You are on page 1of 15

Applied Mathematical Modelling 39 (2015) 6830–6844

Contents lists available at ScienceDirect

Applied Mathematical Modelling


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

Ecological optimization of an irreversible Brayton cycle


with regeneration, inter-cooling and reheating
Santiago del Rio Oliveira ⇑, Vicente Luiz Scalon 1, Vitor Pereira Repinaldo 1
Department of Mechanical Engineering, Faculty of Engineering, Unesp University, Av. Eng. Luiz Edmundo C. Coube 14-01, Neighborhood: Vargem Limpa,
17033-360 Bauru, SP, Brazil

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

Article history: A mathematical model is developed for an irreversible Brayton cycle with regeneration,
Received 18 August 2014 inter-cooling and reheating. The irreversibility are from the thermal resistance in the heat
Received in revised form 30 December 2014 exchangers, the pressure drops in pipes, the non-isentropic behavior in the adiabatic
Accepted 4 February 2015
expansions and compressions and the heat leakage to the cold source. The cycle is
Available online 6 March 2015
optimized by maximizing the ecological function, which is achieved by the search for
optimal values for the temperatures of the cycle and for the pressure ratios of the first stage
Keywords:
compression and the first stage expansion. The advantages of using the regenerator,
Brayton cycle
Ecological optimization
intercooler and reheater are presented by comparison with cycles that do not incorporate
Intercooler one or more of these processes. Optimization results are compared with those obtained by
Regenerator maximizing the power output and it is concluded that the point of maximum ecological
Reheat function has major advantages with respect to the entropy generation rate and the thermal
Irreversibility efficiency, at the cost of a small loss in power.
Ó 2015 Elsevier Inc. All rights reserved.

1. Introduction

Brayton cycle is the ideal thermodynamic cycle used for the gas turbine analysis. Its usefulness varies from stationary
power generation to applications in transport, and also presents an advantageous relationship between high power output
and low weight of machinery. Due to its high applicability the search for improving its performance has been the goal of
several papers, specially the optimizations based on the finite time thermodynamics.
Curzon and Ahlborn [1] conducted one of the first studies in this area, finding that the thermal efficiency of an endore-
pffiffiffiffiffiffiffiffiffiffiffiffiffi
versible Carnot cycle at maximum power is equal to 1  T L =T H . Leff [2] expanded this same analysis for Brayton, Otto,
Diesel and Atkinson cycles, and found that the efficiencies at maximum work to these cycles are equal to that obtained
by Curzon and Ahlborn [1]. Bejan [3] studied the optimum distribution of the conductance between heat exchangers for a
Brayton cycle when the power is maximized and introduced a model to quantify the heat leakage. Ibrahim et al. [4]
optimized the power output for the Carnot and Brayton cycles considering both thermal reservoirs with finite and infinite
thermal capacitance rates. Wu and Kiang [5] studied the effects of incorporating non-isentropic processes in compressor
and turbine for the power output optimization of a Brayton cycle. Chen [6] observed that the addition of other kinds of

⇑ Corresponding author. Tel.: +55 (14) 3103 6119, Cell phone: +55 (14) 99612 6900.
E-mail addresses: santiago@feb.unesp.br (S. del Rio Oliveira), scalon@feb.unesp.br (V. Luiz Scalon), vitor_repinaldo@hotmail.com (V. Pereira Repinaldo).
1
Tel.: +55 (14) 3103 6119.

http://dx.doi.org/10.1016/j.apm.2015.02.029
0307-904X/Ó 2015 Elsevier Inc. All rights reserved.
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6831

Nomenclature

A heat transfer area [m2]


C_ I internal conductance rate of the heat engine [W/K]
C_ p heat capacity rate at constant pressure of the working fluid [W/K]
E_ ecological function [W]
E_ dimensionless ecological function
k specific heat ratio [dimensionless]
N number of transfer units [dimensionless]
p pressure [Pa]
Dp pressure drop [Pa]
Q_ heat transfer rate [W]
Q_ dimensionless heat transfer rate
rP pressure ratio [dimensionless]
S_ g entropy generation rate [W/K]
S_ g dimensionless entropy generation rate
T temperature [K]
T dimensionless temperature
U overall heat transfer coefficient [W/m2 K]
v specific volume [m3/kg]
W _ power output [W]
W _ dimensionless power output
x isentropic temperature ratio for the low pressure compressor [dimensionless]
y isentropic temperature ratio for the high pressure turbine [dimensionless]

