circuit and the airside
cooling coils are required to be furnished with two-way control
valves in order to allow the ow of chilled water to drop under the
reduced load conditions. For a constant air volume system
requiring high latent cooling capacity, the potential of reducing
the chilled water ow rate under part load conditions is rather
limited and so are the pump energy savings.
Gordon et al. [7] and Hydeman et al. [8] developed chiller
models to study variations of chiller COP (coefcient of performance) at different condenser water ow rates. No analysis was

204

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

made on the trade-off between the chiller power, tower fan power
and condenser water pump power under the reduced ow
conditions. Also no control regime was given on how the condenser
water ow should vary in response to the chiller load and wet-bulb
conditions. Variable speed control is increasingly applied to
cooling tower fans to reduce their cycling frequency and allow
better heat rejection control with decreased fan power [9]. The
dynamic characteristics of cooling towers with the varying heat
rejection airow impose more complications in system optimization. Lu et al. [10] presented a model-based optimization strategy
for the condenser water loop of a chiller system. The strategy
involved minimizing the total power of the chillers, condenser
water pumps and cooling tower fans, taking into account the
interaction between the varying air ow rate and condenser water
ow rate for a given heat rejection rate. The inuence of the
condenser water entering temperature on both the chiller power
and cooling tower performance was analysed. Yet they did not
explain the control logic of the condenser water pump and cooling
tower fan for system optimization.
Graves [11] reformulated the GordonNg model [7] in order to
optimize a system designed with two chiller-pump pairs and one
cooling tower. The chiller model was coupled with an Ntueffectiveness (Ntuno. of transfer units) model [12] for evaluating
the cooling tower performance. Drawing on the chillertower
interaction, the modied model was used to analyse how the
system COP varied with the changing condenser water ow. Yet
the tower model discounted the water loss due to evaporation, and
a single NTU value was assumed to represent the tower
performance at different water and airow rates. Two correlations
were identied to facilitate near-optimal system operation: one
was the linear relationship between the cooling water set point
and wet bulb (an analogy to the xed approach method); another
was the linear relationship between the tower fan speed and pump
speed. Benton et al. [13] developed a regression model to represent
the improved cooling tower simulation algorithm (CTSA). The
algorithm simply indicated the cooling tower approach (cooling
water leaving temperature subtracted from wet-bulb temperature) as a dependent variable of the wet-bulb temperature, range
(temperature difference of cooling water), condenser water ow
and fan power. It remains to be ascertained how to evaluate an
optimal set point for cooling water temperature or the optimal fan

speed control in response to dynamic characteristics of cooling

towers at part load operation.
The aforementioned studies have demonstrated important
modelling techniques to analyse chiller system performance and
some insights on optimizing the control of chiller and cooling
tower systems with variable speed control for condenser water
pumps and tower fans. Yet none of the reported models are
comprehensive enough to assess power relationships of chillers,
condenser water pumps and cooling towers together with water
consumption in the heat rejection process, with regard to various
control methods of cooling towers and condenser water pumps.
Furthermore, most of the studies focus only on electricity sayings
without considering the likely trade-off between water and
electricity savings.
More quantitative gures and economic analysis are considered
necessary in order to reap the potential benets of applying
optimization technologies to chiller systems. The objective of this
study is to investigate the economic benets of a water-cooled
chiller system with the following energy-efcient technologies:
variable speed control of cooling tower fans; a constant approach;
a xed and low cooling water leaving temperature; variable
condenser water ow. These technologies are adaptable to most
existing chiller systems with minor modications. This paper rst
describes a hotel and its chiller system. The method to simulate
hourly building cooling loads is presented. A sophisticated chiller
system model was used to analyse how the system COP varies with
different technologies and to predict the annual electricity and
water consumption of the chiller system. An assessment was made
on the water and electricity cost savings resulting from the
individual and mixed uses of the technologies. The signicance of
this study rests on providing more quantitative analysis to
promote water-cooled chiller systems with optimal operating
strategies in order to boost their environmental performance in
terms of annual electricity and water consumption, and, at the
same time, to reduce their operating costs.
2. Evaluation of hourly cooling loads of the hotel
Table 1 summarizes the features of the hotel to be modelled.
The features were compiled into a building description le for the
multi-zone model in the simulation program TRNSYS 15 [14] used

