Energy and Buildings 60 (2013) 100109

Contents lists available at SciVerse ScienceDirect

Energy and Buildings

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

Thermo-economic optimization of an ice thermal energy storage system for

air-conditioning applications
Sepehr Sanaye , Ali Shirazi
Energy Systems Improvement Laboratory (ESIL), School of Mechanical Engineering, Iran University of Science and Technology (IUST), Narmak, Tehran 16844, Iran

a r t i c l e

i n f o

Article history:
Received 19 November 2011
Received in revised form 27 October 2012
Accepted 26 December 2012
Keywords:
Ice thermal energy storage system
Air-conditioning
Thermo-economics
Environmental
Genetic algorithm
Optimization

a b s t r a c t
A major portion of electricity consumption in buildings in residential, administrative, and commercial
sectors is related to air-conditioning (A/C) systems. To reduce and shift the electricity consumption of
A/C systems from on-peak hours to off-peak hours, an ice thermal energy storage (ITES) can be utilized.
In this paper, thermo-economic optimization of an ITES system was carried out for A/C applications. In
order to consider the environmental aspects, a penalty cost was considered for CO2 emission. Applying
the genetic algorithm optimization technique, the optimum values of system design parameters were
obtained. The objective function included the capital and operational costs as well as the penalty cost
due to CO2 emission, without and with costs associated with exergy destruction. The results indicated
that, on average, the amount of electricity consumption and CO2 emission of ITES system were lower 9%
and 9.8%, respectively, in comparison with those of a conventional system. Furthermore, the ITES extra
capital cost could be paid back through savings in electricity cost in 3.43 years.
2013 Elsevier B.V. All rights reserved.

1. Introduction
The large part of electricity consumption in buildings is allocated to A/C systems. In addition, due to the limited resources of
fossil fuels and also strict environmental protection rules, nding
an appropriate way to reduce energy consumption is necessary.
Several methods are currently used to reduce energy consumption in buildings, which can be divided into two main categories
of active and passive methods. The passive techniques include
shading of facades and fenestrations, use of thermal insulation
material, and consideration of a proper orientation for the buildings envelop. Management of the building heat loads through
dynamic tariff strategy, optimum operation design, and use of thermal energy storage (TES) are examples of the active methods [1].
The basic principle behind using TES systems is shifting the electricity consumption of building cooling from on-peak hours (during
daytime) to off-peak hours (during night-time).
TES systems are divided into two major categories including
sensible heat storage (e.g. water and stone) and latent heat storage (e.g. water/ice mixtures and salt hydrates) [2]. In the rst
type, energy is stored by changing the temperature of energy storage media (without phase change). In the second type, energy is
stored by changing the phase of energy storage media at a constant

Corresponding author. Tel.: +98 21 77240192; fax: +98 21 77240192.

E-mail addresses: sepehr@iust.ac.ir (S. Sanaye), ashirazi@mecheng.iust.ac.ir
(A. Shirazi).
0378-7788/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.enbuild.2012.12.040

temperature. Considering the same volume, latent heat storage

systems can store more energy in comparison with sensible heat
storage systems [3]. That is why latent TES systems have been
receiving much more attention in recent years, mainly in ice storage
for use in A/C applications [4]. Although utilizing cold storage systems and their auxiliary equipment incur extra initial costs, these
additional costs are paid back in a short time because of reduction in
electricity consumption in comparison with conventional systems.
Numerous studies have been carried out on modeling of cold
thermal storage systems. MacPhee and Dincer [4] evaluated the
performance of ice storage charging and discharging processes on
the basis of energy and exergy analyses. The results showed that
the exergy analysis is important due to the existing thermodynamic irreversibility in ice storage system. Dincer and Rosen [5]
also reviewed and reported the technical, economical, and environmental advantages of energy storage systems. The mathematical
modeling and optimization of a chilled water thermal storage
system was carried out by Henze et al. [6]. In their investigation, they considered the capital investment costs as the objective
function. The results showed that addition of a TES system to
conventional cooling systems is cost-effective. Henze [7] also evaluated the performance of four control strategies for an ITES system
in a commercial unit. Six parameters were investigated on the
system performance: Storage losses, utility rate structures, rate
periods, penalty for ice making, storage capacity, and the impact
of load forecasting. The results showed that control strategy with
the priority of storage has the lowest operational cost. A thermal
energy storage module for three different types of ITES systems

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

Nomenclature
A
c
celec
C elec
C env
C tot
COP
cp
CRF
E
F
h
i
iph
k

m
N
n
NTU
Obj
p
Q
QC
Q C
QST
Rth
s
T
U
V

W
Z
Z

heat transfer surface area (m2 )

the unit cost of exergy
electricity cost (US$/kWh)
cost rate of electricity consumption (US$/s)
penalty cost rate of CO2 emission (US$/s)
total cost rate (US$/s)
coefcient of performance
specic heat at constant pressure (kJ/(kg K))
capital recovery factor
exergy ow rate (kW)
logarithmic mean temperature difference correction factor
specic enthalpy (kJ/kg)
interest rate (%)
melting latent heat (kJ/kg)
specic heat ratio
mass ow rate (kg/s)
operational hours in a year
system life time (year)
number of transfer unit
objective function
pressure (Pa), extra cost payback period (year)
the time rate of heat transfer (kW)
cooling load (kWh)
cooling load (kW)
stored cooling energy (kWh)
total thermal resistance (m2 K/kW)
specic entropy (kJ/kg K)
temperature (K)
overall heat transfer coefcient (kW/(m2 K))
volume (m3 )
the time rate of energy transfer by work (kW)
capital cost (US$)
capital cost rate (US$/s)