Greek letters
e heat exchanger effectiveness [dimensionless]
g thermal efficiency, isentropic efficiency [dimensionless]
q pressure drop parameter [dimensionless]

Subscripts
0 environment
C compressor
Carnot Carnot cycle
E corresponding to ecological optimization
G global
H hot side
I heat leakage
L cold side
max maximum
R regenerator
s isentropic
T turbine
W corresponding to power optimization

irreversibility, in addition to the thermal resistance between working fluid and reservoirs, makes appear a point of maximum
thermal efficiency which has a finite amount of power, unlike the endoreversible cycles [1,2,7].
Angulo-Brown [8] introduced an ecological criterion, E ¼ W _  T L S_ g , for the optimization of a Carnot cycle, where W
_ is the
power output and S_ g is the entropy generation rate, and obtained the efficiency under conditions of maximum ecological
function is close to the average between the Carnot efficiency and the Curzon and Ahlborn efficiency [1]. Yan [9] suggested
_  T 0 S_ g ; in case the temperature of
that the proposed ecological function would make more sense if the expression was E_ ¼ W
the cold reservoir TL is different from the environment T0, and this modification was accepted by subsequent authors. The
ecological function was used for optimization of endoreversible and irreversible Carnot and Brayton cycles by several
authors [10–14]. Ust et al. [15] also utilized this criterion for the analysis of a Brayton heat engine with regeneration. In
all these studies it is observed that such optimization leads to greater thermal efficiencies along with lower entropy genera-
tion rates, at the cost of a small drop in power, when compared with the same cycle operating at maximum power
conditions.
6832 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

Ust et al. [16] presented a modification of the ecological function named ecological coefficient of performance,
ECOP ¼ W=T_ _
0 Sg ; which is used for the optimization of an irreversible Dual cycle. Such optimization was then expanded to
the analysis of an irreversible Brayton cycle by Ust et al. [17], and also to an irreversible regenerative Brayton cycle by
Ust et al. [18]. The results of using this type of criterion indicate an improvement in the thermal efficiency and in the entropy
generation rate when compared with the optimization of ecological function, but resulting in a considerable decrease in the
power output.
Another optimization criterion usually employed in finite time thermodynamics is the power density, which was first
introduced by Sahin et al. [19] for the optimization of a reversible Brayton cycle. This criterion is given by the ratio between
the power output W _ and the maximum specific volume of the cycle vMAX, P ¼ W= _ v MAX ; and used by Chen et al. [20] for the
optimization of an endoreversible Brayton with variable temperature reservoirs. Chen et al. [21] also did this optimization
for an irreversible intercooled regenerated Brayton cycle with optimal allocation of inventory for heat exchangers. The power
density optimization is shown to have a strong influence in reducing the size of power cycles.
The use of modifications in the Brayton cycle also is very important in improving the performance of a gas turbine and
therefore used together with the concepts of finite time thermodynamics for the cycle optimization. These kind of studies
include regeneration [12,15–19], inter-cooling [18,17–19] and reheating [18,19]. Haseli [22] optimized a regenerative
Brayton cycle using as criterion a modified second law efficiency. Wang et al. [23] optimized the power output of an
endoreversible closed intercooled regenerated Brayton cycle along with an optimal allocation of inventory for heat
exchangers. Also related to these types of reservoirs, Tyagi et al. [24] performed a thermodynamic analysis for a Brayton
cycle with regeneration, inter-cooling and reheating, analytically optimizing the power output and the thermal efficiency.
Sánchez-Orgaz et al. [25] modeled and optimized a Brayton cycle with respect to the power and efficiency incorporating
these same modifications, but with an arbitrary number of compressors and turbines.
Analysis for the maximum efficiency and the maximum power are also performed in models of internal combustion
engines. Zhao and Chen [26] analyzed the performance for an irreversible Dual cycle, which can be reduced to Otto and
Diesel cycles, optimizing the power and the efficiency and comparing the results for the three cycles. Lin and Hou [27] uti-
lized a new model for the types of heat loss found in internal combustion heat engines in order to analyze its influence on the
results for the maximum efficiency and maximum power for an air standard Miller cycle. Huijun et al. [28] developed an
exergoeconomic study of an irreversible closed Brayton cycle combining cooling, heating and power plant.
In this paper, the optimization criterion introduced by Angulo-Brown [8] and called ecological function is used as an
alternative for power and efficiency maximization. The ecological optimization is done to improve the performance of an
irreversible Brayton cycle with regeneration, inter-cooling and reheating through optimal design parameters. The analysis
will be performed in order to demonstrate the effect of addition of these processes in cycle performance. The results of
this kind of optimization will also be compared with the maximization of power output. Thus, the work aims to show the
benefits brought by the ecological optimization and by the addition a regenerator, an inter-cooler and a reheater. The main
difference of this study and those found in the literature is the dual optimization of Brayton cycle using the concept of
ecological function. The first optimization calculates the optimal temperature cycle for each state. The second optimization cal-
culates the compression and expansion ratios optimized in the first stage of the compressor and the first stage of the turbine.

