Energy Conversion and Management 162 (2018) 225–238

Contents lists available at ScienceDirect

Energy Conversion and Management

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

A novel Kalina power-cooling cycle with an ejector absorption refrigeration T

cycle: Thermodynamic modelling and pinch analysis
⁎
Jouan Rashidi, ChangKyoo Yoo
Dept. of Environmental Science and Engineering, College of Engineering, Center for Environmental Studies, Kyung Hee University, Seocheon-dong 1, Giheung-gu, Yongin-Si,
Gyeonggi-Do 446-701, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: A new power and cooling cogeneration cycle is proposed that combines the Kalina power cycle and the ejector
Absorption absorption refrigeration cycle with an ammonia-water mixture as the working fluid. The proposed system, Kalina
Ejector power-cooling with an ejector cycle (KPCE), originates from the Kalina power and cooling cycle (KPCC) and
Kalina cycle introduces an ejector before the evaporator. Thermodynamic analyses from the viewpoints of energy efficiency,
Heat pinch
as well as comparisons between KPCC and KPCE under the same initial conditions, were conducted for the
Power and cooling cogeneration
cycles’ refrigeration output and thermal efficiency. Energy analysis results showed that the KPCE provides a
performance improvement without greatly increasing system complexity. At the same power production level,
the refrigeration output and thermal efficiency of KPCE is 13.5% higher and 17% more than KPCC, respectively.
Energy losses due to inefficient heat recovery design of the system are identified by cross heat pinch analysis. All
three preheaters of the system showed an inefficient design of heat recovery. After redesigning, power, and
power-cooling efficiencies showed 7% and 4.3% increases, respectively. The effect of four important input
parameters including three pressure levels and ammonia mass fraction on the KPCE performance are in-
vestigated to optimize the system. The optimized KPCE performance improved by 17.9% and 13.6% for power
and power-cooling efficiency while the total annual cost of the system could decrease by 6.8%.

1. Introduction of the cycle using the ejector.

Industrial waste heat is a low quality free energy source that can be
recovered and converted into power and/or cooling. Several re-
searchers have investigated the performance of power systems using
low grade heat sources, and most concentrated on Rankine, Kalina, and
other cycles with unconventional working fluids [1,2]. Various ther-
modynamic cycles have been developed to utilize and recover low-
grade heat for cooling purpose, such as absorption, adsorption, de-
siccant, and ejection cycles [3]. Dai et al. proposed a system that
combined the Rankine cycle and the ejector-refrigeration cycle by
adding a turbine between the generator and the ejector [4]. Alexis [5]
performed the first law analysis of a Rankine cycle generating 2 MW
combined with an ejector-refrigeration cycle. Li et al., proposed a Ka-
lina cycle (KC) where an ejector was used as a substitute for the throttle
valve and the absorber [6]. Since the ejector could increase the working
pressure difference, the cycle could obtain a higher power output and
thermal efficiency. Wang et al. proposed a combined power and re-
frigeration cycle where an ejector is placed between the rectifier and
the condenser of the Rankine cycle thereby improving the performance
Low quality and free energy sources such as industrial waste heat
can be recovered and used in both power and cooling production sys-
tems. Several researchers have studied power production performance
using low grade heat sources while concentrating on unconventional
working fluids such as ammonia-water mixture in Rankine, Kalina, or
other cycles [7,8]. Ammonia-water evaporates and condenses as a non-
azeotropic mixture over a range of temperatures, and is therefore able
to achieve a better temperature match between the working substance
and heat sources.
The Kalina cycle KC, introduced by Alexander Kalina, is a fairly new
thermodynamic power cycle using an ammonia-water mixture with the
potential of efficient energy conversion of low-grade heat sources, in-
cluding low-temperature geothermal energy [9], solar energy [10], and
industrial waste heat [11].
Power and cooling cogeneration systems utilizing low grade heat
sources have been studied by various researchers. Rashidi et al. in-
troduced two highly efficient power-cooling cycle integrating Kalina
and absorption refrigeration cycles, KPCC (Kalina power-cooling cycle)
and KLACC (Kalina lithium-bromide absorption cooling cycle), to

⁎
Corresponding author.
E-mail address: ckyoo@khu.ac.kr (C. Yoo).

https://doi.org/10.1016/j.enconman.2018.02.040
Received 9 October 2017; Received in revised form 26 January 2018; Accepted 9 February 2018
0196-8904/ © 2018 Elsevier Ltd. All rights reserved.
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Nomenclature ṁ Mass flow rate (kg/s)

P1 Pump1
Abbreviations Explanation P2 Pump2
ABS Absorber Q Heat (kW)
ACC Annual capital cost S Entropy (kJ/kg K)
C Condenser T Temperature (K)
DES Desorber W Power (kW)
EV Expansion valve X Ammonia mass fraction
EVAP Evaporator Y Exergy ratio
FT Flash tank
FTP Flash tank preheater Indexes
KC Kalina cycle
KPCC Kalina power - cooling cycle e Outlet
KPCE Kalina power-cooling with ejector i Inlet
PH preheater k kth component
SH Superheater L Loss
PA Pinch analysis Out Outlet
TAC Total annual cost p power
Symbol Explanation p,c Power-cooling
fr Mass ratio Q Heat
h Enthalpy (kJ/kg) ref Refrigeration
η Efficiency W Work