Table 1
General information about the hotel and its air-conditioning systems.
General
Gross oor area (GFA) (m2)
Total air-conditioned area (m2)
U-values of wall/window/roof (W/m2 8C)
Shading coefcient of glass

52,020
45,540 (87.5% GFA)
1.9/5.4/0.7
0.55

Area
Area per oor (m2)
Air-conditioned area per oor (m2)
Number of oors
Cooling temperature set point (8C)
Relative humidity (%)
Ventilation rate (L/s/person)
Occupancy (m2/person)
Equipment power density (W/m2)
Lighting power density (W/m2)

Guestrooms
2,010
2,010
18
24
50
7.5
18
12
18

Shops and restaurants

2640
1560
6
22
50
5
5
50
35

Air-conditioning system operating hours

MondayFriday
Saturday
Sunday

01002400
01002400
01002400

08002300
08002300
08002300

Air side system details

Type of air handling units

Fan coil units (FCUs)

CAV AHU and FCUs

Chilled water ow control

Chiller plant operating hours (all days)

2-way valve
01002400

2-way valve

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

Fig. 1. Frequency distribution of the hourly data of building load ratios.

for building system energy analysis. The hotel had average

characteristics in terms of the number of guestrooms, total oor
area and EUI according to the energy end-use surveys for the local
hotel sector [2,4]. Based on a typical example of hourly weather
data under the subtropical climate of Hong Kong, the annual
cooling energy of the hotel was evaluated to be 17,692,414 kWh,
with a total of 8152 cooling hours which account for 93.1% of a total
of 8760 h/year. Fig. 1 shows the frequency distribution of the
hourly building load ratiosa ratio to the peak cooling load of
4753 kW. The building load ratios ranging from 0.5 to 1 account for
47% of the total operating hours. This suggests that the chiller
system needs to operate under part load conditions for most of the
time. Higher building load ratios are generally associated with
higher outdoor temperatures, though a certain range of building
load ratios (especially below 0.5) could be present at a large range
of outdoor temperatures.
3. Method of studying the chiller system
3.1. Arrangement of the chiller system
To meet the peak cooling load of 4753 kW, the total capacity of
the chiller system was designed to be 4800 kW. As Table 2
illustrates, the system contains three identical chillers rated at a
Table 2
Details of the chiller system.
Total cooling capacity (kW)
(Three identical sets of chillers, pumps and cooling towers)

4800

For each chiller

Compressor type
Refrigerant type
Nominal cooling capacity (kW)
Nominal compressor power (kW)
COP at full load
Design chilled water supply/return temperature (8C)
Design chilled water ow rate (L/s)
Design condenser water entering/leaving temperature (8C)
Design condenser water ow rate (L/s)

Centrifugal
R134a
1600
280.7
5.7
7/12.5
72
33/38
87

For each cooling tower

Heat rejection capacity (kW)
Design entering/leaving temperature (8C)
Water ow rate (L/s)
Air volume ow rate (m3/s)
Fan motor power (kW)
Drift loss (% of nominal ow)
Design water evaporation rate (% of nominal ow)
Design wet-bulb outdoor temperature (8C)

2004
34.4/29.4
87
63
22.8
0.2
0.53
28

Rated power of each chilled water pump (kW)

Rated power of each condenser water pump (kW)

