Energy 147 (2018) 36e50

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Thermodynamic efficiency improvement of combined cycle power

plant's bottom cycle based on organic working fluids
Stjepko Katuli 
c*, Mislav Cehil, Daniel Rolph Schneider
Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Ivana Lucica 5, 10000 Zagreb, Croatia

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

Article history: This paper presents thermodynamic optimization of a bottom cycle which uses water and organic fluids
Received 10 June 2017 as working fluids. The heat recovery steam generator is modeled with regard to different pressure levels,
Received in revised form including the reheat of first pressure level, the configurations of heat-exchangers network where heat
4 January 2018
can be exchanged between flue gas and working fluids. Water is chosen as the working fluid of the first
Accepted 4 January 2018
pressure level while an organic fluid for the second pressure level. Thermodynamic optimization of
efficiency of the bottom cycle was conducted considering the variables of the heat-exchangers' inside the
heat recovery steam generator HRSG and the operating parameters of working fluid of each pressure
Keywords:
Heat-exchangers network
level and reheat. A genetic algorithm and a gradient optimization method were used with the ther-
Heat recovery steam generator modynamic model implanted in Matlab. It is shown that by using parallel and serial configurations of
Organic cycle heat-exchangers and water in the first level pressure and organic working fluids in the second pressure
Steam cycle level a better thermodynamic efficiency of the bottom cycle can be achieved.
Thermodynamic optimization © 2018 Elsevier Ltd. All rights reserved.

1. Introduction
The consumption of fossil energy sources is still far greater than
the energy consumption from renewable energy sources. As an
example, the total consumption of primary energy sources in the
United States [1] is showing that oil, natural gas and coal are the
most used primary energy sources, and the average annual increase
in oil consumption is about 1.5%, in coal 3.24% and in natural gas
more than 4% in the period from 2000 to 2011. The generation of
electrical energy from the fossil energy sources is also far greater
compared to generation from renewable energy sources. The pro-
jections of consumption of primary energy sources for electricity
generation in the United States [2] show their increasing tendency,
and natural gas will be the most used for electricity generation with
the total share of around 35% by the 2040. These data indicate that
the generation of electricity using combined cycle power plants
(CCPP) will continue to increase in the future.
The need to improve the thermodynamic efficiency of the CCPP
is emerging as one of the measures to increase energy efficiency
and reduce greenhouse gas emissions. Generally speaking, waste
heat recovery and utilization represents opportunity to reduce total
* Corresponding author.
E-mail address: stjepko.katulic@fsb.hr (S. Katuli
c).
energy use and to decrease CO2 emissions [3]. Particularly inter-
esting is the increase of the CCPP thermodynamic efficiency (hCCPP )
by proper arrangement of heat-exchangers' layout in HRSG, proper
selection of number of pressure levels, adequate selection of
working fluids and operating parameters in order to increase
thermodynamic efficiency of steam turbine part of the power plant
(hST ). By finding optimum values of previously mentioned vari-
ables, for the same amount of generated electricity less fuel would
be consumed and therefore less greenhouse gases would be pro-
duced. According to the scientific literature review, research of
increasing hCCPP can be roughly divided into three categories:
a) research of increasing the thermodynamic efficiency of both the
top cycle (hGT ) and bottom cycle (hST ) [4],
b) research of increasing hCCPP by simultaneously increasing the
thermodynamic efficiency of both cycles (top and bottom) [5],
c) research of organic fluids and their impact on hST [6].
Higher inlet temperatures, at the gas turbine inlet, resulted in
higher hGT [7]. Modern steam-turbine power plants, which are part
of the CCPP, use water at lower pressure levels, which somewhat
reduced negative effect of a pinch point. The problem of using
water in the lower pressure levels is that water at lower heat source
temperatures has low hSC because, in that case, operating param-
eters (pressure and temperature) are also low. By introducing the