Greek letters

thermal efciency
CO2
CO2 emission factor (kg/kWh)
density (kg/m3 )


maintenance factor

relative humidity (%)

absolute humidity (kg water vapor/kg dry air)

Subscripts
air
a
AHU
air handling unit
amb
ambient
chemical
CH
ch
charging
compressor
Comp
Cond
condenser
cooling tower
CT
cv
control volume
CW
chilled water
D
destruction
discharging
dc
EV
evaporator
EX
expansion valve
fuel
F
FP
freezing point
int
initial
number of ow line
k

l
PH
r
ST
sv
t
w
WB

101

leakage
physical
refrigerant
storage tank
salvage value
time
water
wet-bulb

(ice-on-coil external melt, ice harvester, and ice-on-coil internal

melt) was developed by Ihm et al. [8]. In their investigation, they
evaluated the potential cost savings associated with the use of
TES system for various conventional control strategies as well as
TES chiller and storage tank sizes and nally determined the best
control strategy for a building equipped with a TES system. Habeebullah [1] investigated the economic feasibility of retrotting an
ITES system for the unique A/C plant of the Grand Holly Mosque of
Mecca in Saudi Arabia in full storage and partial storage scenarios.
In his study, the operational and the capital investment costs were
considered as the objective function. The obtained results showed
that applying the full storage strategy is more reasonable to reduce
electrical energy consumption. A theoretical model and optimization of an ITES system in large commercial buildings was carried
out by Chen et al. [9]. In their economic analysis, the objective
function included initial cost and energy cost of the system, and
optimal parameters of the system were ultimately obtained. Moreover, designing, installation, and testing of a shell-and-tube TES
unit was performed by Ezana et al. [10]. They analyzed the charging and discharging processes from energy and exergy viewpoints
and drew useful concluding remarks.
Thermo-economic analysis helps designers to reach optimal
point of the system performance through a cost-effective way. This
analysis can play a key role in the design and optimization of a
thermal system. This method can be implemented by applying cost
balance equations and obtaining the exergetic unit cost for each
component in a thermodynamic cycle. In fact, through combining
economic and thermodynamic analysis, this methodology provides
very important information about thermodynamic inefciencies,
and also leads to the optimum conguration of the system.
Domanski and Fellah [11] investigated the useful aspects of
applying thermo-economic analysis in the design and operation
of TES systems. Thermo-economic optimization of a sensible thermal storage system was also performed by Badar et al. [12]. In their
analysis, they minimized the sum of entropy generation cost and
annualized capital cost of system components.
The present work covers the thermo-economic and environmental analyses as well as optimization of an ice storage
air-conditioning system to save energy/cost and reduce CO2 emission. To implement this job, thermal modeling of the system was
performed. Then, an objective function (the sum of capital cost,
operational cost, and penalty cost due to CO2 emission) without
and with the corresponding cost of system exergy destruction was
dened and ultimately minimized subject to a list of constraints.
In summary, the followings are the main contribution of this
paper into the subject:
An ITES system was modeled in detail for A/C applications and
was also analyzed from thermo-economic and environmental
viewpoints.
The optimal design of modeled system with a new objective function as well as a new list of design parameters was carried out,
which is a novel approach in TES system analysis.

102

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

The performance of modeled ITES system was compared with

a conventional cooling system from the viewpoints of electricity
consumption and the amount of CO2 emission (in a year).

where the terms pws and  are water vapor saturation pressure at
ambient dry bulb temperature and relative humidity, respectively.
pws is a function of temperature and can be estimated as follows
[14]:

2. Mathematical modeling
pws =
Schematic diagram of the ITES system which has been considered in the present study is demonstrated in Fig. 1. The whole ITES
system included two main parts:
Charging cycle including evaporator, compressor, condenser,
cooling tower, pump, and expansion valve.
Discharging cycle including air handling unit (AHU), discharging
pump, and ice storage tank.
In charging cycle (vapor compression refrigeration system),
R134a is used as refrigerant, and water/Glycol solution (chilled
water) is the cooling uid in discharging cycle (Fig. 1).
The charging cycle was used to make ice during off-peak demand
hours when the electricity price is low (usually after mid-night).
During on-peak demand hours when the electricity price is high,
the chilled water inside the tubes passed through the ice storage
tank and was pumped into AHU for cooling building (discharging
process). In this study, full storage strategy was considered for modeling of the ITES system. According to this strategy, the total cooling
energy used during on-peak hours is supplied by the storage tank
and the ice maker operates only at the night-time. Therefore, the
whole charging cycle is turned off during on-peak hours.
2.1. Energy analysis
Based on energy analysis, a thermodynamic model of the ITES
system is presented in this section. The following assumptions have
been taken into consideration while developing the system model:
Expansion valve heat losses and pressure drop within the connecting pipes were assumed negligible.
All cooling energy is stored in the water/ice medium.
All kinetic and potential effects were assumed negligible.
The states of the refrigerant at evaporator and condenser outlets
were considered as saturated vapor and saturated liquid, respectively.
The storage tank temperature distribution was assumed constant.
The relations used to model each system component in charging
and discharging cycles are given in Appendix A.The required cooling energy of the building (QC , kWh) can be estimated from its
cooling load (Q C , kW) [13]:

Q C (t) dt

QC =

(1)

tdc

where tdc is the discharging time (h). Assuming that cooling load of
the building is known, the mass ow rate of air within AHU can be
calculated as follows:
Q C
a=
m
h1 h2

(2)