2. Theoretical model

The path taken by the properties of the working fluid of the power plant is shown in the T–s diagram presented by Fig. 1
for an irreversible regenerate Brayton cycle with inter-cooling and reheating. In process 1–2 the working fluid enters the first
compressor at state 1 and suffers an irreversible adiabatic compression to reach state 2, with the process 1–2s representing
the process performed ideally (isentropic process).
The next process 2–3 represents the heat rejection Q_ L2 from the working fluid to a thermal reservoir at constant
temperature TL during the inter-cooling between the compressors. This way, as in process 1–2, the process 3–4 is an
irreversible adiabatic compression with the point 4s indicating the exit state of the second compressor if the process was
performed isentropically.
The working fluid then enters the regenerator at state 4 and suffers a process of heating to the state 4R due to heat
exchange with the exhaust gases leaving the low pressure turbine. This heat transfer rate is given by Q_ R .
From the state 4R occurs a heat addition process Q_ H1 , provided by the thermal reservoir at constant temperature TH, caus-
ing the working fluid to reach state 5. In process 5–6 occurs an irreversible adiabatic expansion in the high pressure turbine.
The reversible process is represented by the path 5–6s. After this the working fluid suffers another heat addition Q_ H2 , due to
the reheating process between turbines, to reach state 7.
The adiabatic expansion that occurs in the low pressure turbine is irreversible for the process 7–8 and reversible for the
process 7–8s. The exhaust gases that leaves the last turbine are cooled in the regenerator to state 8R, providing a heat
transfer rate Q_ R to the working fluid that leaves the high pressure compressor at state 4.
Finally, the working fluid is cooled to initial state 1 rejecting heat Q_ L1 to the thermal reservoir with temperature TL.
There is also considered a heat leakage Q_ I from thermal reservoir at temperature TH to the reservoir at temperature TL.
Equations for the different rates of heat transfer that occurs with the working fluid can be written as [3,10–15,18,20,21]:
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6833

Fig. 1. T–s diagram of an irreversible regenerated inter-cooled reheated Brayton cycle.

Q_ H1 ¼ C_ p eH1 ðT H  T 4R Þ ¼ C_ p ðT 5  T 4R Þ; ð1Þ

Q_ H2 ¼ C_ p eH2 ðT H  T 6 Þ ¼ C_ p ðT 7  T 6 Þ; ð2Þ

Q_ L1 ¼ C_ p eL1 ðT 8R  T L Þ ¼ C_ p ðT 8R  T 1 Þ; ð3Þ

Q_ L2 ¼ C_ p eL2 ðT 2  T L Þ ¼ C_ p ðT 2  T 3 Þ; ð4Þ

Q_ R ¼ C_ p eR ðT 8  T 4 Þ ¼ C_ p ðT 4R  T 4 Þ ¼ C_ p ðT 8  T 8R Þ; ð5Þ
where the effectiveness e for the heat exchangers are defined as [10–14,18,20,21]:

eH1 ¼ 1  expðNH1 Þ; ð6Þ

eH2 ¼ 1  expðNH2 Þ; ð7Þ

eL1 ¼ 1  expðNL1 Þ; ð8Þ

eL2 ¼ 1  expðNL2 Þ; ð9Þ

NR
eR ¼ ; ð10Þ
1 þ NR
and with the number of transfer units being:
U H1 AH1
NH1 ¼ ; ð11Þ
C_ p

U H2 AH2
NH2 ¼ ; ð12Þ
C_ p

U L1 AL1
NL1 ¼ ; ð13Þ
C_ p

U L2 AL2
NL2 ¼ ; ð14Þ
C_ p

U R AR
NR ¼ : ð15Þ
C_ p
6834 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

As the processes in the compressors and turbines are irreversible, the isentropic efficiencies are defined as
[4,11,14,15,18,20–22]:
T 2S  T 1
gC1 ¼ ; ð16Þ
T2  T1