https://doi.org/10.1016/j.energy.2018.01.033
0360-5442/© 2018 Elsevier Ltd. All rights reserved.
 S. Katulic et al. / Energy 147 (2018) 36e50
complex HRSG configurations to a power plant, it became neces-
sary to use advanced optimization methods, which made possible
to find optimal operating parameters of the HRSG depending on the
objective function. Manassaldi et al. [8] made superstructure that
embeds different alternative HRSG configurations. The objective
functions were a) The total net power generation, b) total heat
transfer area. It can be concluded from the paper that the layout of
the heat exchangers was not an optimization variable handled by
the optimization algorithm. The proposed configurations did not
have. Authors did not use parallel heat-exchanger's accommoda-
tion which is also less complex configuration and therefore optimal
solution cannot be found. Also, and it is not clear why flue gas
temperature at the HRSG outlet (stack temperature) is so high. By
using an optimal layout for the heat exchanger, the optimal oper-
ating parameter of stack temperature should be between 60 and
70
C. Mehrgoo et al. [9] have shown how to simultaneously
optimize the operating and geometric design parameters of the
HRSG by using the constructal theory. The objective function was
the total entropy generation. Author's did not optimize heat-ex-
changer's layout. Zhang et al. [10] optimized operation of HRSG
coupled with external heat-exchangers. HRSG was divided into
several sub-units. The position of the heat exchanger was deter-
mined by binary variables. That means that the position of the
heat-exchanger can be (if it exist) only at an advanced determined
location. Also, location of evaporator was fixed which is certain
limitation if optimal solution wants to be found. The mass flow rate
of the first pressure level can be determined using the mass and
energy equilibrium equations so that it is not clear why it was
selected as an optimization variable. Proposed HRSG configurations
do not have heat-exchanger's in their mutual parallel position
which proved to be disadvantage if optimal thermodynamic effi-
ciency of a plant wants to be obtained. Li et al. [11] presented a
method for waste heat utilization. Objective function was the net
power output. Different cycle configurations were evaluated, in
addition working fluid selection among the organic fluids was
performed. The optimization of heat-exchanger's layout was not in
the scope of their work. The results indicate that the regenerative
organic transcritical cycle produces the maximum power output at
source temperatures up to about 500
C. Nadir and Ghenaiet [12]
compared three different HRSG configurations operating at exhaust
gas temperature from 350
C to 650
C. The optimization variables
were operating parameters (pressure, temperature) but not the
heat-exchanger's layout. In their paper stack temperature was 96
C
which is too high compared to well optimized heat-exchanger's
layout and operating parameters. Valdes et al. [13] optimized the
combined cycle with multi pressure HRSG using the cost of
production per unit of generated electricity and annual cash flow as
an objective function. Bassily [14]] conducted numerical cost opti-
mization and irreversibility analysis of CCPP with triple pressure
HRSG. Operating parameters and the irreversibilities of the com-
ponents were analyzed. Koch et al. [15] applied an evolutionary
algorithm to the minimization of the product cost of CCPP. Authors
analyzed CCPP with two-pressure HRSG with a reheater. Authors
concluded that exergy is closely related to the economic value of an
energy carrier. Valdes and Rapun [16] presented a method for the
optimization of the HRSG based on the application of influence
coefficients. Authors concluded that application of influence
coefficients to the thermal system design permits better under-
standing of the effects of the modifications in the variables of these
systems. Katovicz and Bart [17] optimized HRSG with three
pressure levels and a reheater, analyzing influence of the fuel price
on the optimum operating parameters. The objective function was
the net present value of investment. Xiang et al. [18] reported that
today's hST reaches thermodynamic efficiency of 39.2%. The meth-
odology which they applied for calculating thermodynamic
37
efficiency was not mentioned. Cehil et al. [19] presented a novel
method for determining optimal heat-exchanger layouts for HRSG.
The method considers all possible heat-exchanger layouts, of each
pressure level, both in serial and parallel arrangement and water
was working fluid. The maximum hST was set as the objective
function. Bianchi et al. [20] focused on an innovative strategy to
improve waste heat conversion through the integration of a con-
ventional waste-to-heat power plant with a gas turbine. Authors
carried out parametric analysis of the effect of the discharged heat,
from a gas turbine, on the steam mass flow production in a steam
generator. They concluded that this conventional waste-to-heat
power plant provides power output increase up to 80% compared
to mid-size reference case.
Recently, there is a growing interest in exergoeconomic opti-
mization of power plant operation, especially of cogeneration
systems [21], systems with renewable energy sources [22] and
thermal power plants with coal as the primary fuel source [23].
Nadir et al. [24] conducted thermo-economic optimization of
different HRSG configurations for gas turbine outlet temperatures
ranging from 350
C to 650
C. The obtained results were used to
elaborate correlation between net present value and gas turbine
outlet temperature, flue gas mass flow rate, electricity selling price
and number of pressure levels of HRSG. Petrakopoulou et al. [25]
conducted a comprehensive exergy analysis to determine the
potential benefits of using the system with triple pressure HRSG.
Carapelluci and Giordano [26] compared two methods for opti-
mizing the operating parameters of CCPP: a) minimizing the cost
per unit of generated electricity and b) minimizing the objective
function which represents exergoeconomic losses associated with
inefficiencies of thermodynamic processes. Optimization was
performed for different configurations of HRSG (different number
of pressure levels), different gas turbines and different fuel prices.
Bakhshmand et al. [27] conducted exergoeconomic analysis and
optimization of CCPP with triple-pressure HRSG and with a
reheater. The objective function was the total cost rate of the power
plant. Once again, authors did not optimize heat-exchanger's
layout. Sharma and Singh [28] performed exergy analysis of dual-
pressure HRSG. Different physical parameters of HRSG such as fin
height, fin density and fin thickness were varied for analyzing
exergy efficiency at different operating pressures. Authors also
presented in Ref. [29] exergy analysis of dual-pressure HRSG for
varying dead states. Particular sections of the HRSG having
maximum exergy losses have been located. Naemi et al. [30]
performed thermodynamic and thermoeconomic analysis of dual-
pressure HRSG coupled with gas turbine in order to achieve
optimum operating parameters. Heat-exchanger's were only in
serial arrangement.
A power plant with a gas turbine in the top cycle generally has
water as working fluid in the bottom cycle because water has good
physical and thermodynamic properties at high temperatures and
is easily accessible. In systems in which the heat source is at lower
temperatures, it is possible to use other working fluids such as a
mixture of water and ammonia in a process known as the Kalina
cycle [6]. Generally, it can be said that at lower heat source
temperatures organic Rankine cycle (ORC) is an alternative to the
water in the CCPP's bottom cycle because it has better thermody-
namic properties [31]. Marrero et al. [32] conducted optimization of
a combined triple power cycle (gas-steam-ammonia). They
concluded that the ammonia bottoming cycle provides a more
efficient thermal matching for the triple cycle HRSG than the HRSG
of a conventional combined cycle. Carcasci et al. [33] have inves-
tigated the impact of the working fluid such as: toluene, benzene,
cyclohexane and cyclopentane selection on utilization of flue gas
heat at the gas heat from a turbine outlet. HRSG had only one
pressure level. Chacartegui et al. [34] investigated the use of
38 S. Katulic et al. / Energy 147 (2018) 36e50

toluene and cyclohexane in the CCPP's bottom cycle of two CCPP Table 2
configurations with different, commercially available, gas turbines. Optimization constraints.

HRSG consisted of only one pressure level. The authors used a Constraint Explanation
commercially available simulation software Thermoflow GT Pro. Tfg;in  Twf ;out  DTpp to satisfy pinch point temperature
Hettiarachchi et al. [35] presented economic results which show x  xmin anywhere during expansion e water
the influence of working fluid selection on the investment costs. and organic fluid
Four different organic fluids were selected as working fluids. The Tfg;out  Tfg;min minimal allowed temperature at
HRSG outlet e flue gas
authors conclude that the cost of heat-exchangers is by far the most
p < pcrit 0 eco ¼ 1p  pcrit 0 eco ¼ 0 criterion for number of economizers
dominant in the total cost of the investment. Toffolo et al. [36] pcond  prh  p1:pressure level constraint to the reheater pressure
presented a method that improves the selection of organic fluids, in Trh;in  T 0 ¼ f ðprh Þ 1. Pressure level expansion cannot be
Rankine cycle systems, taking into account several criteria at a time. into wet region
The method is applied to comparatively assess the performance of
isobutane and R134a, in the temperature interval between 130
C
and 180
C. The results show that the optimal cycle configuration is selection is essential in order to find HRSG configuration
in most cases subcritical for isobutane and supercritical recuper- which enables finding optimal hST and therefore optimal hCCPP . A
ated for R134a. Song et al. [37] modeled dual-loop ORC system thermodynamic model which has enabled finding optimal HRSG
which utilizes waste heat of a diesel engine. High temperature loop heat exchangers' layout (heat-exchangers of each pressure level can
uses water as a working fluid and low temperature loop uses be in their mutual parallel and/or serial position), optimal working
organic fluid. Yang et al. [38] investigated thermodynamic and parameters, while using organic working fluids in lower pressure
economic performances optimization for an ORC system recovering levels, is described in this paper.
the waste heat of exhaust gas from a diesel engine. Organic fluids
that were used were R1234ze, R245fa, R600, and R600a. The results
2. Thermodynamic model
showed that the maximum thermodynamic efficiency was
obtained while using R1234ze.
Assuming that the fuel is brought only to a gas turbine part of
Performance of a feed pump used in organic Rankine cycles was
CCPP and that there is no additional combustion in HRSG, hCCPP is
the topic of many scientific papers. Landelle et al. [39] investigated
calculated as
reciprocating pump running with an organic fluid, which was
R134a. They concluded that the use of a variable speed drive results PGT þ PST
in losses of nearly 20% of the initial consumption and that the hCCPP ¼ (2.1)
FGT
reciprocating pumps have an excellent volumetric efficiency even
at a high load. Also, pump should operate above its cavitation limit Power generated in the gas turbine cycle (PGT) represents net
which reduces ORC thermal efficiency. Richardson [40] investigated generated power. Also, power generated in the steam turbine cycle
thermodynamic performance of new thermofluidic feed pumps for (PST) represents net generated power. hGT and hST are calculated as
organic Rankine cycle applications.
PGT
By thorough literature survey of papers published in a range of hGT ¼ (2.2)
top energy journals, especially in the past four years, it has not been FGT
found that any of the authors carried out optimization of HRSG
configuration which, beside water in high pressure level integrates PST
hST ¼ (2.3)
organic fluids in lower pressure levels. FGT;out
Simultaneous optimization of HRSG heat-exchangers layout and
optimization of operating parameters of each pressure level was where
carried out. The objective function of optimization is maximization
of hST (Equation (2.15); the optimization variables shown in Table 4) FGT;out ¼ FGT $ð1  hGT Þ (2.4)
i.e. to utilize flue gas heat, which enters HRSG, and convert it, as
It should be noted that in the real process heat flux entering the
much as possible, to electrical energy. Input data are given in
HRSG is lower than FGT;out because there are certain heat losses
Table 1 and optimization constraints are given in Table 2. Simul-
between the gas turbine outlet and the HRSG inlet. These heat
taneous optimization and appropriate organic working fluids
losses are neglected because temperature of a heat flux entering
HRSG was input variable (chosen in advance). From the Equations
Table 1 (2.3) and (2.4) expression (2.5) is obtained
Input data.
PST
Variable Value hST ¼ (2.5)
FGT $ð1  hGT Þ