The difference in specic enthalpy of moist air at points 1 and

2 (h1 h2 ), which depends on ambient dry bulb temperature (T),
absolute humidity (), and specic enthalpy of water vapor (hg )
can be expressed as:
h1 h2 = cp,a (T1 T2 ) + (1 hg1 2 hg2 )

(3)

pws
patm pws

(4)

= 0.622

C1
+ C2 + C3 T + C4 T 2 + C5 T 3 + C6 ln(T )
T

(5)

Since a part of stored cooling energy is lost due to heat transfer

between the storage tank and its surroundings, a thermal efciency
(ST ) was dened for the storage tank. Considering a constant temperature distribution within the tank, the amount of heat leakage
will be a function of inner temperature of the tank, ambient temperature, storage tank heat transfer surface area, and its thermal
resistance. During the charging process, the temperature of the tank
is TST . Therefore, the amount of heat leakage of the tank during
charging process (Ql,ch ) is
Ql,ch = AST

Tamb TST
tch
Rth

(6)

where the terms AST , Rth, and tch are the storage tank heat transfer
surface area, thermal resistance of the storage tank and the charging
time, respectively.
Similarly, the amount of heat leakage of the storage tank during
discharging process is calculated as:
Ql,dc = AST

Tamb Tdc
tdc
Rth

(7)

where Tdc and tdc are discharging temperature and discharging

time, respectively.
Therefore, the cooling energy which should be stored in the
storage vessel (QST ) can be obtained as:
QC
ST

QST =

(8)

where
ST

QST Ql,ch Ql,dc

=
=1
QST

Ql,ch + Ql,dc
QST


(9)

2.2. Exergy analysis

Exergy is dened as the maximum theoretical useful work that a
system can perform in a given state when it comes to the ambient
conditions. The method of exergy analysis is a branch of applied
thermodynamics which analyzes thermal systems from the second
law of thermodynamics standpoint. This type of analysis focuses
not only on the quantity of energy, but also on its quality. The
exergy balance equation for a system (Eq. (10)) accounts for the
time rate of exergy transfer due to heat transfer (E Q ) and work (E W ),
the exergy transfer rate at the control volume inlet (E i ) and outlet
(E e ), and the rate of exergy destruction (E D ) due to thermodynamic
irreversibility within the control volume.
E D = E Q E W +

cv +
W


i

E i

E i

E e

E e =

 
Q j

T0
Tj

(10)

In absence of electromagnetic, electric, nuclear, and surface tension effects and assuming negligible values of change in potential
and kinetic energy, the exergy ow rate of various components
of ITES system is divided into two parts of physical and chemical
exergy [15,16]:
E = E PH + E CH

(11)

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

103

Fig. 1. Schematic diagram of the modeled ITES system.

Physical exergy in the general form is dened as follows:

E PH = m[(h
h0 ) T0 (s s0 )]

3.1. The objective function

(12)

where for liquids (incompressible uids):

p T0
E PH = mc

T
T0

1 ln

 T 

(13)

T0

while for ideal gases:

E

PH

p T0
= mc

T
T0

1 ln

T
T0

k1
ln
+
k

 p 
p0

In this paper, the rst objective function includes the rate of

investment and maintenance costs (Z k ), operational costs (C elec ),
and the rate of penalty cost due to CO2 emission (C env ) which
was included in the objective function to cover the environmental approaches. Therefore, the rst objective function represents
the total cost rate of the system in terms of US dollar per second:
Obj(1) =

Z k + C elec + C env

(15)

(14)

In this paper, since any chemical reaction did not occur, the
chemical exergy was not used in the modeling of ITES system.

The investment cost of each cycle component (Zk ) can be estimated based on the cost functions listed in Table 1 [1,15,1720].
Due to the fact that the values of Zk should be used in form of
investment cost rate (Z (US$/s)), one may write:

3. Thermo-economic-environmental optimization

Z k =

In order to identify the sources of thermodynamic inefciencies, estimate the thermodynamic irreversibility of components,
and obtain optimum design parameters which improve the overall efciency of the ITES system, thermo-economic-environmental
optimization of the system was carried out. To perform this task,
two types of objective functions were introduced. The purpose of
thermo-economic optimization here is to estimate those design
parameters which minimize the objective function.

where N, and CRF are the operational hours of ITES system in a

year, maintenance factor, and the capital recovery factor, respectively. The capital recovery factor depends on the annual interest
rate (i) as well as estimated equipment life time (n) which is dened
as:

Zk CRF
N 3600

CRF =

i(1 + i)
i

(16)

(1 + i) 1

(17)

104

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

Table 1
The cost functions of various equipments in ITES system [1,15,1720].
System component

Capital cost function

Air handling unit (AHU)

ZAHU = 24202 A0.4162

AHU

Pump

0.71
pump
Zpump = 705.48 W
1+

Ice storage tank

ZST = 8.67 10[2.9211 exp(0.1416log VST )]

Evaporator

ZEV = 16648.3 A0.6123

EV

Compressor

ZComp =

r
39.5m
0.9Comp

0.2
1pump

pdc 
pdc 
psuc

ln

psuc

r
= 114.5 m

Expansion valve

ZEX

Condenser

ZCond = (516.621 ACond ) + 268.45

Cooling tower

CT )
ZCT = 746.749 (m

Compression chiller

ZChiller = 150.2 Q Chiller

0.79

(TCT )

Operational cost of the system (including the cost of electricity

consumption of each system component during off-peak and onpeak hours) can be expressed as follows:

Comp + W
pump,CT + W
fan,CT )
C elec = (W

pump,dc + W
fan,AHU )
+ (W

celec,off-peak

3600

celec,on-peak
3600

(18)

The penalty cost of CO2 emission (cCO2 ) was considered as 90 US

dollars per ton of carbon dioxide emissions [21]. Using CO2 emission factor (CO2 ), the amount of CO2 produced in ITES system is
obtained as:
mCO2 [kg] = CO2 [kg/kWh]annual electricity consumption[kWh]
(19)
where CO2 is 0.968 (kg/kWh) [22].
Thus, the rate of penalty cost of CO2 emission is dened as follows:
C env =

(mCO2 /1000) cCO2

Z k + C elec + C env +

C D,k

(21)

where the last term (C D,k = cF,k E D,k ) represents the exergy destruction cost rate in kth system component. This additional term caused
the optimal design parameters to shift toward the thermodynamically more efcient values. The rate of exergy destruction in each
component (E D,k ) was computed through Eq. (10). To calculate the
unit cost of fuel for each component (cF,k ), it is necessary to calculate the unit cost of exergy for each ow line (cj ) through solving
the exergy-cost balance equation in each system component in the
general form of:

(cj E j )k,in + Z k =

(Tin,CT TWT,out )

0.9924

(0.022 TWB,out + 0.39)

2.447

For most of the well-designed equipment, as the amount of

exergy destruction decreases, or in other words, the efciency
increases, the cost of exergy destruction diminishes, and in return,
the cost of capital investment goes up. In thermo-economic optimization, the main point is to achieve the best balance between C D,k
and Z k . In order to obtain this, a parameter named exergo-economic
factor (fk ) was dened for each component, which indicates the
ratio of the capital investment cost (non-exergy related cost) to the
total cost [15]:
fk =

Z k
Z k + cF,k (E D,k + E loss,k )

(23)

where E loss,k is the rate of exergy loss in each component. If the

value of fk for a component is small, it is suggested to improve
the efciency of that component by reducing the rate of exergy
destruction. This is done by increasing the capital investment cost
of that component. On the other hand, a higher value of fk implies
the need for decreasing capital investment cost, even at the expense
of a relative increase in exergy destruction of that component.
3.2. Design parameters

0.57

(20)

N 3600

For the more accurate analysis and greater emphasis on the

thermodynamic inefciencies, one may introduce the following
objective function:
Obj(2) =

(cj E j )k,out

(22)

where E j and cj are the exergy ow rate and the unit cost of exergy
in each ow line, respectively. Applying relation (22) for each system component provided a system of k equations which are given
in Appendix B. The matrices of computing cj values from the constructed exergy-cost balance equations as well as the unit cost of
fuel for each component are also listed in Appendix B. It should be
noted that the term C D,k is a hidden cost that can only be revealed
through thermo-economic analysis.

In this study, the design parameters (decision variables) are

chilled water temperature at AHU inlet (T3 ) and outlet (T4 ), storage
temperature within the ice storage tank (TST ), refrigerant saturated
temperature at evaporator (TEV ) and condenser (TCond ).
The list of mentioned design parameters and their range of variation as well as the constraints for system optimization are listed
in Table 2.
As shown in Table 2, the range of condenser saturated temperature is given by (TWB,out ) + 5 < TCond < 60. This range varies
Table 2
The ITES design parameters and their range of variation as well as the constraints
for system optimization.
Constraints

Reason

3 < T3 < 5
11 < T4 < 13
10 < TST < 0
30 < TEV < 0

Typical data for refrigeration systems

Typical data for refrigeration systems
Typical data for refrigeration systems
Minimum and maximum refrigerant saturation
temperature in evaporator for a wide range of
applications
Minimum and maximum refrigerant saturation
temperature in condenser for a wide range of
applications
For occurring heat transfer between evaporator
and storage tank
To avoid occurring icing phenomenon of
water/glycol solution at discharge cycle

(TWB,out ) + 5 < TCond < 60

TEV < TST

TFP,Glycol < TST

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

105

completely in line with relation between ambient wet-bulb

temperature and condenser saturated temperature. Indeed, the
left-hand-side limit considered in this study (TWB,out + 5) guarantees that Tcond is always ve degrees Centigrade above TWB,out [23].
3.3. Genetic algorithm optimization technique
The genetic algorithm considers an optimization problem as an
evolutionary problem. The rst step in genetic algorithm is to represent a correct solution to the problem by a string of genes. This
string of genes, which represent a solution, is known as a chromosome. Then an initial population of legal chromosome is needed
to start the procedure. The genetic algorithm repeatedly modies
the population of individual solutions. At each step, individuals
are selected at random from the current population to be parents.
These parents are used to produce the children for the next generation. The genetic process will end if there is no change in the
optimal values (populations best tness) for a specied number of
generations.
In this study, genetic algorithm technique was used to optimize
the objective functions mentioned in Eqs. (15) and (21) to obtain
the optimum design parameters of the ITES system.
4. The payback period
Using ITES system, in comparison with conventional cooling
system, imposes additional expenses. These extra expenses arise
from the capital investment and maintenance costs of the storage tank. These additional costs can be compensated over time
with the reduction in electricity consumption (in comparison to
conventional systems) when ITES systems are applied.
The payback period of additional expenses (in years) can be
estimated as follows [24,25]:


(Zop )

(1 + i) 1
i(1 + i)


+ ZSV

i
(1 + i)

=




Zk

(24)

where Zsv is the difference in salvage values of ITES and conventional systems. The salvage value is dened as the estimated value
that an asset will realize upon its sale at the end of its useful life.