produce both outputs simultaneously [12]. They reported a power- To analyze the KPCE the performance of the system is subjected to
cooling efficiency of 18.8% and 20.2% for KPCC and KLACC respec- energy analysis to find the energy losses. KPCE is then compared with
tively. The exergetic efficiency and unit cost of power-cooling genera- the KPCC system to find the effect of ejector on KPCC. Additional losses
tion from exergoeconomic analyses for both KPCC and KLACC systems of energy due to inefficient heat recovery design of the system are
are studied by Rashidi et al. [13]. They concluded that KPCC system can identified through heat pinch analysis. Pinch point and Cross pinch heat
be introduced as the best power-cooling system with 20.5% lower unit exchangers are identified. According to these losses obtained from en-
cost of product and less complexity, but the performance should be ergy and pinch analyses, the process is redesigned to improve the
optimized. With the same purpose, the performance of a combined performance of the equipment and eliminate cross heat transfer. For the
Rankine cycle with an ejector has been investigated [14]. The perfor- final step the effect of initial input parameters on the performance and
mances of different working fluids in a combined organic Rankine cycle total annual cost of KPCE are investigated and subject to the best per-
and ejector refrigeration cycle have been studied by Habibzadeh et al. formance and lowest total cost the system is optimized.
[15].
The Pinch analysis (PA) is a powerful analytical method to identify
2. Materials and methods
and select concrete technical solutions for improving efficiencies and
providing optimum manufacturing solutions [16]. PA with the idea of
2.1. System configuration
setting targets prior to design was first introduced and developed in the
late 1970s and has reported significant changes in energy saving and
Fig. 1a shows a schematic diagram of the KPCC (Kalina power-
several applications in chemical process industries [17]. Researchers
cooling cycle) using an ammonia – water mixture as the working fluid
have performed studies using PA in various power plants to simulate
and a low-temperature heat source [19]. KPCC consists of two loops,
and modify the existing sites [18].
the refrigeration loop and the power loop. The absorber, flash tank, and
According to the literature, performance studies of a combined
condenser play roles in both loops. A portion of the condensed solution
Kalina and ejector cycle as a power-cooling cogeneration cycle are rare.
is sent to the evaporator to generate cooling. Indeed, the cooling system
Some studies have applied the heat integration between Kalina power
operates based on an ammonia-water absorption chiller. The cooling
cycle and ejector absorption refrigeration cycle where power and
load of the system can be controlled by changing the mass flow rate
cooling are generated in two separated sub-cycles. Some other studies
through the evaporator and the evaporator pressure. The cycle works
proposed a combined cycle as a single cycle to produce power and
with three different ammonia concentrations. The primary concentra-
cooling simultaneously with a where the turbine exhaust enters the
tion (XABS) with an intermediate range (47%) at state 1 leaves the ab-
ejector as the primary fluid. The aim of this article is to propose an
sorber and is separated through the flash tank to the highest (XTUR) and
optimized and high efficient combined power and cooling cycle with an
lowest (XLOW) concentration streams (step 4 → 5, 15). The solution with
ejector refrigeration cycle that utilizes the ammonia water solution. In
a concentration of XTUR is divided into two portions to produce power
this study, the Kalina power-cooling cycle with an ejector (KPCE) is
and cooling through the turbine and evaporator, respectively. Cycling
introduced. Unlike most of the studies, the KPCE system is able to
the working fluid heat recovery happens in two heat exchangers (FTP1
produce power and cooling in one single cycle simultaneously. Also it is
and FTP2) before separation in the flash tank, and in another heat ex-
proposed to place the ejector after the condenser of the cycle. The KPCC
changer (PH) before condensation. This system is a flexible cycle,
system which was mentioned before can produce both power and
switching between a power (fr = 0), a power-cooling cogeneration
cooling but the efficiency is not high enough because of heat losses, not
(0 < fr < 1), and a cooling generation cycle (fr = 1).
optimized initial input parameters, and low cooling generation. In order
Fig. 1b represents a schematic of the proposed Kalina power-cooling
to improve the KPCC system, a new cycle (KPCE) is proposed to use the
cycle with an ejector (KPCE) system. The cycle is a combination of KC
ejector in the cycle since the ejector is able to decrease losses while
and ejector absorption refrigeration cycles to simultaneously produce
expansion of the working fluid and recover a portion of working fluid as
power and cooling. The working fluid after pumping to intermediate
the secondary fluid.
pressure (4.7 bar) and heating in two preheaters (FTP1 and FTP2)

226
J. Rashidi, C. Yoo
Fig. 1. Schematic representation of (a) the Kalina power-cooling cycle (KPCC) [12] and (b) the Kalina power and cooling cycle with an ejector (KPCE).
enters the flash tank to be separated into the highest and the lowest
ammonia concentration. The working fluid with the highest ammonia
concentration which is around 73% after giving heat to the other
stream in PH is condensed to be pumped before entering the boiler.
After the condenser, the working fluid (stream 17) enters the ejector to
decrease the pressure drop inefficiencies and increase the cooling sub-
system performance. Two inputs are considered to pass the ejector:
primary flow, which is a portion of the condensed working fluid (state
17), and secondary flow, which is a portion of the evaporated working
fluid in the cooling subsystem. The ejector outlet at state 18 generates
cooling through the evaporator. Following cooling generation, the
working fluid (state 31) is divided into two parts to be recycled in the
system; one enters the ejector as the secondary flow (state 31) and the
other is absorbed by the absorber (state 33). The high pressure super-
heated working fluid after passing the boiler and SH produces power
through the TUR. After recovering the exhaust heat, it is absorbed by
the ABS and recycled through the cycle.
2.2. Thermodynamic modeling
The thermodynamic properties of all states for the base (KPCC) and
227
Energy Conversion and Management 162 (2018) 225–238
proposed (KPCE) systems must be obtained using thermodynamic
modelling. The mass, concentration, and energy conversion balance
equations are presented in this section. A concentration balance equa-
tion should be considered since the ammonia concentration changes
throughout the cycle, especially in the flash tank and the absorber. The
models developed by Rashidi et al. [12], Li et al. [6], and Habibzadeh
et al. [15] are used in this study.
The three pressure levels of the turbine, flash tank, and absorber,
the heat source temperature, the ammonia concentration in the
working fluid, and the ejector geometry in this study are considered as
the input operating data. The values for these parameters were selected
from the studies performed by Rashidi et al. [12] and Ma et al. [20].
The reference values of the turbine, flash tank and absorber pressure
(PTUR, PFT, PABS) are 1500, 470, and 170 kPa, respectively. The highest
concentration of ammonia-water through the cycle is 73% and the heat
source temperature is 280 °C. The surface area of the ejector for the
throat, diffuser, and constant area are 0.12, 1.5, and 0.49 m2, respec-
tively. The input conditions for the KPCC and KPCE systems are shown
in Table 1, including the ammonia mass fractions, mass flow rates,
pressure and temperature levels, and mass ratios of different levels.
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Table 1 ̇ = (wTUR)(ṁ 17)