Tfg,in, C 600, 800
Tsuper,max,
C: Substituting Equations (2.2) and (2.5) into Equation (2.1)
acetone 276 expression (2.6) is obtained.
trichlorofluoromethane 349
water 600, 800 hGT $FGT þ hST $FGT $ð1  hGT Þ
Twf,in,
C 26.5 hCCPP ¼ (2.6)
Tenv,
C 15 FGT
Tcond,
C 25
Tfg,min,
C 27 i.e.
ht, % [] 90
hel$hmec, % 98 hCCPP ¼ hGT þ hST $ð1  hGT Þ (2.7)
hpum, % 84
g, % 0.5 The thermodynamic model describes the operation of the
xmin 0.8 steam-turbine part of CCPP where steam turbine cycle is repre-
qm,fg, kg/s 1
sented by Rankine cycle. The power plant, in its basic form, consists
 S. Katulic et al. / Energy 147 (2018) 36e50 39

of a HRSG (exchanged heat flux from 2 to 3 and from 4 to 5), a steam

turbine (expansion from 3 to 4 and from 5 to 6), a condenser
(driven away heat flux from 6 to 1) and feedwater pump (from 1 to
2). Heat flux exchanged between flue gas and working fluid can be
calculated as

Fex ¼ qm;wf $½ðh3  h2 Þ þ ðh5  h4 Þ (2.8)

Heat flux driven away in the condenser can be calculated as

Faway ¼ qm;wf $ðh6  h1 Þ (2.9)

The resulting net electrical power in the steam turbine is

PST ¼ qm;wf $½ðh3  h4 Þ þ ðh5  h6 Þ$hel $hmec  Ppum (2.10)

where the power required for the feedwater pump drive is calcu-
lated as

Ppum ¼ qm;wf $ðh2  h1 Þ (2.11)

It is important to emphasize that enthalpies h4, h6 and h2

represent real and not isentropic enthalpies at the turbine and
pump outlet. These enthalpies have been calculated using turbine
and pump isentropic efficiencies. hSC represents the degree of
conversion of heat flux, entering by flue gas to the HRSG, into
electrical power (2.13). It depends on the flue gas inlet temperature.

PST
hSC ¼ (2.12)
Fex

PST
hSC ¼
 (2.13)
qm;fg $ hfg;in  hfg;out

Since hST depends on hSC and hHRSG (2.14), that means that hST
will be lower at lower inlet temperatures of flue gas (2.15) Fig. 1. Schematic of bottom cycle of a CCPP.

F ex
hHRSG ¼
 (2.14)
qm;fg $ hfg;in  hfg;Tenv
hST ¼ hHRSG $hSC (2.15)
HRSG has two pressure levels. Of each pressure level heat-
exchangers can be in their mutual parallel and serial arrange-
ment. In thermodynamic model each pressure level, including
reheating, is represented by one row of exchangers which com-
prises of several exchangers arranged in series. An example of such
HRSG is shown in Fig. 1. The default number of possible serial
exchangers in each pressure level is 6, which proved to be a suffi-
cient number for unlimited optimization. Optimization was also
conducted for HRSG with 8 serial exchangers in each pressure level.
In that case optimization results were same as in case with 6 heat
exchangers and optimization process was about 30% more time
consuming because number of optimization variables increased.
Heat-exchangers are represented by economizer, evaporator,
superheater and reheater. Figs. 2 and 3 show possible (among other
possible) layouts of heat exchangers inside HRSG with two pressure
levels (Fig. 2) and two pressure levels with reheater (Fig. 3). It can
be seen that in Fig. 2 heat exchangers, of first and second pressure
level, are only in serial mutual accommodation but they can also be
in parallel position if optimization algorithm decides so. Fig. 3
shows that some of the heat-exchangers are in mutual parallel
position and some are in mutual serial position.
The genetic algorithm, which is a part of Matlab Optimization
Toolbox R2015a, was used for determining exchanged heat flow in
each heat-exchanger of each pressure level, pressure of each
pressure level, reheating pressure, superheating temperature of
each pressure level and working fluid mass flow rate of each
pressure level (except working fluid flow rate of the first pressure
level). Procedure of how does genetic algorithm (Matlab Optimi-
zation Toolbox R2015a) work is given in Appendix A. Working fluid
flow rate of first pressure level is calculated from the heat balance
equation. Each pressure level has its own steam turbine because
not all pressure levels have the same working fluid. Pressure drop of
the working fluid and flue gas is neglected because they depend
on the geometric characteristics of the HRSG and are somewhat
independent of the heat-exchangers’ arrangement. Pressure drop
of the flue gas can be reduced by increasing the cross-section of the
HRSG, as this reduces the velocity of the flue gases. Pressure drop of
the working fluid can be reduced by increasing the number of
parallel tubes in each heat-exchanger.
Flue gas inlet temperature (Tfg,in) of 600
C has been chosen
according to [41] while temperature of 800
C has been chosen
arbitrary in order to see does hST difference between HRSG with
organic fluid in second pressure level and with water in second
pressure level increases with increase of Tfg,in. Maximal allowed
superheating temperature (Tsuper,max) of acetone and R11 was
limited by NIST Transport Properties Database calculator
(REFPROP) which was used for calculating thermodynamic prop-
erties of working fluids. Temperatures Twf,in, Tenv, Tcond, Tfg,min have
been chosen arbitrary. Efficiencies ht, hel$hmec and hpum have been
chosen according to [5]. Heat losses due to radiation (g) have been
chosen according to [42] and minimally allowed moisture content
(xmin) according to [43].
40 S. Katulic et al. / Energy 147 (2018) 36e50

Fig. 2. Schematic of serial accommodation of heat-exchangers.

Fig. 3. Schematic of mixed serial and parallel accommodation of heat-exchangers inside HRSG.