28
21.6

205

cooling capacity of 1600 kW each. This multiple-chiller arrangement is commonly used to enhance exibility and back up capacity
at part load operation. The capacity modulation of each centrifugal
chiller is done by varying the opening of the inlet guide vanes. Each
chiller is coupled with a constant speed chilled water pump to
deliver the rated ow rate to the chiller evaporator. There is a
differential pressure by-pass pipe in the single-loop pumping
system to balance the ow of chilled water supplied by the staged
chillers against the demand of chilled water required by the airside
system coils.
Chiller sequencing was implemented under which all the
running chillers carried the same part load, and no additional
chillers in the chiller arrangement started to operate until each of
the operating chillers was running at full load. Following this
strategy for staging chillers, two and three chillers operated when
the building cooling load exceeded 1600 and 3200 kW, respectively. Each chiller operating was interlocked with one chilled
water pump, one condenser water pump and one cooling tower.
3.2. Calculation of the annual electricity consumption of the chiller
system
A sophisticated chiller system model was developed using
TRNSYS 15 in order to compute the annual electricity consumption
of chillers, pumps and cooling tower fans. The model considered
the real process phenomena, including the capacity control of
compressors and mechanistic relations between chiller components. In simulating the operation of a chiller of any given size, the
compressors and the condenser satised the mass balance of
refrigerant and the energy balance at the evaporator. The cooling
towers were modelled using the Ntu-effectiveness approach with
which a change in the heat transfer effectiveness of the tower at
various combinations of airow and water ow is taken into
account. An algorithm was included in the cooling tower module to
compute the operating variables of the cooling towers based on a
given set point of cooling water leaving temperature. Details about
the development of the model and its validation were given
elsewhere [15].
The hourly chiller power was computed by using the chiller
system model based on the hourly weather data and the load
which each operating chiller carried. The electricity consumption
of the chilled water pumps varied step by step, depending on the
number of chillers operating. The fan power of each cooling tower
operating depends on whether it is on/off or variable speed control
is used and on the set point of the temperature of cooling water
leaving the cooling tower. The overall annual electricity consumption of the chiller plant is the sum of all chiller power, pump
power and cooling tower fan power for all operating hours.
3.3. Prediction of water consumption costs and electricity costs
The make-up water of a cooling tower serves to compensate for
three parts of water loss: evaporation, drift and bleed-off. The
cooling tower module within the system model considered the
overall mass and energy balance of cooling water and air to
estimate the evaporation loss. The drift rate and bleed-off rate
were assumed to be 0.2% and 0.6%, respectively, of the cooling
water circulation rate, with regard to the use of the traditional
chlorination water treatment technology. The make-up water cost
of HK$4.58/m3 and the sewage charge of HK$1.2/m3 from bleed-off
were used to calculate the overall water consumption and
discharge cost for the operation of cooling towers, based on the
guidelines stated in the pilot scheme for the wider use of fresh
water in evaporative cooling towers for energy-efcient airconditioning systems [16]. The calculation of the electricity cost of
the chiller plant was based on the maximum demand tariff of a

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

206

local power company and the overall electricity consumption

calculated from the chiller system model. The tariff structure
involves different rates for the monthly demand charges and
energy charges. The demand charge is HK$42.1/kVA in the month
for the rst 400 kVA and HK$41.1/kVA for the next additional kVA.
The energy charge is HK$1.023/unit (kWh) for the rst 200 units
supplied per month per kVA of maximum demand and HK$0.962
per unit for each additional unit supplied.
4. Detailed analysis of chiller system performance with various
energy-efcient technologies
Simulation was carried out on eight operating schemes given in
Table 3. Schemes 14 represent a constant speed conguration for
both the cooling tower fan and condenser water pump. Under this
conguration, the tower fan is cycled on and off to deliver the heat
rejection airow required to meet a given temperature of cooling
water leaving the cooling tower (Tctwl) for any heat rejection. The
condenser water pump is staged continuously to provide the
chiller operating with the rated ow of condenser water for all
loading conditions. For Schemes 58, variable speed control is
applied to both the tower fan and condenser water pump in order
to improve their energy efciency at part load operation. Under
this control, the fan speed is modulated in a range of 10100% full
speed in response to a given Tctwl to continuously adjust the airow
rate for any heat rejection with fan power savings. The variable
speed pump provides a varying condenser water ow rate which
changes in direct proportion to the chiller load in order to maintain
a constant temperature difference across the condenser. The
minimum pump speed was set at 50% of full speed to ensure that at
least 50% of the design ow rate passed through the condenser of
each operating chiller. Pump energy can be saved when the chillers
are operating under part load conditions.
With regard to the conventional operation of cooling towers,
the fans are cycled on/off or controlled at variable speed to
maintain the temperature of cooling water leaving the tower (Tctwl)
at its set point. The set points of 29.4 and 18.3 8C used for Tctwl in
the Schemes 1, 2, 5 and 6 are based on the upper and lower limits of
part load rating conditions given in ARI standard 550/590 [17]. The
higher set point is typically used to maintain high condensing
temperature with high compressor power in all operating
conditions while the lower set point helps to keep the lowest
compressor power at the expense of more fan power. A xed
approach of 4 8C was used in Schemes 3 and 7, considering that it