Fig. 2. The typical commercial building cooling load prole and the chiller load
capacity for both ITES and conventional systems in a day.

Therefore, Zsv is given as a percentage of the difference

between



investment costs of ITES and conventional systems. 
Z is
k k
the additional investment cost for substituting ITES system instead
of the conventional system. Zop is savings in annual operational
cost (the difference in annual operational cost of conventional and
IETS systems). Finally, i and p are the interest rate and the payback
period.
Eq. (24) is a non-linear equation in terms of p which was solved
by NewtonRophson iterative numerical method.
5. The case study
The investigated case study was a commercial building in
Ahwaz, a city in south of Iran. The working hours of the building
were 7 AM to 7 PM. The daily cooling load of the building is shown
in Fig. 2. The chiller cooling load capacity for ITES and conventional
(system with capability of load change (increase or decrease) steps
equal to 25% of nominal cooling load) systems are also shown in
Fig. 2. Fig. 3 illustrates the variation of maximum values of ambient
temperatures during a year in Ahwaz [26]. Table 3 outlines the
approximate thermal energy stored in the storage tank during

Table 3
The building required cooling load, the ITES cooling and chiller load capacities for our case study during 24 h of a day.
Time of day (h)

Process

Storage (kW)

Building load (kW)

Chiller capacity in
ITES system (kW)

Chiller capacity in conventional

system (kW)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

Charging
Charging
Charging
Charging
Charging
Charging
Charging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging
Discharging

Charging

1513.4
1513.4
1513.4
1513.4
1513.4
1513.4
1513.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1513.4

0
0
0
0
0
0
0
305
404
569
779
1131
1490
1756
1926
1750
1125
751
0
0
0
0
0
0

1513.4
1513.4
1513.4
1513.4
1513.4
1513.4
1513.4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1513.4

0
0
0
0
0
0
0
481.5
481.5
963
963
1444.5
1926
1926
1926
1926
1444.5
963
0
0
0
0
0
0

106

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

Table 4
r,
The comparison of computed values of system operating parameters including m
Comp , the compressor power consumption, and
the refrigerant mass ow rate, W
COP, coefcient of performance, obtained from modeling of vapor compression
refrigeration system with the corresponding values reported in Ref. [30].
Inputs
TEV ( C)
TCond ( C)
Q EV (kW)

Fig. 3. Variation of the minimum and maximum values of ambient temperatures in

a year for Ahwaz [26].

charging and discharging processes, the building cooling load

as well as the chiller capacity of ITES and conventional systems
quantitatively. As is shown, during off-peak hours when the
cooling load of commercial building is zero, the ice maker stores
the building required cooling energy in the storage tank. During
the day (on-peak hours), the refrigeration system is turned off
(discharging process), and the storage tank provides the building
required cooling load.
The refrigerant used in the modeling of the refrigeration system is R314a, and all of its properties were obtained from in-house
developed software. The thermal resistance (Rth) of the storage
tank is 1980 m2 K/kW [27]. The comfort temperature and the relative humidity of the room as well as the ambient pressure are
assumed to be 21 C, 0.55, and 100 kPa, respectively. The electricity
cost during on-peak hours was 0.09 US$/kWh and during off-peak
hours was 0.06 US$/kWh [3].
To determine CRF (Eq. (17)), the annual interest rate, approximate life time of the system, and the maintenance factor were
considered as 14% [28], 15 years, and 1.06, respectively. The salvage
value ( Zsv ) was considered to be 10% of the difference between
the capital costs of two systems [29]. The annual operational hours
of ITES system (N) from March to December (Fig. 3) for the charging
(12 PM to 7 AM) and discharging (7 AM to 7 PM) cycles were 2100
and 3600 h, respectively.
6. Discussion and results
6.1. Modeling and optimization results
6.1.1. Model verication
From our modeling procedure presented in this paper, the basic
parameters of the refrigeration system, including the refrigerant
mass ow rate, compressor power consumption, and the coefcient
of performance (COP) were obtained from the modeling results
and compared with the corresponding values given in Ref. [30]. As
shown in Table 4, in this comparison, the difference percent points
were less than 1.5%, which is an acceptable value.
6.1.2. Optimization results
In order to achieve the optimal performance of the ITES system
from thermo-economic and environmental viewpoints, the optimum design parameters (decision variables) were computed by
considering two objective functions (1) and (2).

20
40
25.9

Outputs

Reported

Modeling
results

Difference
(%)

r
m
Comp
W
COP

0.2
9
2.87

0.2001
9.1102
2.8430

0.05
1.22
0.94

The genetic algorithm optimization was applied for 100 generations, using a search population size of 100 individuals, cross over
probability 0.9, and gene mutation probability 0.1.
Applying objective function (1) including the cost of investment
and maintenance, operational cost, and penalty cost due to CO2
emission, the optimum design parameters were obtained to minimize the objective function. In the next step, objective function
(2) was used to optimize the system performance, and optimum
design parameters of the system were determined accordingly. In
this case,
 of exergy destruction cost rate of system compo
the sum
nents
was also added to the objective function. In other
C
k D,k
words, in the second objective function, the hidden costs due to
exergy destruction of the system components were re-emphasized.
The optimal values of system design parameters for optimization
of objective functions (1) and (2) are given in Table 5. As shown in
Table 5, the optimum value of TEV is lower than TST about 2.6 C,
which is a typical temperature difference based on the results
reported in Ref. [4].
The amount of total exergy efciency of the system was computed based on the optimum design parameters obtained from
objective functions (1) and (2), with total efciencies of 30.34%
(with the total cost of 1.3441 MUS$) and 32.49% (with the total
cost of 1.4337 MUS$), respectively. The results show that objective
function (2) considers hidden costs related to the system thermodynamic irreversibility. Therefore, objective function (2) may
be considered more appropriate than objective function (1) when
the electricity and investment cost (and therefore the total cost) is
relatively high.
The difference between results of using objective functions (1)
or (2) can also be analyzed for optimum values of Tcond . In fact, with
increasing Tcond , the compressor electricity consumption increases,
which results in a decline in COP of the system. It should be noted
that objective function (1) does not sense the cost of irreversibilities and exergy destruction in various parts of the system. On the
other hand, objective function (2) considers hidden costs related to
irreversibilities and exergy destruction within system components,
which increases with rise in Tcond . Thus, the optimum value of Tcond
obtained from optimization of objective function (2) decreased
compared to objective function (1) to shift the system design to
a thermodynamically more efcient point.