T 4S  T 3
gC2 ¼ ; ð17Þ
T4  T3

T5  T6
gT1 ¼ ; ð18Þ
T 5  T 6S

T7  T8
gT2 ¼ : ð19Þ
T 7  T 8S
Two isentropic temperatures ratios x and y related to the low pressure compressor and high pressure turbine,
respectively, are also defined. These ratios are also functions of the pressure ratios of these processes [18,20–22]:
 k1
T 2S p2 k k1
x¼ ¼ ¼ rPC1
k
; ð20Þ
T1 p1

 k1
T5 p5 k k1
y¼ ¼ ¼ r PT1
k
: ð21Þ
T 6S p6

The irreversibility that arises due to the heat leakage Q_ I is considered using the linear model proposed by Bejan [3]:

Q_ I ¼ C_ I ðT H  T L Þ; ð22Þ

where C_ I is the rate of internal conductance of the heat engine.


Therefore, the heat transfer rate provided by the high temperature reservoir is given by Q_ H , which is the sum of Eqs. (1),
(2) e (22):

Q_ H ¼ Q_ H1 þ Q_ H2 þ Q_ I ; ð23Þ

and the heat transfer rate rejected to the low temperature reservoir é defined by Q_ L , which is the sum of Eqs. (3), (4) e (22):

Q_ L ¼ Q_ L1 þ Q_ L2 þ Q_ I : ð24Þ
The heat transfer processes are considered non isobaric with pressure drops given by Dp. These pressure drops are
quantified by the following parameters [22]:
 ðk1Þ  ðk1Þ
p5 k pH1  DpH1 k
qH1 ¼ ¼ ; ð25Þ
p4 pH1

 ðk1Þ  ðk1Þ
p7 k pH2  DpH2 k
qH2 ¼ ¼ ; ð26Þ
p6 pH2

 ðk1Þ  ðk1Þ
p1 k pL1  DpL1 k
qL1 ¼ ¼ ; ð27Þ
p8 pL1

 ðk1Þ  ðk1Þ
p3 k pL2  DpL2 k
qL2 ¼ ¼ : ð28Þ
p2 pL2
From Eqs. (1)–(5) and (16)–(21) the following temperature relations are obtained:

T 8R ¼ ðT 4  T 8 ÞeR þ T 8 ; ð29Þ

T 1 ¼ T 8R ð1  eL1 Þ þ eL1 T L =T H ; ð30Þ

T 2S ¼ xT 1 ; ð31Þ

T 2 ¼ T 1 þ ðT 2S  T 1 Þ=gC1 ; ð32Þ
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6835

T 3 ¼ T 2 ð1  eL2 Þ þ eL2 T L =T H ; ð33Þ

T 4S ¼ ðT 4  T 3 ÞgC2 þ T 4 ; ð34Þ

T 4R ¼ ðT 8  T 4 ÞeR þ T 4 ; ð35Þ

T 5 ¼ T 4R ð1  eH1 Þ þ eH1 ; ð36Þ

T 6S ¼ T 5 =y; ð37Þ

T 6 ¼ ðT 6S  T 5 ÞgT1 þ T 5 ; ð38Þ

T 7 ¼ T 6 ð1  eH2 Þ þ eH2 ; ð39Þ

T 8S ¼ T 7  ðT 7  T 8 Þ=gT2 ; ð40Þ
where the relations are written in dimensionless form dividing them by TH to generalize the results.
The combination of Eqs. (29)–(40) results in temperature relations as functions only of the dimensionless T 4 and T 8 :