Win (19)
Reference input data of the KPCC and KPCE systems [12].
wTUR = ηtur (h10−h11) (20)
Parameter Unit Value

Turbine inlet pressure (PTUR) kPa 1500

Intermediate pressure (PFT) kPa 470 2.2.2. KPCE components modelling
Absorber pressure (PABS) kPa 170 Eqs. (1)–(13) are used to obtain the KPCE modelling for the com-
Evaporator pressure (PEVAP) kPa 170
Turbine ammonia mass fraction (XTUR) % 72
ponents that are used in both systems. A schematic diagram of the
Absorber ammonia mass fraction (XABS) % 47 ejector from the KPCE system with constant mixing area is shown in
Flash tank dilute ammonia mass fraction (XLOW) % 14.7 Fig. 2. The primary flow enters the ejector at PFT (section P) and its
Absorber mass flow rate kg/s 4 pressure decreases while the velocity increases by flowing through the
Turbine Inlet temperature (TTUR) °C 280
nozzle. The secondary flow (state 33) is sucked into the ejector, ac-
Absorber Outlet temperature (TABS) °C 20
Evaporator outlet temperature (TEVA) °C 10 celerated, and mixed in the same pressure with primary flow at Section
Mass flowrate ratio (fr) – 0.34 2. Mixing happens between Sections 2 and 3, where the pressure in-
creases due to the constant mixing area. The normal shock accrues at
section (sh) since the pressure increases and the velocity drops to a
2.2.1. KPCC components modelling subsonic level.
The governing equations of the KPCC system are expressed as fol- The Mach number is equal to 1 in sections t and 2 for the primary
lows. and secondary flows, respectively.
The mass balance and concentration balance equations for all The thermodynamic model of the ejector is established based on the
components can be obtained using Eqs. (1) and (2). mass, momentum, and energy balance equations. The simulation is
conducted in different steps, as shown in Fig. 3. In each step, the
∑ ṁ in = ∑ ṁ out (1)
pressure of the section is assumed, and then according to the thermo-
dynamic equation, the assumption is repeated until the last equation of
∑ ṁ in xin = ∑ ṁ out xout (2)
each step is satisfied. The known parameters in this process are the
where, ṁ is the mass flowrate and x is the concentration of ammonia in primary flow properties, as well as the ejector area of different parts. It
the stream. The In and out subscripts refer to the inlet and outlet is assumed that the expansion, suction, and compression processes are
streams, respectively. The energy balance equations for all components non-isentropic. So three isentropic efficiencies for these processes are
are summarized in Table 2. The enthalpy and entropy for NH3eH2O considered, which are, ηn = 0.9, ηs = 0.85, and ηd = 0.85 related to the
have been evaluated by the correlations presented in Appendix A expansion, suction, and compression, respectively. A mixing loss factor
[12,21]. (ηm = 0.85) also is considered, which accounts for the inefficiency in
As can be seen, Eq. (3) determines the absorber heat (Qabs) which is momentum transfer between the primary and secondary flows due to
equal to the difference of inlet and outlet enthalpies. The power con- friction between flows and between the fluid and ejector wall. The
sumption of the pump is obtained from Eq. (4). The pump efficiency (ηp) values of the mentioned efficiencies are selected from studies per-
and turbine efficiency (ηtur) are considered to be equal to 0.7 and 0.85 formed by Ma et al., and Wang et al. [20,22].
[22]. Eqs. (5) and (6) are used to calculate the enthalpy and tempera-
tures of heat exchanger outlet streams. In Eq. (6) ε refers to the heat 2.3. Economic modeling
exchanger effectiveness. These equations are applied for all heat ex-
changers of the KPCE system. In order to estimate the total annual cost, TAC, the economic model
The first law efficiency is defined by dividing the output of the cycle was developed using models presented by Campos Rodríguez et al.
by the required input. It is given by Eqs. (14) and (15) [23]. [24], Janghorban Esfahani et al. [25], Rashidi et al. [12], and Jiang
̇ + Qeva
Wnet ̇ et al. [26].
ηp,c = The total annual cost is given by Eq. (21) [25]
̇
Qin (14)
TAC = ACC + AOC (21)
̇
Wnet
ηPower =
Qiṅ (15) where ACC and AOC are the annual capital cost and annual operating
cost. The equations to find ACC of the system are presented below,
where, ηpower and ηp,c are power and power-cooling efficiencies when while AOC is explained in Appendix B.
considering only power production and power-cooling cogeneration, To find the ACC, the total cost investment, TCI, is multiplied with
respectively. Wnet is the net produced power of the cycle and Qin is the the amortization factor given in Eq. (22) [12].
total consumed heat of the cycle which are obtained as follows:
̇ = WTUR
Wnet ̇ −ẆP1−ẆP 2 (16) Table 2
Energy balance equations for components of the KPCC system [12].
and
Component Energy balance equation
̇ = Qboiler
Qin ̇ ̇ + Q̇FT
+ Qsh (17)
Absorber ̇ = ṁ 12 h12 + ṁ 17 h17−ṁ 1 h1
Qabs (3)
That means total consumed heat of the cycle is equal to sum of Solution pumps (4)
Ẇ pump = ṁ in vin (Pout −Pin)/ ηpump
boiler, superheater (sh), and flash tank (FT) required heat.
Preheaters Qph = ṁ hot (hin,hot −hout ,hot ) = ṁ cold (hout ,cold−hin,cold ) (5)
Further, the coefficient of performance can be obtained by dividing
ε = Qph/ cp,min (Tin,hot −Tin,cold) (6)
the cooling output by the required power. Here, the required power is Flash tank (8)
Q̇FT = ṁ 5 h5 + ṁ 15 h15−ṁ 4 h4
determined by the quantity of decreased power from only the power Condenser (9)
QĊ ,K = ṁ 6 (h6−h13)
system to the power-cooling system. Boiler ̇ (10)
Qboiler = ṁ 8 (h9−h8)
Q̇ Superheater QSḣ = ṁ 9 (h10−h9) (11)
COPkpcc = eva Turbine ̇ = ηtur ṁ 10 (h11−h10) (12)
̇
Win (18) Wtur
Evaporator ̇ = ṁ 18 (hin−hout )
Qeva (13)
where

228
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Fig. 2. Schematic representation of the ejector.