helps offer fan power savings when the chiller load could drop with
the ambient wet-bulb temperature (Tawb) and that this control
provides more energy savings compared with the xed set point
control [9]. For Schemes 4 and 8, the algorithm of simulating tower
fan control searched for an optimal Tctwl to minimize the sum of
compressor power, condenser water pump power and cooling
tower fan power for any given operating condition. The optimal fan
speed for maximum system performance was determined along
with the optimal Tctwl. Scheme 1 is regarded as the baseline
because it is conventionally applied to existing chiller systems.
Schemes 28 represent individual and mixed uses of the energyefcient technologies.
Comparisons were made on how the system COP varied under
Schemes 28 in relation to the baseline. The system COP means the
chiller load output divided by the total input power of the chiller,
condenser water pump and cooling tower fan. All the schemes had
the same power of chilled water pumps and so it was excluded
from the calculation of system COP. As Fig. 2 illustrates, lowering
Tctwl to 18.3 8C could not bring about an absolute increase in the
system COP, though the compressor power was lowest for most
operating conditions among the other operating schemes. Indeed,
the system COP could drop by up to 11% when the chiller load
reduced to a part load ratio of 0.2. This is because the tower fans
operated at full speed continuously and the increase of the fan
power exceeded the reduction of the compressor power. Except
the 18.3 8C Tctwl control, the system COP could be improved by
applying variable speed control to the condenser water pump. This
implies that modulating the condenser water ow in direct
proportion to the chiller load is a viable means to optimize the
trade-off between the compressor power and condenser water
pump power.
The system COP could increase by various degrees for all
operating conditions when using the 4 8C approach or implementing the optimal Tctwl, regardless of whether the on/off or variable
speed control was used for the tower fans and condenser water
pumps. When the wet-bulb temperature was below 28 8C, the
percentage increase in the system COP was more noticeable at
higher chiller loads under the 4 8C approach control. The system
COP increased by 1.416.1% by using the optimum Tctwl and
variable speed control for both the fans and pumps. Such an
increase tends to be signicant at the minimum and maximum
loading conditions and be moderate at a part load ratio of around
0.5 with a wet-bulb temperature of below 28 8C.
5. Economic benets of the energy-efcient technologies

Table 3
Operating schemes of different controls of condenser water pump and cooling
tower.
Scheme

Notation

Description

1 (baseline)

Tctwl29.4CS

Tctwl18.3CS

App4CS

Tctwl,opCS

Tctwl29.4VS

Tctwl18.3VS

App4VS

Tctwl,opVS

Tctwl set at 29.4 8C with

constant speed congurations
Tctwl set at 18.3 8C with
constant speed congurations
Approach (Tctwl  Tawb) xed
at 4 8C with constant speed
congurations
Optimum Tctwl for maximum
system COP with constant
speed congurations
Tctwl set at 29.4 8C with
variable speed congurations
Tctwl set at 18.3 8C with variable
speed congurations
Approach (Tctwl  Tawb) xed
at 4 8C with variable speed congurations
Optimum Tctwl for maximum system
COP with variable speed congurations