Table 5
The optimum values of design parameters (decision variables) by using objective
functions (1) and (2).

T3 ( C)
T4 ( C)
TST ( C)
TEV ( C)
TCond ( C)

Optimum values
(objective function (1))

Optimum values
(objective function (2))

3.86
12.39
2.51
5.20
38.05

3.31
12.80
1.68
4.27
36.11

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

107

Table 6
The investment cost, exergy destruction cost, exergy destruction and exergo-economic factor for various components of the ITES system optimized at various ambient
temperatures 30, 35, 40, 45 and 50 C.
Ambient temperature

AHU

ST

EV

Comp

EX

Cond + CT

30

Investment cost (MUS$)

Exergy destruction cost (MUS$)
Exergy destruction (kW)
Exergo-economic factor (%)

0.2238
0.2501
79.54
22.4

0.0963
0.1343
97.25
20.2

0.1093
0.1905
83.08
52.1

0.2285
0.1677
176.53
56.3

0.0007
0.0062
18.68
14.8

0.1883
0.5536
278.06
25.3

35

Investment cost (MUS$)

Exergy destruction cost (MUS$)
Exergy destruction (kW)
Exergo-economic factor (%)

0.2249
0. 2575
97.16
20.8

0. 0964
0. 1392
98.71
19.2

0.1095
0.1996
83.88
50.3

0.2481
0.1779
182.01
55.6

0.0007
0.0068
22.18
14.2

0.2246
0.5715
285.31
24.7

40

Investment cost (MUS$)

Exergy destruction cost (MUS$)
Exergy destruction (kW)
Exergo-economic factor (%)

0.2262
0.2652
119.21
19.9

0.0965
0.1434
100.19
18.4

0.1096
0.2091
84.75
49.4

0.2741
0.1885
187.06
54.9

0.0008
0.0075
26.17
13.8

0.2586
0.5886
293.95
24.1

45

Investment cost (MUS$)

Exergy destruction cost (MUS$)
Exergy destruction (kW)
Exergo-economic factor (%)

0.2273
0.2624
142.92
19.3

0.0967
0.1476
102.19
17.9

0.1098
0.2194
85.49
48.8

0.2909
0.2076
192.49
54.2

0.0008
0.0083
30.87
12.9

0.2906
0.6063
302.77
23.7

50

Investment cost (MUS$)

Exergy destruction cost (MUS$)
Exergy destruction (kW)
Exergo-economic factor (%)

0.2285
0.2799
171.51
18.4

0.0969
0.1520
104.24
17.3

0.1099
0.2305
86.48
48.1

0.3093
0.2199
198.06
53.6

0.0009
0.0092
36.42
11.9

0.3285
0.6245
311.86
23.1

6.2. Investment costs, exergy destruction costs and

exergo-economic factor
The amount of investment cost, exergy destruction cost, exergy
destruction rate, and exergo-economic factor for various components of ITES system when the ambient temperature was varied
from 30 C to 50 C are presented in Table 6 (for the design parameters obtained from optimization of the objective function (2)). It
was observed that the investment cost of each component increases
due to the increase in the capacity size of system components when
the ambient temperature rises. For instance, with increase of the
ambient temperature from 40 C to 45 C, the compressor power
consumption, as well as the evaporator and condenser heat transfer surface area and their investment cost increase (6.1%, 0.2% and
8.3%, respectively). The exergy destruction cost of each component (as well as the exergy destruction rate) also goes up when the
ambient temperature rises due to increase in the entropy generation of the system as a result of thermodynamic irreversibility. It is
observed that the highest values of thermodynamic irreversibility
and exergy destruction cost belong to the condenser and the cooling tower (which varies from 0.5536 MUS$ to 0.6245 MUS$ for the
various ambient temperatures). The next highest exergy destruction costs were related to AHU (increasing from 0.2501 MUS$ to
0.2799 MUS$ at the various ambient temperatures), evaporator
(from 0.1905 MUS$ to 0.2305 MUS$), and the compressor (from
0.1677 MUS$ to 0.2199 MUS$), respectively.
Table 6 also shows that the amount of exergo-economic factor for each component decreases when the ambient temperature
increases. This is due to more intense increase of exergy destruction rate of each component at higher temperatures in comparison
with the corresponding capital investment cost (Eq. (23)).