T 1 ¼ a1 T 4 þ b1 T 8 þ c1 ; ð41Þ

T 2S ¼ xða1 T 4 þ b1 T 8 þ c1 Þ; ð42Þ

T 2 ¼ a2 T 4 þ b2 T 8 þ c2 ; ð43Þ

T 3 ¼ a3 T 4 þ b3 T 8 þ c3 ; ð44Þ

T 4S ¼ a4 T 4 þ b4 T 8 þ c4 ; ð45Þ

T 5 ¼ a5 T 4 þ b5 T 8 þ c5 ; ð46Þ

T 6S ¼ ða5 T 4 þ b5 T 8 þ c5 Þ=y; ð47Þ

T 6 ¼ a6 T 4 þ b6 T 8 þ c6 ; ð48Þ

T 7 ¼ a7 T 4 þ b7 T 8 þ c7 ; ð49Þ

T 8S ¼ a8 T 4 þ b8 T 8 þ c8 : ð50Þ
The dimensionless coefficients that arose from the combination procedure are listed here:
a1 ¼ eR ð1  eL1 Þ b1 ¼ ð1  eL1 Þð1  eR Þ c1 ¼ eL1 T L =T H
a2 ¼ a1 ðx  1 þ gC1 Þ=gC1 b2 ¼ b1 ðx  1 þ gC1 Þ=gC1 c2 ¼ c1 ðx  1 þ gC1 Þ=gC1
a3 ¼ a2 ð1  eL2 Þ b3 ¼ b2 ð1  eL2 Þ c3 ¼ c2 ð1  eL2 Þ þ eL2 T L =T H
a4 ¼ a3 ð1  gC2 Þ þ gC2 b4 ¼ b3 ð1  gC2 Þ c4 ¼ c3 ð1  gC2 Þ
:
a5 ¼ ð1  eH1 Þð1  eR Þ b5 ¼ eR ð1  eH1 Þ c5 ¼ eH1
a6 ¼ a5 ½ð1  gT1 Þy þ gT1 =y b6 ¼ b5 ½ð1  gT1 Þy þ gT1 =y c6 ¼ c5 ½ð1  gT1 Þy þ gT1 =y
a7 ¼ a6 ð1  eH2 Þ b7 ¼ b6 ð1  eH2 Þ c7 ¼ c6 ð1  eH2 Þ þ eH2
a8 ¼ a7 ðgT2  1Þ=gT2 b8 ¼ ½b7 ðgT2  1Þ þ 1=gT2 c8 ¼ c7 ðgT2  1Þ=gT2

The relations for T 8R and T 4R , given respectively by Eqs. (29) and (35), did not need to be modified because these relations
are already functions of T 4 and T 8 . Using the second law of thermodynamics to the working fluid path and considering the
cycle 1-2S-3-4S-4R-5-6S-7-8S-8R-1 the following relation is obtained [10,12,18,21]:
 k1
T3 T5 T7 T1 p3 p5 p7 p1 k
¼ : ð51Þ
T 2S T 4S T 6S T 8S p2 p4 p6 p8
Eq. (51) is simplified substituting Eqs. (20), (21), (25)–(28) and using dimensionless forms for the temperatures:

T 3 T 7 y ¼ T 4S T 8S xqG ; ð52Þ
6836 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

where qG is the global parameter for the pressure drops:


qG ¼ qH1 qL1 qH2 qL2 : ð53Þ
Substituting Eqs. (44), (45), (49) and (50) in Eq. (52) a quadratic equation is obtained:

A1 T 24 þ A2 T 4 þ A3 ¼ 0; ð54Þ
where:

A1 ¼ a4 a8 qG x  a3 a7 y; A2 ¼ a9 T 8 þ a10
;
A3 ¼ a11 T 28 þ a12 T 8 þ a13
and:
a9 ¼ ða4 b8 þ a8 b4 ÞqG x  ða3 b7 þ a7 b3 Þy a10 ¼ ða4 c8 þÞa8 c4 qG x  ða3 c7 þ a7 c3 Þy
a11 ¼ b4 b8 qG x  b3 b7 y a12 ¼ ðb4 c8 þ b8 c4 ÞqG x  ðb3 c7 þ b7 c3 Þy :
a13 ¼ c4 c8 qG x  c3 c7 y

Solving Eq. (54) we arrive at an expression for T 4 as a function only of temperature T 8 :


qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
A2 þ A22  4A1 A3 a9 T 8  a10 þ ða9 T 8 þ a10 Þ  4A1 ða11 T 28 þ a12 T 8 þ a13 Þ
T4 ¼ ¼ : ð55Þ
2A1 2A1
With Eq. (55) all the temperature relations previously obtained can now be described as a function of T 8 only. Using the
first law of thermodynamics the power output is given by the difference between the rates of heat transfer Q_ H and Q_ L ,
provided by Eqs. (23) and (24):
_ ¼ Q_ H  Q_ L ¼ Q_ H1 þ Q_ H2  Q_ L1  Q_ L2 :
W ð56Þ
Substituting Eqs. (1)–(4) into Eq. (56) yields:
_ ¼ C_ p eH1 ðT H  T 4R Þ þ C_ p eH2 ðT H  T 6 Þ  C_ p eL1 ðT 8R  T L Þ  C_ p eL2 ðT 2  T L Þ:
W ð57Þ

Dividing Eq. (57) by C_ p T H and utilizing Eqs. (29), (35), (43) and (48) an equation for the dimensionless power output is
obtained:

W _ C_ p T H ¼ d1 T 4 þ d2 T 8 þ d3 ;
_ ¼ W= ð58Þ
where:
d1 ¼ ðeH1  eL1 ÞeR  eH1  a6 eH2  a2 eL2
d2 ¼ ðeL1  eH1 ÞeR  eL1  b6 eH2  b2 eL2
  :
d3 ¼ eH1 þ eL1 TTHL þ ð1  c6 ÞeH2 þ TTHL  c2 eL2

The entropy generation rate for the cycle is:

Q_ L Q_ H Q_ L1 þ Q_ L2 þ Q_ I Q_ H1 þ Q_ H2 þ Q_ I
S_ g ¼  ¼  : ð59Þ
TL TH TL TH
Eq. (59) can be rewritten as a function only of the temperatures by substituting in it Eqs. (1)–(4) and Eq. (22):

C_ p eL1 ðT 8R  T L Þ þ C_ p eL2 ðT 2  T L Þ þ C_ I ðT H  T L Þ C_ p eH1 ðT H  T 4R Þ þ C_ p eH2 ðT H  T 6 Þ þ C_ I ðT H  T L Þ


S_ g ¼  : ð60Þ
TL TH
Dividing Eq. (60) by C_ p and substituting Eqs. (29), (35), (43) and (48) into Eq. (59) a dimensionless entropy generation rate
is obtained:

S_ g ¼ S_ g =C_ p ¼ d4 T 4 þ d5 T 8 þ d6 ; ð61Þ
where:

d4 ¼ ðeR eL1 þ a2 eF2 Þ TTHL þ ð1  eR ÞeH1 þ a6 eH2


d5 ¼ ½ð1  eR ÞeL1 þ b2 eL2  TTHL þ eR eH1 þ b6 eH2 :
hqffiffiffiffi qffiffiffiffii2  
C_ I TH TL TH
d6 ¼ C_ TL
 T H  eH1 þ eH2 ðc6  1Þ  eL1 þ eL2 c2 T L  1
p
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6837

An equation for the dimensionless ecological function is defined as:

_ C_ p T H ¼ W
E_ ¼ E= _  ðT 0 =T H ÞS_ g : ð62Þ
Replacing Eqs. (58) and (61) in Eq. (62) leads to an equation for the ecological function that will be used to perform the
optimization:

E_ ¼ d1 T 4 þ d2 T 8 þ d3  ðT 0 =T H Þðd4 T 4 þ d5 T 8 þ d6 Þ: ð63Þ
_ given by Eq. (58), and the
The thermal efficiency g can be evaluated by the ratio between the dimensionless power W,
dimensionless heat transfer rate Q_ H , obtained by dividing Eq. (23) by C_ p T H :