The composition of fuel (natural gas) is given in Table 3. Specific a higher investment costs. Pinch point, within the HRSG, is caused
net calorific value of fuel is LHV ¼ 41.57 MJ/kg and is l ¼ 3. The list by a change in the specific heat capacity of the working fluid. The
of optimization variables in case of a dual pressure HRSG with a HRSG could utilize flue gas heat flux maximally if flue gas was
reheater is given in Table 4. cooled to the inlet temperature of a working fluid. Flue gas cannot
Pinch point is defined as the minimum temperature difference be cooled to the working fluid inlet temperature because of pinch
between flue gas and the working fluid (Fig. 4). Pinch point can point phenomenon. By selection of operating parameters, such as
occur inside economizer or at the economizer outlet i.e. at the pressure and temperature, and appropriate heat-exchangers'
evaporator inlet. Generally, it can be said that, for smaller values of layout, flue gas temperature at the HRSG outlet can be controlled.
pinch point utilization of the flue gas heat flux is better. Lower The assumption that the pinch point occurs only at the evapo-
values of pinch point result in greater heat-exchanger size, and thus rator inlet is not true. Specific heat capacity of the working fluid,
approaching saturation temperature, changes faster (increases) as
operating pressure increases. To find a position of pinch point,
Table 3 inside the HRSG, it is necessary to know the temperature profile of
Fuel composition. working fluid and flue gas and the value of pinch point. Pinch point
Component Vol, % value can be determined in advance (thermodynamic optimization)
or it can be optimization variable (exergoeconomic optimization).
Methane (CH4) 85.0
Ethane (C2H6) 5.0 In order to find the temperature profile of working fluid and flue
Propane (C3H8) 3.0 gas, each heat-exchanger, of each pressure level and reheater, was
Carbon dioxide (CO2) 3.5 divided into 200 segments. This way, accurate temperature profiles
Nitrogen (N2) 3.5 can be calculated in each heat-exchanger and thus the exact
Hydrogen sulfide (H2S) 0.0
location of pinch point can be found. The temperature profile of
 S. Katulic et al. / Energy 147 (2018) 36e50 41

Table 4
Optimization variables for dual pressure HRSG with a reheater.

v (1) e v (5) proportion of heat flux exchanged between flue gas and working fluid in each heat-exchanger, first pressure level
v (6) position of evaporator in heat-exchangers network, first pressure level
v (7) e v (11) proportion of heat flux exchanged between flue gas and working fluid in each heat exchanger, second pressure level
v (12) position of evaporator in heat-exchangers network, second pressure level
v (13) fresh steam pressure, first pressure level
v (14) fresh steam/vapor pressure, second pressure level
v (15) superheating temperature, first pressure level
v (16) superheating temperature, second pressure level
v (17) ratio of second pressure level fluid mass flow and first pressure level fluid mass flow
v (18) e v (22) proportion of heat flux exchanged between flue gas and working fluid in each heat-exchanger, reheater
v (23) position of reheater in heat-exchangers network
v (24) reheat pressure
v (25) reheat temperature

direction from the flue gas inlet towards the flue gas outlet
from the HRSG (opposite to the calculation of a working fluid
temperature profile). Each segment of the heat-exchanger is
indicated by j.

X
b
qm;k $Dhk
hfg ðj þ 1Þ ¼ hfg ðjÞ  (2.19)
1g
k¼1


Tfg ðj þ 1Þ ¼ f hfg ðj þ 1Þ (2.20)

X
b
qm;k $Dhk
Ffg ðj þ 1Þ ¼ Ffg ðjÞ þ (2.21)
1g
k¼1

Equation (2.19) is used for determining enthalpy of a flue gas in

Fig. 4. Position of a pinch point inside economizer.
working fluid and flue gas within the HRSG can be determined by
solving the following equations:
a) The equations are used for determining the temperature profile
of working fluid of ith pressure level and heat flux delivered to
the working fluid. Temperature profile is calculated in the
direction from the flue gas outlet towards the flue gas inlet to
the HRSG. Each segment of the heat-exchanger is indicated by j.
hwf;i ðj þ 1Þ ¼ hwf ;i ðjÞ þ Dhi


Twf;i ðj þ 1Þ ¼ f hwf ;i ðj þ 1Þ; pwf ;i
heat-exchanger segment. Product qm;k $Dhk represents total
exchanged heat, in each heat-exchanger segment, between flue gas
and a working fluids. The function between flue gas temperature
and enthalpy (2.20) is calculated by using the function written in
Matlab, which is part of a described thermodynamic model. Total
exchanged heat of the flue gas in each segment of a heat-exchanger
is given by Equation (2.21). Illustration of determining enthalpy,
temperature and exchanged heat of a working fluid and flue gas in a
heat-exchanger is given in Fig. 5. The mass flow of the working fluid
in pressure levels different from the first pressure level is deter-
mined by the optimization algorithm. The mass flow of a working
fluid in first pressure level is calculated by using “fzero” function, in
a way that temperature profiles of flue gas and working fluid in
each pressure level are calculated and after that, the program finds
(2.16) minimal temperature difference (2.22) between those temperature
profiles (pinch point), which has to be equal to the temperature
difference imposed by the model before the process of optimization
(2.17) (2.23) (thermodynamic optimization).

DTpp;i ¼ Tfg;i  Twf ;i (2.22)

Fwf ;i ðj þ 1Þ ¼ Fwf ;i ðjÞ þ Dhi $qm;i
Specific enthalpy in each segment of a heat-exchanger (2.16) is
calculated and then the temperature at the boundary of each
segment (2.17). Specific enthalpy increment in heat-exchanger
segment of each pressure level (i) is given by variable Dhi which
is determined by optimization algorithm. The temperature of
working fluid is given as a function of operating pressure and
enthalpy and can be calculated using REFPROP. Absolute exchanged
heat of a working fluid in each heat-exchanger segment is calcu-
lated by Equation (2.18).
b) The equations are used for determining the temperature profile
of flue gas and heat flux exchanged with the working fluid
and the environment. Temperature profile is calculated in the
(2.18)
 
min DTpp;i ¼ DTpp;imposed (2.23)
3. Results
Thermodynamic optimization was conducted for a steam-
turbine power plant with dual pressure (2 PL) HRSG and dual
pressure HRSG with a reheater (2 PL þ RH). A detailed analysis was
conducted for dual pressure HRSG with a reheater for the flue gas
inlet temperature of 600
C. The results for the dual pressure HRSG
with a reheater for the flue gas inlet temperature of 800
C are
given in Table 5 and results for dual pressure HRSG without a
42 S. Katulic et al. / Energy 147 (2018) 36e50

working fluid flue gas

n (n-1) (n-2) 3 2 1 1 2 3 (n-2) (n-1) n

Twf,1(n) Twf,1(n-1) Tfg(1)

Tfg(2)

b qm, k hk
Tfg(3) htot
1 g rad
Twf,1(n-2) k 1

flue flue flue

Temperature

Temperature
gas gas gas
+ h1 + h1 + h1 + h1 - htot - htot - htot - htot

Tfg(n-2)
Twf,1(2)
Twf,1(3) Tfg(n-1)
Twf,1(1)
Tfg(n)

wf,1(n) wf,1(n-1) wf,1(n-2) wf,1(2) wf,1(2) wf,1(1) fg(1) fg(2) fg(3) fg(n-2) fg(n-1) fg(n)

Heat flux Heat flux

Fig. 5. Illustration of determining enthalpy, temperature and exchanged heat of a working fluid and flue gas in a heat-exchanger.

Table 5
Results of thermodynamic optimization of dual pressure HRSG with a reheater, Tfg,in ¼ 800
C.