cooling energy provided by the conventional system is more than

the required demand, which is clearly shown in Fig. 2. It should be
mentioned that the conventional chiller system considered in this
paper is a four steps chiller which is commonly available at the local
market and is capable of load change equal to 25% of its nominal
cooling load in each step, which results in producing more cooling
energy than the required demand. However, the ITES system provided the exact required cooling load by running the chilled water
pump at the appropriate times to control the room temperature by a
thermostat. Thus, although the ITES system has a lower evaporation
temperature compared to the conventional one (which leads to an
increment in its electrical consumption), owing to the fact that the
conventional system produces an excessive cooling energy (more
than required cooling load), the overall electrical consumption of
the ITES system is less than the conventional one.
Moreover, on average, 0.577 106 kg carbon dioxide emission
was prevented to enter the atmosphere in comparison with that of
the conventional system (9.8% decrease) in our case study.
Finally, according to Eq. (24), the payback period of the extra cost
of ITES system relative to a conventional system was estimated 3.43

6.3. Performance comparison of ITES system with a conventional

system
To compare the performance of ITES and conventional cooling
systems, the amount of electricity consumption in two systems
were compared with similar input conditions. The results of this
comparison are presented in Fig. 4. The results show that the
electricity consumption of ITES system is lower than that of conventional system (in this study about 9%). The reason is that the

Fig. 4. The comparison of electricity consumption between ITES and conventional

systems in a year.

108

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

years. Therefore, the additional expenses imposed by the installation of the storage tank in the ITES system can be compensated
in less than four years due to reduction in electricity consumption
when ITES system was applied.
7. Conclusions

Condenser heat transfer rate:

r (h7 h8 )
Q Cond = m
Refrigeration system COP
COP =

Thermo-economic modeling and optimum design of an ITES system was performed for A/C applications. In order to consider the
environmental aspects, a penalty cost was considered for CO2 emission. The whole system consisted of charging (including a vapor
compression refrigeration cycle) and discharging cycles (including
a storage tank and an air handling unit).
The objective function in the system optimization was the total
cost (including the capital investment cost, operational cost, and
the penalty cost due to CO2 emission) without and with the corresponding cost of total exergy destruction by which the objective
functions (1) and (2) were introduced, respectively. The optimum
values of design parameters obtained from the objective function
(2) indicated more thermodynamically efcient design (with the
higher total cost).
The results of a case study demonstrated that, on average, the
amounts of annual electricity consumption and CO2 emission of the
ITES system, relative to a conventional cooling system decreased
by 9% and 9.8%, respectively. In addition, the extra cost of ITES system relative to a conventional one was compensated and recovered
within 3.43 years.

(A.6)

Q EV
Comp
W

(A.7)

Cooling tower mass ow rate:

CT = 43.2 103 Q Cond
m

(A.8)

Evaporator heat transfer surface area:

AEV =

p )min
NTU (mc
UEV

(A.9)

Condenser heat transfer surface area:

ACond =

Q Cond
U F TLMTD

(A.10)

Appendix B.
The exergy-cost balance equations for various components of
ITES system based on Eq. (22) are written as follows:
Air handling unit (AHU):
c1 E 1 + c3 E3 + Z AHU = c2 E 2 + c4 E 4

Appendix A.

(B.1)

Ice storage tank:

A.1. Air handling unit

The overall heat transfer coefcient, heat transfer surface area
and pressure drop for air handling unit (AHU) were obtained from
Ref. [3].
A.2. Ice storage tank

Q
c4 E 4 + cST E CW
+ Z ST = c3 E 3 + cST,int E ST,int

(B.2)

where E ST,int is the initial exergy ow rate in storage tank before

starting charging process.
Evaporator:
cST,int E ST,int + c5 E 5 + Z EV = cST E ST,ch + c6 E 6

The following relations are used to determine the storage tank

volume and its heat transfer surface area [4,31]:
VST =

3600 QST
w cp,w (Tdc TFP,w ) + w iph + ice cp,ice (TFP,w TST )

AST = 6

 V 2/3
ST

(A.1)

Compressor:
Comp + Z Comp = c7 E 7
c6 E 6 + cw,Comp W

(A.2)

c8 E 8 + Z EX = c5 E 5

The relations used to model vapor compression refrigeration

system (charging cycle) are [32,33]:
Evaporator heat transfer rate:
(A.3)

Refrigerant mass ow rate:

Condenser:
c7 E 7 + c10 E 10 + Z Cond = c8 E 8 + c9 E 9

(B.6)

c9 E 9 + Z CT = c10 E 10

A.X = B

E
(A.4)

Refrigeration cycle compressor power consumption:

(A.5)

(B.7)

The matrices of computing unit cost of exergy values (c) which

are obtained from exergy-cost balance equations as well as the unit
cost of fuel for each component are as follows:

Comp = m
r (h7 h6 )
W

(B.5)

Cooling tower:

A.3. The refrigeration cycle

Q EV
r=
m
h6 h5

(B.4)

Expansion valve:

It should be mentioned that the shape of ice storage tank was

assumed to be cylindrical (with diameter equal to the height) in
order to minimize the heat leakage rate [4].

QST
Q EV =
tch

(B.3)

00

A= 0
0

0
0

(B.8)
(E 4 E 3 )
(E 3 E 4 )
0
0
0
0
0

0
0
(E 6 E 5 )
E 6
E 5
0
0

0
0
0
E 7
0
E 7

0
0
0
0
E 8
E 8

0
0
0
0
0
(E 9 E 10 )
(E 10 E 9 )

0
(E ST,int E Q CW )

(E ST,ch E ST,int )
0
0
0
0

S. Sanaye, A. Shirazi / Energy and Buildings 60 (2013) 100109

Z AHU
Z ST
Z EV

B = cw,Comp WComp + ZComp

Z EX

ZCond
Z CT

cF,AHU = c3

c2
c3
c
5

X=
c7
c8
c
9
cST
(B.9)

cF,ST = cST

(B.10)

cF,EV = c5

(B.11)

cF,Comp = cw,Comp

(B.12)

cF,EX = c8

(B.13)

cF,Cond+CT = c9

(B.14)