_
W _
W
g¼ _ ¼ : ð64Þ
Q H =ðC_ p T H Þ Q_ H

3. Results and discussion

To check the ability of the model to deliver results for more simple cycles and test the validity of the proposed model,
results were compared with the works of other authors through changes in design parameters. These changes eliminate
specific process cycle and the same parameters can be calculated for models where they were not included. To eliminate
the reheating process the following input data are used for: y = 1, qH2 = 1 e NH2 = 0. As for the inter-cooling, the parameters
related to the heat exchanger of this process and the low-pressure compressor were changed: x = 1, qL2 = 1 e NL2 = 0. With
this, both the processes related to the reheating and intercooling as related to the high pressure turbine and the low pressure
compressor are eliminated, which means that the model is reduced to an irreversible Brayton cycle only with regeneration.
The check is then made by comparing the results of this simple cycle with those obtained in the work of [17], that have
studied an irreversible regenerative Brayton cycle. Performing this comparison, the results were equivalent, indicating the
ability of the model to eliminate inter-cooling and reheating processes. To transform the model into an endoreversible
Brayton cycle with regeneration, the irreversibility due to heat loss and non-isentropic behavior in compressor and turbine
must be removed. This is done by setting C_ I =C_ p ¼ 0 and gC2 = gT2 = 1.
The results for the reduced model for an endoreversible Brayton cycle with regeneration showed up in accordance with
those presented in the work of [15]. For the elimination of the regeneration process, it can be used a value which reduces the
cycle to an even simpler model. For cases without regeneration the model was in line when compared to results obtained by
[12], for a Brayton cycle with heat loss and non-isentropic behavior of the turbine and compressor, and the results also agree
with the models developed by [4,11] for an endoreversible Brayton cycle.
In this section the following design parameters were used: g ¼ g ¼ g ¼ g ¼ 0:9; k ¼ 1:4; C_ I =C_ p ¼ 0:02;
C1 C2 T1 T2
x ¼ y ¼ 1:5; qH1 ¼ qH2 ¼ qL1 ¼ qL2 ¼ 0:97; N H1 ¼ N H2 ¼ N L1 ¼ N L2 ¼ N R ¼ 4 and TH/TL = 5. These parameters are utilized to
construct all of the figures, except in cases where there is some variation, in which the same will be appointed in the text
and in the graphics. Furthermore, it is assumed that the environment temperature is equal to temperature TL.
It was observed that the cycle temperatures are given as function of T 8 : Therefore, by varying the value of this tempera-
ture we obtain different results for the configuration of temperatures and so for the power plant performance. From that, the
relation between ecological function E_ and power output W, _ shown in Fig. 2 for different values of TH/TL, can be obtained
varying values for T 8 in Eqs. (58) and (63).
The graph begins and ends at a negative value for the ecological function, meaning that in these points the exergy
destruction rate T 0 S_ g is greater than the power output. However, negative values of the ecological function have no meaning
_ where the ecological function E_ becomes greater than zero. It is also
in reality. This situation changes for higher values of W
observed in Fig. 2 that exists one point of maximum. In this point there is a unique value for T 8 , and the search for this opti-
mum value that will maximize the ecological function E_ is the goal of the first optimization.
The increase in TH/TL rises these maximum values for both W _ and E;
_ increasing the region where E_ is positive. It can also
be noticed that the increase in TH/TL does not increase significantly the distance between the point of maximum ecological
function and the point of maximum power.
Fig. 3 presents the maximum ecological function with T 8 optimized as a function of the pressure ratio rpT1 for different
values of N. The parameter N is related to the numbers of transfer units of each heat exchanger so that
N = NH1 = NH2 = NL1 = NL2 = NR.The values of E_ MAX grow vertiginously to the point of maximum and then decreases almost
as fast as when E_ MAX is increasing. E_ MAX increases considerably with the value of N, indicating an appreciable improvement
in the cycle performance. However, this increase in E_ MAX is most effective when it occurs for low values of N. Values of rpT1
that lead to the optimum point of E_ MAX also increase with N, varying between about 2 and 4.
6838 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

0.2

−0.2

−0.4

−0.6
TH /T L =4
TH /T L =5
TH /T L =6
TH /T L =7
−0.8
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Fig. 2. Ecological function E_ versus power W
_ for different values of TH/TL.

N = 2
0.12 N = 3
N = 4
N = 5
0.1

0.08
E max

0.06

0.04

0.02

0
2 4 6 8 10 12 14 16 18 20
rpT 1
Fig. 3. Maximum ecological function E_ MAX versus pressure ratio rpT1 for different values of N.

Fig. 4 shows the maximum ecological function E_ MAX versus the pressure ratio rpC1 for different values of N. The same con-
clusions can be obtained with respect to the values of N since it leads to higher values of E_ MAX and to higher values of rpC1 that
optimize E_ MAX .
The difference between the optimum pressure ratios is that the values of rpC1 are always greater than the values of rpT1.
This is observed in Fig. 4, where rpC1 begins with values close to 4 for lower N and then approaches 6 for higher values of N.
Both graphics show that the correct choices for the values of pressure ratios are very important to ensure a better cycle
performance and raise even more the ecological function. It is also seen that an arbitrary choice for pressure ratio can lead to
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6839

N = 2
0.12 N = 3
N = 4
N = 5
0.1

0.08
E max

0.06

0.04

0.02

0
2 4 6 8 10 12 14 16 18 20
rpC 1
Fig. 4. Maximum ecological function E_ MAX versus pressure ratio rpC1 for different values of N.

0.4

0.35

0.3

0.25
Ẇ E2

0.2

0.15

0.1

0.05
CRT
CRT T
CCRT
CCRT T
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R
_ E;2 under maximum ecological function versus regenerator effectiveness eR for different cycle modifications.
Fig. 5. Power output W

an operation point where the value of E_ MAX is considerably less than what would be obtained if used optimum values for the
pressure ratios.
Thus, to guarantee a better performance for the power cycle is necessary not only the optimization of the temperature T 8 ,
but also the determination of optimal values for rpC1 and rpT1, which will improve further the results.
These optimization processes are numerically performed through MATLAB using the fminsearch command, which allows
us to find optimal values for one or more variables, which in this case are T 8 ; rpC1 and rpT1. This optimization is designated by
subscript 2, indicating that the cycle is optimized twice.
In Fig. 5 it can be seen the power W _ E;2 under maximum ecological function versus the regenerator effectiveness for four
different cycles. The cycle CRT refers to a simple cycle with compressor, regenerator and turbine. The cycle CRTT has a
6840 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