Fig. 3. The simulation modeling process of the ejector.

Table 3
Pressure and Mach number in the ejector sections.

Section Mach number Pressure Section Mach number Pressure

Primary flow Secondary flow Primary flow Secondary flow Mixed flow Mixed flow

P 0.2 2 3 1.16 0.38

S 0.99 0.2 0.71 Sh 1.16 0.50
1 1.19 0.79 4 0.28 0.79
2 1.44 1.001322 0.54 0.38 d 0.28 0.79

229
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Fig. 4. Pressure and velocity profiles of the ejector.

absorber, desorber, boiler and 4superheater, which are assumed to be a

shell and tube heat exchanger. The cost of these exchangers is calcu-
lated according to their heat transfer area (A) given by Eq. (24) [28].
Qk
Ak =
Uk ·LMTDk (24)

where Qk is the heat transfer flow rate through component k, U is the

heat transfer coefficient, and LMTD is the logarithmic mean tempera-
ture difference. U values for different components, given in units of kW/
m2 K, are shown in Table B2.
The logarithmic temperature difference LMTD for the heat ex-
changers is obtained by Eq. (25) where subscripts h, c, i, and e refer to
hot, cold, inlet and outlet streams.
(Th,i−Tc,i )−(Th,e−Tc,e )
LMTDk = T −T
ln T h,i − Tc,i (25)
h,e c,e

Having the heat transfer area, the cost of all mentioned heat ex-
changers can be obtained using the power law relation given by Eq.
(26) [29].
0.6
A
Zh = ZR . h·⎛ h ⎞
⎜ ⎟

⎝ AR ⎠ (26)

Fig. 5. Comparison of (a) the power and power-cooling, efficiencies and (b) the power where Zh is the capital cost of the heat exchanger with a heat transfer
and cooling outputs of KPCC and KPCE. area of Ah in a reference year. ZR,h is the reference cost of the heat
exchanger in the reference year which has a reference heat transfer area
i·(1 + i)n of AR. The reference costs of the heat exchangers for the reference year
CRF =
(1 + i)n−1 (22) 2000 are listed in Table B3 [30].
The capital cost of the pump is calculated using Eq. (27) [31].
where CRF is the amortization factor, i and n are the interest rate and
0.8
plant life cycle which are assumed to be 15% and 20 years, respectively Cpump = 1120·Ẇ pump (27)
[27].
Total cost investment (TCI) is calculated by Eq. (22), which is a where Ẇ pump is required pump power.
summation of the direct cost (DC), including onsite and offsite costs, ̇ )0.7
Ctur = 4405 × (Wtur (28)
indirect costs (IDC), and other outlays (OO).
̇ is the
Eq. (28) gives the capital cost of the turbine [32]. where Wtur
TCI = DC + IDC + OO (23)
output power of the turbine. Using Eq. (12) and chemical engineering
Eqs. (B-1)-(B-16) describe the TCI calculation that are summarized in plant cost (CEPCI), all of the obtained costs were updated to represent
Table B1 including the cost of pipes, electrical controls, the monitoring equivalent costs for the desired year of 2016 [30].
system, civil structural and architectural work, service facilities, con-
CI
tingencies, and other outlays. Zk,2013 = Zk ⎛ 2016 ⎞
⎜ ⎟

⎝ CI2000 ⎠ (29)

2.3.1. Equipment cost where Zk,2016 is the component cost in year 2016, Zk is the component
This section explains how equipment costs are defined. Heat ex- cost in year 2000 and CI2016, and CI2000 are equipment cost indexes in
changers including the condensers, evaporators, preheaters, flash tank, year 2016 and 2000, respectively. CI2000 and CI2016 are equal to 394.1

230
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Fig. 6. (a) The total cost of equipment (b) annual capital, operating, and total cost of KPCE and KPCC systems.

Table 4 process simulation demonstrated in Fig. 3 have been summarized in

Stream data of the KPCE system. Table 3. The pressure and velocity profiles of ejector flows are shown in
Fig. 4. As can be seen after mixing, a normal shock takes place and the
Stream No. Type Ts (K) Tg (K) ΔH (kW) m.cp (kW/K)
velocity drops significantly. The pressure and mass flow rate of sec-
1: H1 (12:13) Hot 337.1 330 −343.35 48.35915 ondary flow besides the properties of ejector outlet, were obtained after
2: H2 (14:15) Hot 382.8 331 −577.3058 11.1449 modeling and are presented in Table 3.
3: C2 (8:11) Cold 380.1 433.1 1279.528 24.14203
Using the thermodynamic properties of the working fluid and the
4: H3 (5:7) Hot 382.8 293 −3943.253 43.9115
5: C3 (18:31) Cold 264.6 283.1 468.704 25.33535 energetic analysis presented in Eqs. (1)–(19), the energetic performance
6: H4 (34:35) Hot 289.3 274.6 −468.6905 31.88371 of the KPCE system was obtained. As can be seen in Fig. 5, the power
7: H5 (63:1) Hot 325.4 288.1 −3246.555 87.03901 efficiency of KPCE with 13% shows no change when compared with the
8: C1 (2:4) Cold 321.5 321.5 3580.578 82.61602 KPCC system, while the power-cooling efficiency (25%) was increased
due to the 13.5% greater cooling output of KPCE, which is a result of
using the ejector. This observation means that KPCE uses the potential
and 568.8, respectively [33].
of power and cooling generation more effectively than the KPCC
system. This increase of cooling output happens for two reasons; less
3. Results and discussion pressure drop losses occurs in the ejector when compared with expan-
sion valves, and the mass flowrate of ejector outlet entering the eva-
3.1. Energy and economic analyses porator is more than mass flow rate of expansion valve outlet since a
portion (22%) of evaporated working fluid is recovered and entered the
Energy and economic analyses were conducted to determine the ejector as the secondary fluid. therefore, the mass flow rate of eva-
efficiencies including power and power-cooling efficiencies, coefficient porator of KPCE is 1.22 times more than that of KPCC system which
of performance, generated cooling, and total annual cost in the KPCE results in higher enthalpy difference in the evaporator and higher
system and compare these values with the KPCC system to introduce cooling generation.
the best power-cooling Kalina cycle. The ejector modeling results using Fig. 6 summarizes the economic analysis results for both KPCE and