Variable Value

working fluid water acetone R11

pressure level 1st 2nd reheater 1st 2nd reheater 1st 2nd reheater

ɳSC, % 45.88 44.45 46.11 25.39 46.06 23.27

DTpp
C 0 0.6 0 0.03 0 0.27
p, bar 607.14 590.85 40.46 785.68 71.6 144.31 693.3 109.7 72.82
pcond, bar 0.03 0.03 0.03 0.31 0.03 1.06
x 0.8 0.8 0.8 0.91 0.8 0.97
Tsuper,
C 789.7 765.7 341.18 799 264.46 519.97 799 256.69 417.27
qm, kg/s 0.226 0.001 0.226 0.22 0.062 0.22 0.223 0.149
Фeco, kW 457.8 1.69 451.5 38.09 455.1 29.91
Фeva, kW 0 0 0 0 0 0
Фsuper, kW 372.1 1.32 3.8 344.5 2.74 25.69 362 5.05 11.46
Фshare,% 99.19 0.36 0.45 92.29 4.73 2.98 94.62 4.05 1.33
PST,tot, kW 398.7 1.39 399.2 11.03 399.8 9.43
Ppum,tot, kW 16.14 0.058 20.32 0.666 18.2 1.301
Ptot, kW 382.5 1.33 378.9 10.36 381.6 8.13
Tfg,out,
C 52.68 27.82 26.86
ɳST,tot, % 42.85 43.46 43.51
ɳSC,tot, % 45.88 45.13 45.14
ɳHRSG, % 93.41 96.29 96.4

reheater are given in Table 6. The optimization of single pressure
HRSG would be trivial and therefore not interesting for the opti-
mizing of heat-exchangers' layout. It should be noted that in the
thermodynamic optimization pressures of a working fluid achieved
very high values (somewhere over 700 bar) and that the imposed
pinch temperature was 0
C, which are not realistic for practical
use. The results of thermodynamic efficiency obtained by exer-
goeconomic optimization (thermoeconomic optimization) can be
compared to those obtained by the thermodynamic optimization.
Comparing those results it can be seen how inefficient is a real
power plant compared to the theoretical power system. Working
fluids used in second pressure level were water, acetone and
trichlorofluoromethane (R11). In every HRSG the configuration
working fluid of first pressure level was water. Fig. 1 shows a steam-
turbine power plant with dual pressure HRSG and a reheater.
3.1. Results of thermodynamic optimization with water as a
working fluid in second pressure level and Tfg,in ¼ 600
C
(2 PL þ RH)
Superheating temperature of first pressure level is 592
C and of
reheating 375
C (T-s diagram, Fig. 6). It can be seen that the
reheating temperature has not reached the temperature of fresh
steam superheating. The optimization algorithm has chosen this
pressure in first pressure level and a temperature of reheating,
although there is probably a solution with lower pressure in the
 S. Katulic et al. / Energy 147 (2018) 36e50 43

Table 6
Results of thermodynamic optimization of dual pressure HRSG without a reheater, Tfg,in ¼ 600
C and Tfg,in ¼ 800
C.

Variable Value

Tfg,in 600
C 800
C

working fluid water acetone R11 water acetone R11

pressure level 1st 2nd 1st 2nd 1st 2nd 1st 2nd 1st 2nd 1st 2nd

ɳSC, % 41.07 26.93 41.06 25.35 41.08 25.01 44.89 44.97 45 25.37 45.01 22.95
DTpp,
C 0 0.01 0.02 0 0.01 0 0 0.5 0 0 0 0
p, bar 241.4 9.58 238.1 67.68 241.4 180.3 707.8 693 719.9 67.4 677.8 197.2
pcond, bar 0.032 0.032 0.032 0.307 0.032 1.06 0.032 0.032 0.032 0.307 0.032 1.06
x 0.8 0.81 0.8 0.91 0.8 1.04 0.8 0.8 0.8 0.91 0.81 0.94
Tsuper,
C 598.6 197.3 599 261.3 599 297.6 792.3 797 798.8 261.2 799 271.9
qm, kg/s 0.151 0.027 0.149 0.182 0.151 0.444 0.231 0.0003 0.229 0.037 0.227 0.132
Фeco, kW 297.9 17.3 295.6 112.3 298.1 91.1 471.7 0.549 468.1 22.7 462.5 27.2
Фeva, kW 0 54.5 0 0 0 0 0 0 0 0 0 0
Фsuper, kW 209.8 1.32 209 8.09 210 25.14 367.7 0.435 367.8 1.67 370.4 3.5
Фshare, % 87.4 12.6 80.7 19.3 81.4 18.6 99.9 0.1 97.2 2.8 96.4 3.6
PST,tot, kW 212.9 19.7 211.4 32.37 213 35.43 396.1 0.464 395.5 6.55 392.9 9.12
Ppum,tot, kW 4.32 0.03 4.23 1.86 4.33 6.35 19.24 0.022 19.4 0.374 18.1 2.07
Ptot, kW 208.6 19.7 207.2 30.51 208.7 29.08 376.9 0.442 376.1 6.18 374.8 7.06
Tfg,out,
C 70.28 27.85 28.5 49.07 30.05 26.78
ɳST,tot, % 34.7 36.17 36.19 42.12 42.68 42.63
ɳSC,tot, % 39.29 38.03 38.09 44.89 44.45 44.22
ɳHRSG, % 88.4 95.12 95.01 93.82 96.03 96.41

Fig. 6. T-s diagram of a working cycle e dual pressure HRSG (water in second pressure level) with a reheater, Tfg,in ¼ 600
C.