Drawing on the chiller system model, calculations were made

on the annual electricity and water consumption of the chiller
system with the eight operating schemes. The electricity and water
consumption data was normalized by the total air-conditioned
oor area of the hotel in terms of kWh/m2 and L/m2, respectively.
Table 4 shows the calculation results. The annual electricity
consumption of chilled water pumps is 9.4 kWh/m2 for all the
operating schemes because the same sequencing of the chillers
was applied for all the schemes. The 18.3 8C cooling water
temperature control (as in Schemes 2 and 6) calls for up to a
21.5% increase in the annual water consumption and the highest
fan electricity consumption. However, it helps minimize the
annual chiller electricity consumption, resulting in a net electricity
saving of 2.02.6% in relation to the baseline. The use of variable
speed control for the condenser water pumps enables their annual
electricity consumption to drop by 54.7% and annual water
consumption by up to 14.8%. This is because the condenser water
ow rate is allowed to drop with the chiller load at part load
operation. Using optimal cooling water temperature with variable
speed control for the tower fans and condenser water pumps, the

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

207

Fig. 2. Percentage change of system COP in Schemes 28 in relation to the baseline. (a) Wet-bulb temperature = 16 8C, (b) wet-bulb temperature = 20 8C, (c) wet bulb
temperature = 24 8C, (d) wet bulb temperature = 28 8C.

A simple change of the cooling water temperature set point

from 29.4 to 18.3 8C could reduce the annual operating cost by
HK$39,188 (0.11% from baseline) for the chiller system studied.
This change results in the lowest cost savings among the other
energy-efcient technologies but causes no additional investment
cost and therefore can be identied to be the most economic means
for system performance improvement. A much higher saving of
HK$458,021 in the annual operating cost can be achieved by
implementing Scheme 8 with which variable speed control was
used for the condenser water pumps and cooling tower fans
instead of the conventional on/off fan control with a set point of
29.4 8C for the cooling water leaving temperature. This saving
helps decide how much investment cost the optimal control is
worth investing. If the hotel owner accepts a maximum payback of
up to 1 year for a chiller system improvement programme, the
ceiling on the investment cost for the optimal control could be
HK$458,021 which is expected to be sufcient for purchasing six
variable speed controllers for the tower fans and condenser water

trade-off between the chiller power, pump power and fan power
was optimized, resulting in an 8.6% decrease in the annual system
electricity consumption. Furthermore, the optimal control in
Scheme 8 could bring about a 14.8% reduction in the annual
water consumption.
To investigate the economic benets of the energy-efcient
technologies for the chiller system, the annual operating costs and
life cycle costs were calculated and compared. Operating costs
(such as repair and maintenance costs) other than the electricity
and water charges were assumed to be the same value for all the
schemes and they were excluded in this cost analysis. Regarding
the calculation of the life cycle operating costs, a life span of 15
years was assumed for the chiller system and a discount rate of 10%
was considered throughout the lifespan. Table 5 gives the
calculation results. The annual electricity cost dictates the overall
operating cost based on its overwhelming proportion and so
reducing electricity use is an effective means to lower the
operating cost.

Table 4
Annual chiller plant electricity and water consumption per unit of air-conditioned oor area of the hotel.
3

8 (% saving from baseline)

Normalized annual electricity consumption (kWh/m2)

Chiller
68.5
63.2
Chilled water pump
9.4
9.4
Condenser water pump
7.2
7.2
Cooling tower fan
3.8
7.3
Total
88.9
87.1

Scheme

1 (baseline)

66.6
9.4
7.2
4.7
87.9

63.2
9.4
7.2
5.2
85.0

70.9
9.4
3.3
1.2
84.7

66.7
9.4
3.3
7.2
86.6

69.1
9.4
3.3
1.5
83.2

67.0 (2.24%)
9.4 (0%)
3.3 (54.17%)
1.6 (58.84%)
81.3 (8.63%)

Normalized annual water consumption (L/m2)

313.3

341.7

353.0

245.6

316.0

258.4

266.9 (14.81%)

380.6

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

208

Table 5
Annual operating costs and life cycle costs per unit of air-conditioned oor area of the hotel for the chiller system with eight operating schemes.
Scheme