231
J. Rashidi, C. Yoo
Fig. 7. Composite curve of the heat exchanger network: (a) without utilities, (b) with utilities, (c) grand composite curve, and (d) available temperature difference of the heat exchanger
network in the KPCE system.
KPCC systems in $/yr. The total annual cost of KPCE was obtained
equal to $8.4∗106/yr while it was 9.3∗106 for the KPCC system. Both
annual operating and capital cost for KPCE were less than KPCE system,
by comparing these values it can be resulted the KPCE is more economic
than the KPCC system. The reason is related to the lower absorber,
boiler and condenser cost of KPCE since the need for cooling water in
these two components were decreased. As can be seen the cost for other
equipment of KPCE are almost equal to those of KPCC system.
3.2. Heat pinch analysis
To begin the pinch analysis, the stream data shown in Table 4 was
extracted for the process and the composite curves of the process could
be calculated as presented in Fig. 7. As shown in Table 4, the heat
exchanger network initially consists of five hot and three cold streams.
It summarizes the source (Ts) and target temperature (Tg) of heat ex-
changer network streams, as well as their corresponding enthalpy dif-
ference (ΔH). In the last column, m.cp is the ratio of ΔH to ΔT. To
232
Energy Conversion and Management 162 (2018) 225–238
continue the pinch analysis, the minimum temperature difference ΔTmin
was set to 5 K in order to determine the heat recovery targets. Fig. 7a
shows the composite curve of hot and cold streams, while Fig. 7b de-
monstrates the composite curve with hot utility. Fig. 7c and d demon-
strates the grand composite curve and available temperature difference.
Initial results show that the required heating and cooling duties with
the minimum temperature difference of 5 K are 1464.99 and
4718.64 kW, respectively. The KPCE system with current heat ex-
changers and heat recoveries required 4201 kW of hot and 7594 kW of
cold utility, which are much higher than calculated amounts from pinch
analysis. The minimum number of heat exchangers required to achieve
the available heat recovery is 12. The pinch point temperature was
calculated to be 334.6 K, while a hot utility with an initial temperature
of 445 K was considered to fulfil the required heat.
The system before modification had several pinch problems where
the cold stream crossed the hot stream at three points.
Fig. 8 shows the grid diagram of heat exchanger network. As can be
seen based on the pinch analysis, 9 heat exchangers including hot and
J. Rashidi, C. Yoo
Fig. 8. Grid diagram of the heat exchanger network in the KPCE system.
cold utilities, condenser, absorber and flash tank to improve the heat
recovery potential for the current process were detected. In the base
system heat is recovered through 3 internal heat exchangers; FTP1,
FTP2, and PH, while after pinch analysis 3 more internal heat ex-
changer are recommended which resulted in 2876 and 2736 kW de-
crease in hot and cold utilities, respectively. In the KPCE system before
pinch analysis all three internal heat exchangers crossed the pinch
point. To solve this problem the stream number 2 after pumping is
divided in three streams before pinch point to be used as a cooling
stream of flash tank outlet (state 14) and turbine exhaust (state 13).
This stream after the pinch point can be heated in two separate paths
using both flash tank outlets (states 5 and 14) through 3 heat ex-
changers.
Since the hot and cold utilities of the KPCE after heat pinch analysis
has been decreased significantly, the performance of the system has
been improved which shows 30% of thermal efficiency and 59% of
power-cooling efficiency.
The enthalpy temperature profile of heat exchangers clarifies how
hot and cold streams release and receive heat while their temperature
may decrease and increase, respectively. For a proper model of heat
exchanger hot and cold streams shouldn’t cross each other. Moreover
the hot stream profile has to be located upper than cold stream profile
that means the temperature of hot stream has to be higher than cold
stream for all points of the heat exchanger. Fig. 9 demonstrates the
enthalpy-temperature profile of all 9 heat exchangers including hot and
cold streams. As can be seen all heat exchangers are working properly
with no cross point and enough temperature difference between hot
and cold streams while before doing pinch analysis there were cross
points in some of them.
233
Energy Conversion and Management 162 (2018) 225–238
3.3. Parametric analysis
Parametric analysis was performed to evaluate the effects of key
parameters on the energy efficiency, cooling generation, and TAC of the
KPCE system. Key parameters include the ammonia mass fraction of the
working fluid (XWF), absorber pressure (PABS), turbine inlet pressure
(Ptur), and flash tank pressure (PFT). The variables change ranges are
summarized in Table 5.
In the parametric analysis, one parameter was varied while the
others were kept constant. Table 6 summarizes the effect of four men-
tioned parameters on three variables including power efficiency,
power-cooling efficiency, and total annual cost. In the last column re-
gression equations for each variable is presented. As can be seen the
most effective parameter on ηp,c was ammonia-water concentration,
that 1% increase in Xwf resulted in 8.9% decrease of ηp,c while the
lowest effective parameter on ηp,c was Ptur with an indirect relation of
0.14. It means the system has a better power-cooling efficiency when
Xwf has the lowest value in the defined range (0.73) but turbine inlet
pressure variation through the defined range is not highly effective on
the cogeneration efficiency.
More detailed results of parametric analysis are shown in Fig. 10. As
can be seen in Fig. 10a and b power and power-cooling efficiency
reached a peak when Pabs is 122.6 kPa while total annual cost was
fluctuating and decreased slightly for higher absorber pressure. Fig. 10c
and d shows that both efficiencies remained almost unchanged with a
slight up and down but, the total annual cost decreased for Ptur increase.
The next two plots have a peak when Pft is equal to 442 kPa, that means
the system imposed a higher cost for a better performance at this point.
Fig. 10g and h shows how system efficiencies and total cost change
when Xwf increases. It can be seen, although power-cooling efficiency
decreased suddenly the TAC sustain a gradual decrease.
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Fig. 9. Enthalpy profiles of heat exchangers 1 through 9.