first pressure level and higher temperature of reheating, which

gives a very similar value of thermodynamic efficiency. It should
also be noted that the expansion in the turbine of the first pressure
level reached the line of dry-saturated steam. There is a possibility
that better optimization result would be obtained if expansion
entered in the region of wet steam, but one of the boundary con-
ditions, imposed in the thermodynamic model, was that if HRSG
has a reheater, then expansion of the first pressure level turbine
cannot enter wet steam region because that would additionally
complicate the optimization process. Also, if expansion in the
turbine of the first pressure level entered wet steam region, then a
new pinch point would occur (isothermal process), which would
act negatively on the cooling of the flue gases. The mass flow of the
first pressure level, which is 0.147 kg/s, is slightly smaller than in
configurations without a reheater. This is a result of introduction of
a reheater which takes over part of the flue gas heat flux which, in
case of HRSG without a reheater, was taken over by a mass flow of
first pressure level. It can be seen in the T-Ф diagram (Fig. 7, x-axis
represents total heat flux which exchanged flue gas) that reheating
process starts at a temperature which is lower than the Fig. 7. T-Ф diagram of a working cycle e dual pressure HRSG (water in second pressure
level) with a reheater, Tfg,in ¼ 600
C.
44 S. Katulic et al. / Energy 147 (2018) 36e50
Fig. 8. T-Ф diagram of each pressure level and reheater (water in second pressure level), Tfg,in ¼ 600
C.
temperature at which pinch point of first pressure level occurs.
Usually, reheating process starts at the temperatures which are
above the temperature at which pinch point of first pressure level
occurs and probably this optimization result is not global optimum,
but also, as it turned out during the optimization, there are several
combinations of operating parameters that give very similar results
of the thermodynamic efficiency. T-Ф diagrams for each pressure
level and reheater are shown in Fig. 8.
As shown in Ref. [44], by not introducing the second pressure
level to the single pressure HRSG with a reheater, temperature of
the flue gas at the HRSG outlet would be around 200
C. This
justifies the introduction of the second pressure. The pressure of
second pressure level is 11.61 bar and superheating temperature is
212
C. Working fluid mass flow is 0.019 kg/s. The moisture con-
tent at the turbine outlet is almost minimally allowed and its
value is x ¼ 0.81. By introduction of the second pressure level flue
gas at the HRSG outlet was cooled to 81.13
C. In this case, tem-
perature of the flue gas at the HRSG outlet is higher than in dual
pressure HRSG without a reheater. This is because introduction of
a reheater necessarily reduces working fluid mass flow of first
pressure level, so that there is less working fluid mass flow below
the pinch point temperature of first pressure level which results in
higher outlet temperature. The pressure of the second pressure
level is somewhat higher than in case of HRSG without a reheater
and therefore hSC is higher too, but on the contrary, hHRSG is
slightly lower. By conducted optimization the compromise
between hSC and hHRSG , according to the Equation (2.15), is found
which resulting in hST ¼ 35.35% and a total exchanged heat flux of
569.7 kW.
3.2. Results of thermodynamic optimization with acetone as a
working fluid in second pressure level and Tfg,in ¼ 600
C
(2 PL þ RH)
In this case, working fluid in second pressure level is acetone.
Superheating temperature of first pressure level and of reheating is
598
C. Working pressure of first pressure level is 518 bar, which is
considerably higher than in 3.1. This is because the temperature of
reheating is almost at its allowed maximum of 600
C. The moisture
content at the reheater's steam turbine outlet is x ¼ 0.88. If working
pressure of the first pressure level was lower than 518 bar then, the
moisture content at the reheater's steam turbine outlet would be
higher than x ¼ 0.88, which would affect hSC . As it can be seen from
the Fig. 9, hSC decreases with a decrease of moisture content at the
steam turbine outlet. The moisture content at the reheater's outlet
is affected by superheating temperature and by decreasing this
temperature, the moisture content decreases. It has to be empha-
sized that values greater than 1 do not actually represent moisture
content (moisture content is defined between 0 and 1), but only
show that expansion in the turbine was into saturated steam
region.
The optimization algorithm is trying to find a combination of
pressure of the first pressure level, reheater's superheating tem-
perature and moisture content at the steam turbine outlet, what
will ensure maximum thermodynamic efficiency of the steam-
turbine plant. The pressure of reheating is 70.4 bar. Working fluid
mass flow of the first pressure level, and therefore also of reheating,
is 0.122 kg/s which is slightly less than in the configuration without
reheater due to the introduction of a reheater, as explained in 3.1. As
 S. Katulic et al. / Energy 147 (2018) 36e50
Fig. 9. Thermodynamic efficiency of a steam cycle as a function of superheating
0
temperature and moisture content and fraction hhh
00
h0
at the turbine outlet (water as a
working fluid).
it can be seen from the T-Ф diagram (Figs. 11 and 12), temperature
profiles of working fluid of the first pressure level and a reheater
almost adhere to the temperature profile of flue gas, which is an
indicator of good utilization of the flue gas heat.
Working pressure of the second pressure level is 46.4 bar which
is very close to the critical pressure. It can be seen from the T-s
diagram (Fig. 10) that second pressure level does not have a
superheater, but acetone enters turbine as dry superheated vapor.
The temperature at the turbine inlet is 234.5
C, which corresponds
to the saturation temperature at 46.4 bar. The moisture content at
the turbine outlet is x ¼ 0.91. As seen in the T-s diagram, turbine
expansion line touches line x ¼ 0.8 which suggests that, in this case,
for a given working pressure of 46.4 bar, maximum possible
moisture content at the turbine outlet is achieved. The mass flow
rate of acetone is 0.225 kg/s. In this case, as in 4.1, the heat of the
flue gas is utilized at temperatures below pinch point occurrence of
first pressure level (390
C). The flue gas at the HRSG outlet has a
temperature of 30.15
C, which is for 50
C lower than in the case
where water was working fluid in the second pressure level. hSC of
Fig. 10. T-s diagram of a working cycle e dual pressure HRSG (acetone in second pressure level) with a reheater, Tfg,in ¼ 600
C.
45
Fig. 11. T-Ф diagram of a working cycle e dual pressure HRSG (acetone in second
pressure level) with a reheater, Tfg,in ¼ 600
C.
the second pressure level is 24.95%, which is less than in 4.1
(27.79%). However, hST is higher (37.24%), which is the consequence
of high hHRSG (94.75%), in contrast to hHRSG in 4.1 (86.69%). Total
exchanged heat flux is 622.6 kW.
3.3. Results of thermodynamic optimization with R11 as a working
fluid in second pressure level and Tfg,in ¼ 600
C (2 PL þ RH)
In this case, a working fluid in the second pressure level is R11.
The working pressure of the first pressure level is 469 bar and of
reheating 51.9 bar. The superheating temperature of the first pres-
sure level is 596
C and of reheating 581
C. The working fluid mass
flow rate of the first pressure level is 0.124 kg/s. The moisture
content at the turbine outlet of a reheater is x ¼ 0.89. As it can be
seen from the T-s diagram (Fig. 13) expansion in the high-pressure
turbine of the first pressure level reached the line of dry-saturated
steam. Pinch point of the first pressure level occurs at a tempera-
ture of about 385
C and the working fluid of the second pressure
level utilizes the heat of the flue gas below this temperature, as it
can be seen in the T-Ф diagram (Figs. 14 and 15). hSC of the first
pressure level is 43.51% and is almost the same as in 4.2.
Operating parameters of the second pressure level are 272.13
C
and 145 bar. Operating pressure in this case is supercritical. The
46 S. Katulic et al. / Energy 147 (2018) 36e50

Fig. 12. T-Ф diagram of each pressure level and reheater (acetone in second pressure level), Tfg,in ¼ 600
C.

Fig. 13. T-s diagram of a working cycle e dual pressure HRSG (R11) in second pressure level) with a reheater, Tfg,in ¼ 600
C.

mass flow of R11 is 0.554 kg/s. The moisture content at the turbine
outlet is x ¼ 0.98. In Fig. 16 it can be seen that the maximum hSC , for
a superheating temperature of 272.13
C, is achieved if vapor at the
turbine outlet is in the superheated region. As with Fig. 9, in Fig. 16
values greater than 1 do not actually represent moisture content,
but only show that expansion in the turbine was into saturated
vapor region. Superheated vapor at the turbine outlet can be ach-
ieved by reducing the operating pressure of R11, but reducing the
operating pressure increases flue gas temperature at the HRSG
outlet (Fig. 17) and thus hHRSG decreases. The optimization algo-
rithm was trying to find a compromise between the hSC and hHRSG . It
is shown on T-Ф diagram that the change in temperature of the
working fluid of the second pressure level, in the region of total
exchanged heat flux from 170 kW to 310 kW, is practically negli-
gible, so the superheater of the second pressure level, in that region,
can be omitted. Total exchanged heat flux of flue gas is 622.2 kW. In
Table 9 optimization results are shown.
4. Conclusion
As it can be seen from this analysis, the optimal solution consists
of a compromise between the hSC and hHRSG . Objective function was
 S. Katulic et al. / Energy 147 (2018) 36e50 47

Fig. 14. T-Ф diagram of a working cycle e dual pressure HRSG (R11 in second pressure
level) with a reheater, Tfg,in ¼ 600
C.

Fig. 16. Thermodynamic efficiency of a cycle as a function of superheating tempera-

0
ture, moisture content and fraction hhh
00
h0
at the turbine outlet (R11 as a working fluid).

maximization of hST so there is optimal correlation between hHRSG

and hSC that gives optimal hST . This is why it is said that solution
consists of compromise between hHRSG and hSC . Optimization
algorithm does not recognize hHRSG and hSC separately but opti-
mizes only objective function (hST ). High hSC does not guarantee a
high hST . Therefore, it is necessary to find operating parameters of
all pressure levels and heat-exchangers' layout which will give
maximum possible hST . Pinch point, which occurs mostly inside the
economizer, disables better utilization of flue gas heat. In order to
additionally reduce negative effect of pinch point occurrence,
together with the heat-exchangers' layout which is determined by
the optimization algorithm, organic working fluids are used in
lower pressure levels. Comparing constraint variables, given
in Table 2, and comparing them with their values, presented in

Fig. 15. T-Ф diagram of each pressure level and reheater (R11 in second pressure level), Tfg,in ¼ 600
C.
48 S. Katulic et al. / Energy 147 (2018) 36e50

Table 8
Results of thermodynamic optimization with acetone as a working fluid in second
pressure level and Tfg,in ¼ 600
C (2 PL þ RH).