References
[1] B.A. Habeebullah, Economic feasibility of thermal energy storage systems,
Energy and Buildings 39 (2007) 355363.
[2] I. Dincer, Thermal energy storage systems and applications in buildings, Energy
and Buildings 34 (2002) 377388.
[3] S. Sanaye, A. Fardad, M. Mostakhdemi, Thermo-economic optimization of an
ice thermal storage system for gas turbine inlet cooling, Energy 36 (2011)
10571067.
[4] D. MacPhee, I. Dincer, Performance assessment of some ice TES systems, International Journal of Thermal Sciences 48 (2009) 22882299.
[5] I. Dincer, M.A. Rosen, Energetic, environmental and economic aspects of thermal energy storage systems for cooling capacity, Applied Thermal Engineering
21 (2001) 11051117.
[6] G.P. Henze, B. Biffar, D. Kohn, M.P. Becker, Optimal design and operation of
a thermal storage system for a chilled water plant serving pharmaceutical
buildings, Energy and Buildings 40 (2008) 10041019.
[7] G.P. Henze, Parametric study of a simplied ice storage model under conventional and optimal control strategies, Journal of Solar Energy Engineering 125
(2003) 212.
[8] P. Ihma, M. Krarti, G.P. Henze, Development of a thermal energy storage model
for Energy Plus, Energy and Buildings 36 (2004) 807814.
[9] H.-J. Chen, D.W.P. Wangs, S.-L. Chen, Optimization of an ice-storage air
conditioning system using dynamic programming method, Applied Thermal
Engineering 25 (2005) 461472.

109

[10] M.A. Ezana, M. Ozdogana, H. Gunerhanb, E. Aytunc, A. Hepbasli, Energetic and

exergetic analysis and assessment of a thermal energy storage (TES) unit for
building applications, Energy and Buildings 42 (2010) 18961901.
[11] R. Domanski, G. Fellah, Thermo-economic analysis of sensible heat, thermal
energy storage systems, Applied Thermal Engineering 18 (1998) 693704.
[12] M.A. Badar, S.M. Zubair, A.A. Al-Farayedhi, Second-law based thermo-economic
optimization of a sensible heat thermal energy storage system, Energy 18
(1993) 641649.
[13] M.M. Hussain, I. Dincer, S.M. Zubair, A feasibility study of using thermal energy
storage in a conventional air-conditioning system, International Journal of
Energy Research 28 (2004) 955967.
[14] ASHRAE Handbook, Fundamentals, Psychrometrics, 2009, pp. 116 (chap. 1).
[15] A. Bejan, G. Tsatsaronis, M. Moran, Thermal Design and Optimization, John
Wiley & Sons, Inc., New York, 1996, pp. 113134, 405459.
[16] T.J. Kotas, The Exergy Method of Thermal Plant Analysis, Krieger Publishing
Company, FL, 1995, pp. 2951.
[17] P. Roosen, S. Uhlenbruck, K. Lucas, Pareto optimization of combined cycle power
system as a decision support tool for trading off investment vs. operating costs,
International Journal of Thermal Sciences 42 (2003) 553560.
[18] G. Wall, Optimization of refrigeration machinery, International Journal of
Refrigeration 14 (1991) 336340.
[19] R. Selbas, O. Kizilka, A. Sencan, Thermo-economic optimization of sub-cooled
and superheated vapor compression refrigeration cycle, Energy 31 (2006)
21082128.
[20] M.H. Panjeshahi, A. Ataei, Application of an environmentally optimum cooling
water system design in water and energy conservation, International Journal
of Environmental Science and Technology 5 (2) (2008) 251262.
[21] http://www.ecocostsvalue.com/httpdocs/content/html/2-emissions.html
[22] J. Wang, Z. Zhai, Y. Jing, C. Zhang, Particle swarm optimization for redundant building cooling heating and power system, Applied Energy 87 (2010)
36683679.
[23] R.J. Dossat, Principles of Refrigeration, 3rd ed., Prentice-Hall, Inc., New Jersey,
1991, pp. 287298.
[24] S. Sanaye, M. Aghaei Meybodi, M. Chahartaghi, Modeling and economic analysis
of gas engine heat pumps for residential and commercial buildings in various
climate regions of Iran, Energy and Buildings 42 (2010) 11291138.
[25] K.K. Humpherys, S. Kattell, Basic Cost Engineering, Marcel Dekker, New York,
1981.
[26] http://www.irimo.ir/English/statistics/synopH/Ahwaz.txt
[27] M.A. Rosen, I. Dincer, N. Pedinelli, Thermodynamic performance of ice thermal
energy storage systems, Journal of Energy Resources Technology 122 (2000)
205211.
[28] http://www.cbi.ir (The ofcial website of Iranian Central Bank).
[29] http://www.mim.gov.ir (The ofcial website of ministry of industries).
[30] I. Dincer, Refrigeration Systems and Applications, John Wiley & Sons, Ltd.,
England, 2003, pp. 109210.
[31] B. Kilkis, S. Kakac, Energy Storage System, vol 167, Kluwer Academic Publishers
Series E: Applied Sciences, 1989, pp. 703720.
[32] R.J. Dossat, Principles of Refrigeration, third ed., Prentice-Hall, New Jersey,
1991, pp. 325389.
[33] ASHRAE Handbook, HVAC Systems and Equipment, Cooling Towers, 2008, pp.
120 (chap. 39).