Table 5 According to parametric analysis the system has been optimized

Range of effective parameters on analyzed systems. subject to the performance and total annual cost. The optimized power
efficiency, power-cooling efficiency, and TAC as well as their corre-
Parameter Range Unit
sponding parameters values in comparison to their initial values are
PABS 80–350 kPa summarized in Table 7. The parametric analysis and optimization of the
PTur 800–2000 kPa KPCE could increase ηp and ηp,c by 10.2% and 8.9%, and decrease TAC
PFT 400–480 kPa by 10.3%. optimized values for input parameters was 122.6 kPa,
XWF 0.73–0.8 –
1000 kPa, 442.1 kPa, and 0.73 for absorber pressure, turbine inlet

Table 6
Influence of effective parameters on the KPCE system.

Variable Parameter Change of variables Polynomial regression equation

Power efficiency Pabs 1.9% ηp = −3*10−6 (Pabs)2 + 0.0012 Pabs + 0.0269

PTur −0.31% ηp = −2*10−8 (Ptur)2 + 3*10−5 Ptur + 0.107
PFT −0.52% ηp = −1*10−5 (PFT)2 + 0.0124 PFT −2.442
XWF −0.7% ηp = 0.846(Xwf)2 –1.6352 Xwf + 0.8591
Power –cooling efficiency Pabs 1.04% ηp,c = −3*10−6 (Pabs)2 + 0.0012 Pabs + 0.1591
PTur −0.14% ηp,c = −2*10−8 (Ptur)2 + 3*10−5 Ptur + 0.2394
PFT 0.70% ηp,c = −1*10−5 (PFT)2 + 0.0121 PFT − 2.4074
XWF −8.9% ηp,c = 8.6517(Xwf)2 –15.17 Xwf + 6.7117
Total annual cost Pabs −0.36% TAC = −4.819(Pabs)2 + 1529.9 Pabs + 850,806
PTur −0.88% TAC = 0.4531(Ptur)2 –1557.7 Ptur + 2*106
PFT −2.3% TAC = −95.175(PFT)2 + 79,798 PFT − 2*107
XWF −0.45% TAC = 8*106 (Xwf)2 – 107Xwf + 6*106

234
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Fig. 10. KPCE energy and cost variations with respect to the effects of Pabs on (a) efficiency, (b) ACC and TAC, Ptur on (d) efficiency and (e) ACC and TAC, PFT on (f) efficiency and (g) ACC
and TAC, Xwf on (h) efficiency and (i) ACC and TAC.

pressure, flash tanks pressure, and ammonia-water concentration. The 4. Conclusion

total improvement of the KPCE system are presented in Table 7 which
shows almost 18% of power efficiency and 6.8% of total annual cost Kalina power-cooling with an ejector cycle (KPCE) is proposed as a
improvement. new power and cooling system which is working with a mixture of
ammonia-water mixture as the working fluid. It is a combination of the

235
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

Table 7
Comparison between effective parameters and performance of KPCE before and after heat pinch analysis and optimization.

Pabs (kPa) Ptur (kPa) PFT (kPa) Xwf ηp ηp,c TAC ($/yr)

Initial values 170 1500 470 0.72 12.9 25 8.4*106

After heat pinch 170 1500 470 0.72 13.8 26.07 8.7*106
Optimized values 122.6 1000 442.1 0.73 15.21 28.41 7.8*106
Total improvement 17.9% 13.6% −6.8%

Kalina power and cooling cycle (KPCC) ejector absorption refrigeration
cycle. KPCE has been analyzed thermodynamically and compared with
KPCC the same initial conditions for the cycles’ refrigeration output and
thermal efficiency. Energy losses due to inefficient heat recovery design
of the system are identified by cross pinch heat analysis where the
optimum heat recovery structure of the system is obtained. Parametric
analysis has been done to find the effect of initial input parameters on
the system. The system is optimized based on parametric analysis to
find the highest performance with lowest total cost. The most important
results are summarized as follows:
– Thermodynamic analysis results showed that the KPCE is able to
achieve 25% of power-cooling efficiency, while thermal efficiency is
around 13%.
– At the same power production level, the refrigeration output and
thermal efficiency of KPCE is 13.5% higher and 17% more than
KPCC, respectively.
– It is evident that the ejector can improve the performance of the
combined power and cooling cycle without greatly increasing
system complexity.
– Through heat pinch analysis three internal heat exchangers crossed
pinch point so 3 more heat exchanger has been added to the cycle.
– Improving heat exchanger network resulted a significant perfor-
mance improvement.
– After redesigning, the improved KPCE showed 7% and 4.3% in-
creases in thermal and power-cooling efficiencies, respectively.
– The ammonia mass flowrate was the most effective parameter on the
power-cooling efficiency while turbine inlet pressure had the lowest
effect on the system performance. Higher values of all three pressure
levels resulted in worst efficiency and less total cost.
– After parametric analysis, the optimized KPCE showed 18%, 13.6%,
and 6.8% improve in power, power-cooling, and total annual cost
when compared to the initial proposed KPCE.
Acknowledgements
This work was supported by the National Research Foundation of
Korea (NRF) grant funded by the Korea government (MSIT) (NRF-
2017R1E1A1A03070713).