Variable Value

Tfg,in 600
C

working fluid acetone

pressure level 1st 2nd reheater

ɳSC, % 43.75 24.95

DTpp
C 0.00 0.12
p, bar 518.30 46.40 70.45
pcond, bar 0.03 0.31
x 0.88 0.91
Tsuper,
C 598.30 234.58 597.51
qm, kg/s 0.122 0.225 0.122
Фeco, kW 245.3 129.8
Фeva, kW 0.00 13.84
Фsuper, kW 127.8 3.70 102.2
Фshare,% 59.92 23.67 16.41
PST,tot, kW 215.3 38.35
Ppum,tot, kW 7.43 1.57
Fig. 17. Temperature difference between a working fluid and flue gas at the HRSG Ptot, kW 207.9 36.77
outlet as a function of superheating temperature, operation pressure and pinch point Tfg,out,
C 30.15
temperature (DTpp ¼ 0
C) e R11 as a working fluid. ɳST,tot, % 37.24
ɳSC,tot, % 39.30
ɳHRSG, % 94.75
Tables 5e9, it can be seen for which values of constraint variables
optimum solution was found. Optimum solutions were found in the
vicinity of the boundary of constraints DTpp and Tfg,out. Values of Table 9
other constraints were not all ways on the boundary. Results of thermodynamic optimization with R11 as a working fluid in the second
The optimization results show that from the thermodynamical pressure level and Tfg,in ¼ 600
C (2 PL þ RH).
point of view the use of acetone and R11 in lower pressure levels, Variable Value
for the same operating conditions, has an advantage over the water
Tfg,in 600
C
i.e. by using acetone and R11 higher hST can be achieved. According
working fluid R11
to the obtained results it can be concluded that:
pressure level 1st 2nd reheater
a) By using acetone, in a case of Tfg,in ¼ 600
C, and R11 and by ɳSC, % 43.51 23.73
enabling optimization algorithm to determine heat-ex- DTpp
C 0.00 0.08
changers' layout, it is possible to achieve higher hST and p, bar 469.20 145.36 51.59
therefore higher hCCPP. pcond, bar 0.03 1.06
x 0.89 0.98
b) Higher hST can be achieved at higher temperatures of the flue Tsuper,
C 595.79 272.13 581.07
gas at the HRSG inlet. qm, kg/s 0.124 0.554 0.124
Фeco, kW 249.5 112.2
Фeva, kW 0.00 0.00
Фsuper, kW 136.5 21.36 102.8
Фshare,% 62.02 21.47 16.51
Table 7 PST,tot, kW 219.6 38.11
Results of thermodynamic optimization with water as a working fluid in second Ppum,tot, kW 6.88 6.39
pressure level and Tfg,in ¼ 600
C (2 PL þ RH). Ptot, kW 212.7 31.71
Variable Value Tfg,out,
C 30.25
ɳST,tot, % 37.20
Tfg,in 600
C ɳSC,tot, % 39.26
ɳHRSG, % 94.74
working fluid water

pressure level 1st 2nd reheater

ɳSC, % 42.12 27.79 c) hST differences between power plant when considering
DTpp
C 0.47 0
p, bar 402.93 11.61 40.56
acetone or R11 and water, at lower pressure levels, decrease
pcond, bar 0.03 0.03 as temperature of the flue gas at the HRSG inlet increases.
x 0.82 0.81 d) Achieving only high hSC and not hHRSG , is not enough for
Tsuper,
C 591.6 212.52 375.7 achieving high hST , but it is necessary to find a compromise
qm, kg/s 0.147 0.019 0.147
between them.
Фeco, kW 294.2 13.25
Фeva, kW 0 38.81
Фsuper, kW 171.5 1.3 50.66 Appendix A
Фshare,% 81.74 9.37 8.89
PST,tot, kW 224.5 14.86
The following outline summarizes how the genetic algorithm
Ppum,tot, kW 7 0.027
Ptot, kW 217.4 14.83 works [45]:
Tfg,out,
C 81.13
ɳST,tot, % 35.35 1. The algorithm begins by creating a random initial population.
ɳSC,tot, % 40.78 2. The algorithm then creates a sequence of new populations. At
ɳHRSG, % 86.69
each step, the algorithm uses the individuals in the current
 S. Katulic et al. / Energy 147 (2018) 36e50 49

generation to create the next population. To create the new References

population, the algorithm performs the following steps:
a Scores each member of the current population by computing [1] DOE/EIA-0383 Annual energy outlook 2013 with projections to 2040. 2013.
April 2013.
its fitness value. [2] DOE/EIA-0383 Annual energy outlook 2014 with projections to 2040. 2014.
b Scales the raw fitness scores to convert them into a more April 2014.
usable range of values. [3] Lu H, Price L, Zhang Q. Capturing the invisible resource: analysis of waste heat
potential in Chinese industry. Appl Energy 2016;161:497e511.
c Selects members, called parents, based on their fitness. [4] Mohagheghi M, Shayegan J. Thermodynamic optimization of design variables
d Some of the individuals in the current population that have and heat exchangers layout in HRSGs for CCGT, using genetic algorithm. Appl
lower fitness are chosen as elite. These elite individuals are Therm Eng 2009;29:290e9.
[5] Bassily AM. Modeling, numerical optimization, and irreversibility reduction of
passed to the next population. a triple-pressure reheat combined cycle. Energy 2007;32:778e94.
e Produces children from the parents. Children are produced [6] Poullikkas A. An overview of current and future sustainable gas turbine
either by making random changes to a single paren- technologies. Renew Sustain Energy Rev 2005;9:409e43.
[7] Hunt JR. The history of the industrial gas turbine (Part 1 the first fifty years
tdmutationdor by combining the vector entries of a pair of
1940-1990). Bedford: Institution of Diesel and Gas Turbine Engineers; 2011.
parentsdcrossover. [8] Manassaldi JI, Arias AM, Scenna NJ, Mussati MC, Mussati SF. A discrete and
f Replaces the current population with the children to form the continuous mathematical model for the optimal synthesis and design of dual
next generation. pressure heat recovery steam generators coupled to two steam turbines.
Energy 2016;103:807e23.
3. The algorithm stops when one of the stopping criteria is met. [9] Amidpour Mehrgoo M. Constructal design and optimization of a dual pressure
heat recovery steam generator. Energy 2017;124:87e99.
Nomenclature [10] Zhang J, Liu P, Zhou Z, Ma L, Li Z, Ni W. A mixed-integer nonlinear
programming approach to the optimal design of heat network in a poly-
generation energy system. Appl Energy 2014;114:146e54.
[11] Li C, Wang H. Power cycles for waste heat recovery from medium to high
temperature flue gas sources e from a view of thermodynamic optimization.
Appl Energy 2016;180:707e21.
Variable Description Unit [12] Nadir M, Ghenaiet A. Thermodynamic optimization of several (heat recovery
g heat losses due to radiation e steam generator) HRSG configurations for a range of exhaust gas tempera-
tures. Energy 2015;86:685e95.
h enthalpy J/kg
[13] Valdes M, Duran D, Rovira A. Thermoeconomic optimization of combined
LHV lower heating value MJ/kg
cycle gas turbine power plants using genetic algorithms. Appl Therm Eng
p pressure bar
2003;23:2169e82.
P electrical power of a combined cycle power plant W [14] Bassily AM. Numerical cost optimization and irreversibility analysis of the
qm mass flow kg/s triple-pressure reheat steam-air cooled GT commercial combined cycle power
s entropy J/(kg plants. Appl Therm Eng 2012;40:145e60.
K) [15] Koch C, Cziesla F, Tsatsaronis G. Optimization of combined cycle power plants
T temperature K,
C using evolutionary algorithms. Chem Eng Process 2007;46:1151e9.
v (i), optimization variable [16] Valdes M, Rapun JL. Optimization of heat recovery steam generators for
i ¼ 1,..,n combine cycle gas turbine power plants. Appl Therm Eng 2001;21:1149e59.
x moisture content [17] Kotowicz J, Bartela L. The influence of economic parameters on the optimal
Dh enthalpy difference W values of the design variables of a combined cycle plant. Energy 2010;35:
DT temperature difference