1 (baseline)

8 (saving from baseline)

Normalized annual operating cost (HK$/m )

Electricity
95.2
93.3
Water
5.9
7.0
Total
101.2
100.3

94.2
6.2
100.4

91.0
6.4
97.4

89.5
4.6
94.1

91.4
5.8
97.2

87.9
4.8
92.7

86.1 (9.57%)
5.0 (15.91%)
91.1 (9.94%)

Normalized life cycle operating cost (HK$/m2)

Electricity
724.4
709.5
Water
45.0
53.5
Total
769.5
762.9

716.7
47.0
763.7

692.1
48.6
740.7

680.4
35.1
715.4

695.5
43.9
739.3

668.5
36.7
705.2

655.1 (9.57%)
37.9 (15.91%)
693.0 (9.94%)

pumps and the associated control system. After the operating cost
savings fully recoup the implementation cost, the building owner
can enjoy a prot of around HK$3,025,722 due to the optimal
control throughout the rest of the functional life of the system. It is
envisaged that the implementation of optimal control would be
more economically attractive if there is a subsidized scheme or
rebate for purchasing energy-efcient equipment.
6. Implementation of load-based speed control for system
optimization

place for a set of operating conditions in terms of various

combinations of PLRs from 0.2 to 1 at 0.1 intervals and wet-bulb
temperatures from 16 to 28 8C at 4 8C intervals. Using regression
analysis, a linear relationship between the optimum fan speed
(Sfan,op) and chiller PLR can be obtained, as given by Eq. (2) with the
coefcient of determination (R2) of 0.9215. Sfan,full represents the
full speed of the tower fans. Such an adjustment of fan speed
should be generic for any given chiller and cooling tower design
but each specic design has its own set of constant coefcients.
Sfan;op 0:7281PLR 0:1776Sfan;full

In this analysis, improved system performance relies on

applying variable speed control to the condenser water pumps
and cooling tower fans. For system optimization, the condenser
water ow rate should vary in direct proportion to the chiller load.
This results in the control algorithm of pump speed (Spump,op)
shown in Eq. (1), given that speed is directly proportional to ow
rate according to the pump laws. The minimum speed is set at half
of the full speed (Spump,full) to ensure the minimum condenser
water ow required when the chiller load in terms of part load
ratio (PLR) drops to below 0.5.
Spump;op

PLRSpump;full
0:5S pum p; full

if PLR > 0:5

if PLR  0:5

(1)

Following the traditional control of cooling water temperature,

the controller for tower fan speed modulation has to evaluate the
optimum set point (Tctwl,op) and operates the fan at the right speed
to meet that set point. The evaluation of Tctwl,op is clumsy and its
adjustment involves measurements of ambient wet-bulb temperature and chiller load. Based on this simulation analysis, it is
possible to apply load-based speed control for cooling tower fans in
order to achieve optimum system operation. Fig. 3 shows data of
the optimum fan speed at which the maximum system COP took

(2)

The load-based speed control for the tower fans brings about a
slight increase of 0.4 kWh/m2 from 81.3 kWh/m2, the lowest
annual electricity consumption of the chiller system in Scheme 8.
The proposed control eliminates the need to determine the
optimum cooling water temperature based on the ambient wetbulb temperature which is likely to be suffered from imprecise
measurement. Furthermore, the optimal control of the whole
system can be highly simplied. This is because the sequencing of
chillers, pumps and tower fans and their individual speed controls
can be based entirely on the chiller load conditions only.
7. Conclusions
This paper presents the economic benets of a water-cooled
chiller system with the individual and mixed uses of energyefcient technologies: variable speed congurations for condenser
water pumps and cooling tower fans; improved cooling water
control methods (constant set point of 18.3 8C, a xed approach of
4 8C to wet-bulb temperature and optimized method to minimize
the overall power). Drawing on a sophisticated model, the
characteristics of improved chiller system COP with various
energy-efcient technologies were identied. Load-based speed
control is introduced for the tower fans and condenser water
pumps to achieve system optimization with minimum power
consumption. Implementing the optimum control to a chiller
system could bring about a saving of HK$10.1/m2 in the annual
operating cost per unit of the total air-conditioned oor area of a
reference hotel in Hong Kong. It is expected that load-based speed
control is generic for all types of multiple-chiller systems with full
or partial use of variable speed drives for the system components.
The economic analysis reported here can help hotel owners realize
the likely reduction in operating costs and decide which
technology is cost-effective or even prot-making with respect
to their budget for implementing chiller system improvement
programmes.
Acknowledgement

Fig. 3. Load-based speed control for cooling tower fans.