Appendix A

A1. Ammonia/water thermodynamic properties calculations

For a pure component, one needs pressure and temperature to specify the state of the substance. However, for a binary mixture, a third
independent variable is needed to specify the equilibrium state.
The Peng–Robinson equation of the state is [21]:
RT a (T )
p= − ,
(v−b) v (v + b) + b (v−b) (A.3)
where the constants in terms of critical data are:
R2Tc2
a (Tc ) = 0.45724
Pc (A.4)
RT
b (Tc ) = 0.07780 c .
Pc (A.5)
At any other temperature, the constants are calculated by:
a (T ) = a (Tc ) α (Tr ,ω) (A.6)

b (T ) = b (Tc ), (A.7)

where α = {1 + m (1−Tr0.5)}0.5 (A.8)

m = 0.37464 + 1.54226ω−0.26992ω2. (A.9)

The exentric factors (ω) for ammonia and water have been taken from Ohe [34]. To determine the mixture properties, an empirically determined
binary mixture interaction constant kij characterizing the mixture formed by components i and j is included in the mixing rule for calculation of the
constant a.
The mixing parameters were defined by the following mixing rules:
N N
a= ∑∑ x i x j aij
i=1 j=1 (A.10)
N
b= ∑ x i bi,
i=1 (A.11)

236
J. Rashidi, C. Yoo Energy Conversion and Management 162 (2018) 225–238

where aij = (1−kij ) ai0.5 aj0.5. (A.12)

Eq. (A.3) has been shown to be a cubic polynomial in the compressibility factor (Z):
Z 3−(1−B ) Z 2 + (A−3B2−2B ) Z −(AB−B2−B3) = 0, (A.13)

ap bp pv
where A = ,B = ,Z = .
R2T 2 RT 2 RT (A.14)
In the two-phase region, the largest root of the equation is the compressibility factor of the vapor phase, whereas the smallest positive root
corresponds to the compressibility factor in the liquid phase, and the third one has no physical meaning. The fugacity coefficient (ϕk) of component k
in a mixture can be calculated by:

bk A ⎡ 2 ∑ x i aik bk ⎤ Z + 2.414B
ln (φk ) = (Z −1)−ln (Z −B )− − ln ⎡ ⎤
b 2 2B ⎢
⎣ a b⎥⎦ ⎣ Z −0.414B ⎦ (A.15)

Appendix B

B1. Capital cost

Heat transfer coefficient (U values) for different heat exchangers of the cycle are summarized in Table B2 which are used to find the area of heat
exchangers and their capital cost.
The reference capital cost of each component in year 2000 are presented in Table B3. These values are used in Eq. (26) to find the equipment cost
in 2016.

Table B1
Equations for capital cost estimation [13].

Equations Description

DC = ONSC + OFSC ONSC is the onsite cost and OFSC is the offsite cost (B-1)
ONSC = Ceq + Cpipe + CEC Ceq, Cp and CEC are the cost of equipment, piping and electric control monitor system, respectively (B-2)
Ceq = Chx + Ctur + Cpump Chx, Ctur and Cpump are the cost of heat exchangers, turbine and pumps, respectively (B-3)
Cpipe = 0.05 × (Ceq) (B-4)
CEC = 0.3 × (Ceq + Cpipe ) (B-5)
OFSC = Cstr + Cser + Ccon Cstr, Cser and Ccon are the cost of the civil structural and architectural work, service facilities, and contingencies, respectively (B-6)
Cstr = 0.2 × (ONSC ) (B-7)
Cser = 0.25 × (ONSC ) (B-8)
Ccon = 0.15 × (ONSC ) (B-9)
IDC = Ceng + Cconst + Cicon Ceng, Cconst and Cicon are the cost of engineering and supervision, construction including contractor’s profit, and indirect contingencies, (B-10)
respectively
Ceng = 0.15 × (DC ) (B-11)
Cconst = 0.15 × (DC ) (B-12)
Cicon = 0.2 × (DC ) (B-13)
OO = Cstart + CW Cstart and CW are the cost of startup and working capital, respectively (B-14)
Cstart = 0.06 × (DC + IDC ) (B-15)
CW = 0.05 × (DC + IDC ) (B-16)

Table B2
Heat transfer coefficient of plant components [24,35].

Component U (W/m2 K) [35] U (kW/m2 K) [24]

Solution heat exchanger 1000 Thermal compr:1.8

NH3superheat:0.2
Evaporator 900 1.2
Condenser 1100 1.1
Refrigerant (condensing – water) 650 –
Refrigerant (evaporating– water) 510 –
Absorber – 0.785

Table B.3
The reference costs of components (AR=100 m2) [30].

Component Reference cost (2000) $

Desorber 17,500
Absorber 16,500
Solution heat exchanger 12,500
Evaporator 16,000
Condenser 8000
Expansion valve 300