C 911e9.
Greek Description Unit [18] Xiang W, Chen Y. Performance improvement of combined cycle power plant
based on the optimization of the bottom cycle and heat recuperation. Energy
Letters
Convers Manag 2007;16:84e9.
h thermodynamic efficiency 
[19] Cehil M, Katulic S, Schneider DR. Novel method for determining optimal heat-
l air excess ratio e exchanger layout for heat recovery steam generators. Energy Convers Manag
Ф heat flux W 2017. https://doi.org/10.1016/j.enconman.2017.03.033.
Subscript Description [20] Bianchi M, Branchini L, De Pascale A. Combining waste-to-energy steam cycle
CCPP combined cycle power plant with gas turbine units. Appl Energy 2014;130:764e73.
cond condenser [21] Zare V, Mahmoudi SMS, Yari M, Amidpour M. Thermoeconomic analysis and
crit critical optimization of an ammonia-water power/cooling cogeneration cycle. Energy
eco economizer 2012;47:271e83.
el electrical [22] Baghernejad A, Yaghoubi M. Exergoeconomic analysis and optimization of an
env environment integrated solar combined cycle system (ISCCS) using genetic algorithm.
eva evaporator Energy Convers Manag 2011;52:2193e203.
ex exchanged heat [23] Xiong J, Zhao H, Zhang C, Zheng C, Luh PB. Thermoeconomic operation opti-
mization of a coal-fired power plant. Energy 2012;42:486e96.
fg flue gas
[24] Nadir M, Ghenaiet A, Carcasci C. Thermo-economic optimization of heat
GT gas turbine
recovery steam generator for a range of gas turbine exhaust temperatures.
HRSG heat recovery steam generator
Appl Therm Eng 2016;106:811e26.
in inlet [25] Petrakopoulou F, Tsatsaronis G, Morosuk T, Carassai A. Conventional and
j,i,k counter advanced exergetic analyses applied to a combined cycle power plant. Energy
LHV lower heating value 2012;41:146e52.
max maximal [26] Carapelluci R, Giordano L. A comparison between exergetic and economic
mec mechanical criteria for optimizing the heat recovery steam generators of gas-steam power
min minimal plants. Energy 2013;58:458e72.
out outlet [27] Bakhshmand SK, Saray RK, Bahlouli K, Eftekhari H, Ebrahimi A. Exer-
PL pressure level goeconomic analysis and optimization of a triple-pressure combined cycle
pp pinch point plant using evolutionary algorithm. Energy 2015;93:555e67.
pum pump [28] Sharma M, Singh O. Exergy analysis of the dual pressure HRSG for varying
physical parameters. Appl Therm Eng 2017;114:993e1001.
rh reheater
[29] Sharma M, Singh O. Exergy analysis of dual pressure HRSG for different dead
SC steam cycle
states and varying steam generation states in gas/steam combined cycle
share share of exchanged heat of each pressure level in total
power plant. Appl Therm Eng 2016;93:614e22.
exchanged heat [30] Naemi S, Saffar-Avval M, Kalhori SB, Mansoori Z. Optimum design of dual
ST steam turbine pressure heat recovery steam generator using non-dimensional parameters
super superheater based on thermodynamic and thermoeconomic approaches. Appl Therm Eng
t turbine 2013;52:371e84.
tot total [31] Yamamoto T, Furuhata T, Arai N, Mori K. Design and testing of the organic
wf working fluid rankine cycle. Energy 2001;26:239e51.
50 S. Katulic et al. / Energy 147 (2018) 36e50
[32] Marrero IO, Lefsaker AM, Razani A, Kim KJ. Second law analysis and optimi-
zation of a combined triple power clcle. Energy Convers Manag 2002;43:
557e73.
[33] Carcasci C, Ferraro R, Miliotti E. Thermodynamic analyses of an organic
Rankine cycle for waste heat recovery from gas turbines. Energy 2014;65:
91e100.
[34] Chacartegui R, Sanchez D, Munoz JM, Sanchez T. Alternative ORC bottoming
cycles FOR combined cycle power plants. Appl Energy 2009;86:2162e70.
[35] Hettiarachchi MHD, Golubovic M, Worek WM, Ikegami Y. Optimum design
criteria for an Organic Rankine Cycle using low-temperature geothermal heat
sources. Energy 2007;32:1698e706.
[36] Toffolo A, Lazzaretto A, Manente G, Paci M. A multi-criteria approach for the
optimal selection of working fluid and design parameters in Organic Rankine
Cycle systems. Appl Energy 2014;121:219e32.
[37] Song J, Gu C. Performance analysis of a dual-loop organic Rankine cycle (ORC)
system with wet steam expansion for engine waste heat recovery. Appl
Energy 2015;156:280e9.
[38] Yang MH, Yeh RH. Thermodynamic and economic performances optimization
of an organic Rankine cycle system utilizing exhaust gas of a large marine
diesel engine. Appl Energy 2015;149:1e12.
[39] Landelle A, Tauveron N, Revellin R, Haberschill P, Colasson S, Roussel V.
Performance investigation of reciprocating pump running with organic fluid
for organic Rankine cycle. Appl Therm Eng 2017;113:962e9.
[40] Richardson ES. Thermodynamic performance of new thermofluidic feed
pumps for Organic Rankine Cycle applications. Appl Energy 2016;161:75e84.
[41] Franco A, Casarosa C. On some perspectives for increasing the efficiency of
combined cycle power plants. Appl Therm Eng 2002;22:1501e18.
[42] Ganapathy V. Industrial boilers and heat recovery steam generators design,
applications and calculations. New York: Marcel Dekker Inc; 2002.
[43] Casarosa C, Donatini F, Franco A. Thermoeconomic optimization of heat
recovery steam generators operating parameters for combined plants. Energy
2004;29:389e414.
[44] http://www.energy.siemens.com/us/pool/hq/energy-topics/technical-papers/
Enhanced%20WaterSteam%20Cycle%20for%20Advanced%20Combined%
20Cycle%20Technology.pdf.
[45] https://www.mathworks.com/help/gads/how-the-genetic-algorithm-works.
html?requestedDomain¼www.mathworks.com.