The work described in this paper was supported by a grant from

the central research grant of The Hong Kong Polytechnic
University, Project A/C Code: G-U565.

F.W. Yu, K.T. Chan / Energy and Buildings 42 (2010) 203209

References
[1] W.W. Chan, J.C. Lam, Prediction of pollutant emission through electricity consumption by the hotel industry in Hong Kong, International Journal of Hospitality
Management 21 (2002) 381391.
[2] F.W. Yu, K.T. Chan, Electricity end-use characteristics of air-cooled chillers in
hotels in Hong Kong, Building and Environment 40 (2005) 143151.
[3] S. Deng, J. Burnett, A study of energy performance of hotel buildings in Hong Kong,
Energy and Buildings 31 (2000) 712.
[4] F.W.H. Yik, J. Burnett, I. Prescott, Predicting air-conditioning energy consumption
of a group of buildings using different heat rejection methods, Energy and
Buildings 33 (2001) 151166.
[5] T. Hartman, All-variable speed centrifugal chiller plants, ASHRAE Journal 43 (9)
(2001) 4353.
[6] W.P. Bahneth, E.B. Peyer, Varying views on variable-primary ow, HeatPiping
Air Conditioning Engineering 76 (3) (2004) S59.
[7] J.M. Gordon, K.C. Ng, H.T. Chua, C.K. Lim, How varying condenser coolant ow rate
affects chiller performance: thermodynamic modeling and experimental conrmation, Applied Thermal Engineering 20 (2000) 11491159.
[8] M. Hydeman, P. Sreedharan, N. Webb, S. Blanc, Development and testing of a
reformulated regression-based electric chiller model, ASHRAE Transaction 108 (2)
(2002), paper no. HI-02-18-2.

209

[9] H. Crowther, J. Furlong, Optimizing chillers and towers, ASHRAE Journal 46 (7)
(2004) 3440.
[10] L. Lu, W. Cai, Y.C. Soh, L. Xie, S. Li, HVAC system optimizationcondenser water
loop, Energy Conversion and Management 45 (2004) 613630.
[11] R.D. Graves, Thermodynamic modeling and optimization of a screw compressor
chiller and cooling tower system, Msc Thesis, The Ofce of Graduate Studies of
Texas A&M University, 2003.
[12] J.E. Braun, G.T. Diderrich, Near-optimal control of cooling towers for chilled water
systems, ASHRAE Transaction (1990), SL-90-13-3.
[13] D.J. Benton, M. Hydeman, C.F. Bowman, P. Miller, An improved cooling tower
algorithm for the CoolToolsTM simulation model, ASHRAE Transaction 108 (1)
(2002), paper no. AC-02-9-4.
[14] Solar Energy Laboratory, TRNSYS: A Transient System Simulation Program (Reference Manual), University of Wisconsin/Madison Press, Madison, WI, 2000.
[15] F.W. Yu, K.T. Chan, Optimization of water-cooled chiller system with load-based
speed control, Applied Energy 85 (2008) 931950.
[16] Electrical, Mechanical Services Department, Pilot Scheme for Wider Use of Fresh
Water in Evaporative Cooling Towers for Energy-efcient Air Conditioning
Systems, Author Press, Hong Kong, 2000.
[17] Air-Conditioning & Refrigeration Institute, Standard 550/590: Performance Rating of Water Chilling Packages Using the Vapor Compression Cycle, Author Press,
Arlington, Virginia, 2003.