237
J. Rashidi, C. Yoo
B2. Operation and maintenance cost
Operation costs are assumed to include the steam, electricity, and labor costs. The steam cost due to heating energy is the main part of the
operating cost and depends significantly on steam conditions, source, and the cost evaluation method. Steam cost is evaluated as the cost of the
electricity that can be produced in a steam turbine, the cost of the saturated steam at pressures of 0.25 MPa and 0.5 MPa are $9.6/ton and $11.4/ton,
respectively, where electricity cost is $0.07/kWh. In the current study, using the same method, the cost of steam is estimated. cost of electricity,
cooling water labor and maintenance as well as plant operation factor and CI are obtained from previous studies which is done by Rashidi and Yoo
[36] The annual operating cost, AOC, is obtained using the following unweighted linear equation.
AOC = Csteam + Ccw + Celec + CL + CM
where Csteam is the steam cost Ccw is the cooling water cost, Celec is the electricity cost, CL is labor cost, and CM is the maintenance cost.
References
[1] Gu W, Weng Y, Wang Y, Zheng B. Theoretical and experimental investigation of an
organic rankine cycle for a waste heat recovery system. Proc Inst Mech Engineers,
Part A: J Power Energy 2009;223:523–33.
[2] Chen H, Goswami DY, Stefanakos EK. A review of thermodynamic cycles and
working fluids for the conversion of low-grade heat. Renew Sustain Energy Rev
2010;14:3059–67.
[3] Aridhi E, Abbes M, Mami A. Solutions based on renewable energy and technology to
improve the performance of refrigeration systems. J Renew Sustain Energy
2016;8:065906.
[4] Dai Y, Wang J, Gao L. Exergy analysis, parametric analysis and optimization for a
novel combined power and ejector refrigeration cycle. Appl Therm Eng
2009;29:1983–90.
[5] Alexis G. Performance parameters for the design of a combined refrigeration and
electrical power cogeneration system. Int J Refrigeration 2007;30:1097–103.
[6] Li XG, Zhang QL, Li XJ. A kalina cycle with ejector. Energy 2013;54:212–9.
[7] Nemati A, Nami H, Ranjbar F, Yari M. A comparative thermodynamic analysis of orc
and kalina cycles for waste heat recovery: A case study for cgam cogeneration
system. Case Stud Therm Eng 2017;9:1–13.
[8] Eller T, Heberle F, Brüggemann D. Second law analysis of novel working fluid pairs
for waste heat recovery by the kalina cycle. Energy 2017;119:188–98.
[9] Saffari H, Sadeghi S, Khoshzat M, Mehregan P. Thermodynamic analysis and opti-
mization of a geothermal kalina cycle system using artificial bee colony algorithm.
Renew Energy 2016;89:154–67.
[10] Sun F, Ikegami Y, Arima H, Zhou W. Performance analysis of the low temperature
solar-boosted power generation system—part ii: Thermodynamic characteristics of
the kalina solar system. J Sol Energy Eng 2013;135:011007.
[11] Zare V, Mahmoudi S, Yari M. On the exergoeconomic assessment of employing
kalina cycle for gt-mhr waste heat utilization. Energy Convers Manage
2015;90:364–74.
[12] Rashidi J, Ifaei P, Esfahani IJ, Ataei A, Yoo CK. Thermodynamic and economic
studies of two new high efficient power-cooling cogeneration systems based on
kalina and absorption refrigeration cycles. Energy Convers Manage
2016;127:170–86.
[13] Rashidi J, Yoo C. Exergetic and exergoeconomic studies of two highly efficient
power-cooling cogeneration systems based on the kalina and absorption refrigera-
tion cycles. Appl Therm Eng 2017.
[14] Zhang K, Chen X, Markides CN, Yang Y, Shen S. Evaluation of ejector performance
for an organic rankine cycle combined power and cooling system. Appl Energy
2016;184:404–12.
[15] Habibzadeh A, Rashidi M, Galanis N. Analysis of a combined power and ejector-
refrigeration cycle using low temperature heat. Energy Convers Manage
2013;65:381–91.
[16] Ataei A, Yoo C. Combined pinch and exergy analysis for energy efficiency optimi-
zation in a steam power plant. Int J Phys Sci 2010;5:1110–23.
[17] Smith R. Chemical process: Design and integration. John Wiley & Sons; 2005.
[18] Linnhoff B, Alanis FJ. Integration of a new process into an existing site: A case study
in the application of pinch technology. J Eng Gas Turbines Power
238
Energy Conversion and Management 162 (2018) 225–238
(B.17)
1991;113:159–69.
[19] Janghorban Esfahani I, Rashidi J, Ifaei P, Yoo C. Efficient thermal desalination
technologies with renewable energy systems: A state-of-the-art review. Korean J
Chem Eng 2016;33:351–87.
[20] Ma Z, Bao H, Roskilly AP. Thermodynamic modelling and parameter determination
of ejector for ejection refrigeration systems. Int J Refrigeration 2016.
[21] Nag P, Gupta A. Exergy analysis of the kalina cycle. Appl Therm Eng
1998;18:427–39.
[22] Wang J, Dai Y, Zhang T, Ma S. Parametric analysis for a new combined power and
ejector–absorption refrigeration cycle. Energy 2009;34:1587–93.
[23] Hasan AA, Goswami DY, Vijayaraghavan S. First and second law analysis of a new
power and refrigeration thermodynamic cycle using a solar heat source. Sol Energy
2002;73:385–93.
[24] Rodríguez CEC, Palacio JCE, Sotomonte C, Leme M, Venturini O, Lora E, Cobas V,
dos Santos D, Dotto F, Giallucca V. Exergetic and economic analysis of kalina cycle
for low temperature geothermal sources in Brazil. In: The 25th international con-
ference on efficiency, cost, optimization, simulation and environmental impact of
energy, ECOS; 2012. p. 167–79.
[25] Esfahani IJ, Lee S, Yoo C. Evaluation and optimization of a multi-effect evapor-
ation–absorption heat pump desalination based conventional and advanced exergy
and exergoeconomic analyses. Desalination 2015;359:92–107.
[26] Jiang L, Gu Z, Feng X, Li Y. Thermo-economical analysis between new absorptio-
n–ejector hybrid refrigeration system and small double-effect absorption system.
Appl Therm Eng 2002;22:1027–36.
[27] Esfahani IJ, Yoo C. Exergy analysis and parametric optimization of three power and
fresh water cogeneration systems using refrigeration chillers. Energy
2013;59:340–55.
[28] Price S. A consideration of cycle selection for meso-scale distributed solar-thermal
power. Georgia Institute of Technology; 2009.
[29] Rogdakis E. Thermodynamic analysis, parametric study and optimum operation of
the kalina cycle, Fuel and Energy Abstracts, Elsevier; 1996. p. 234.
[30] Wang J, Wang J, Zhao P, Dai Y. Thermodynamic analysis of a new combined
cooling and power system using ammonia–water mixture. Energy Convers Manage
2016;117:335–42.
[31] Kaita Y. Thermodynamic properties of lithium bromide–water solutions at high
temperatures. Int J Refrigeration 2001;24:374–90.
[32] Ibrahim O. Thermodynamic properties of ammonia-water mixtures, ASHRAE
Transactions: Symposia; 1993. p. 1495.
[33] Watanabe K, Dooley R. The international association for the properties of water and
steam, release on the iapws formulation 1995 for the thermodynamic properties of
ordinary water substance for general and scientific use, Fredericia: Denmark; 1996.
[34] Ohe S. Vapour–liquid equilibrium data at high pressure, Physical Sciences Data 42,
Elsevier, Kodansha; 1990. p. 311–4.
[35] Goswami D. Solar thermal power: Status of technologies and opportunities for re-
search. Heat Mass Transfer 1995;95:57–60.
[36] Wang Y, Lior N. Thermoeconomic analysis of a low-temperature multi-effect
thermal desalination system coupled with an absorption heat pump. Energy
2011;36:3878–87.