compressor, a regenerator and a reheating process between two turbines and the cycle CCRT has an inter-cooling process
between two compressors, a regenerator and a turbine. The cycle represented by CCRTT is the most complete and has
two compressors, a regenerator and two turbines. This last cycle has therefore both inter-cooling and reheating processes.
Results for simplest cycles are obtained eliminating reheating and inter-cooling processes through changes in the design
parameters of the heat engine. Thus, the results become the same as that would be obtained if these processes had not been
introduced in the mathematical modeling.
To eliminate in inter-cooling process and also leave the cycle operating with just one compressor the following input data
are used: rpC1 = 1, qL2 = 1 and NL2 = 0. For the elimination of the reheating process and the high pressure turbine the following
values are used: rpT1 = 1, qH2 = 1 and NH2 = 0.
Fig. 5 shows a slightly increase in W _ E;2 from eR equal to zero to eR near the value of 0.5. After that the power starts to
decrease until the value of eR equal to one, where it reaches a value equal to the power when there is no regeneration, in
other words, when eR equal to zero. Thus, eR does not show a considerable influence in W _ E;2 .

Cycle CCRT and CRTT have little difference between them when compared their values for W _ E;2 ; with CRTT presenting a
slightly advantage. But the combination with two compressors and two turbines of the CCRTT clearly shows the benefits of
_ E;2 , caused by the combination of these two modifications,
adding more stages in both processes. This, because the raise in W
is much higher than the increase induced when added just the inter-cooling process in the simple cycle CRT. The same could
be said when is just the reheating process that is added.
Fig. 6 shows the thermal efficiency gE,2 at maximum ecological function versus the regenerator effectiveness eR. It is noted
that gE,2 has a small decrease for low values of eR and then increase rapidly for high values of eR.
CCRTT cycle presents the highest values of gE,2 followed by CCRT, CRTT and CRT, respectively. The addition of the inter-
cooling process has a noticeable advantage with respect to values of gE,2 when compared with the addition of the reheating
process. When eR is close to unity, the values for CRT exceed those for CRTT.
Fig. 7 presents the entropy generation rate S_ gE;2 under maximum ecological function versus the regenerator effectiveness
eR. S_ gE;2 for small values of eR is seen as being greater than for cycles without regeneration when eR is equal to zero. But for
larger values of eR ; S_ gE;2 becomes significantly less than when the cycle presents no regeneration. This occurs when the
effectiveness is greater than about 0.7, which shows the advantage of adding this process for higher values of eR.
The highest values of S_ gE;2 are given by CCRTT cycle followed respectively by CRTT and CCRT. In this graph it becomes clear
the gain brought by CRTT in comparison with CCRT for this type of optimization, since the cycle with inter-cooling presents
greater thermal efficiencies along with lower values for entropy generation rates, which is closer to the values for the CRT
cycle.

0.55

0.5
ηE2

0.45

0.4

CRT
CRT T
CCRT
CCRT T
0.35
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R
Fig. 6. Thermal efficiency gE,2 under maximum ecological function versus regenerator effectiveness eR for different cycle modifications.
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6841

1.6

1.4

1.2
Ṡ gE2

0.8

0.6
CRT
CRT T
CCRT
CCRT T
0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R
Fig. 7. Entropy generation rate S_ gE;2 under maximum ecological function versus regenerator effectiveness eR for different cycle modifications.

0.95

0.94

0.93

0.92
Ẇ E2 Ẇ M AX2

0.91

0.9

0.89

0.88

0.87
N = 3
0.86 N = 4
N = 5
N = 6
0.85
4 5 6 7 8 9 10
TH /T L
_ E;2 =W
Fig. 8. Ratio W _ MAX;2 versus ratio TH/TL for different numbers of transfer units N.

It is also observed that in the ecological optimization the regenerator effectiveness eR has a key role in improving the
results for cycle efficiency and entropy generation rate. The variation of eR does not show a significant influence only when
analyzed the power output for the optimized cycle.
The next graphs relate the results for the ecological optimization with the results for the power output optimization. For
both optimizations are found optimum values for T 8 ; rpC1 and rpT1 . Therefore, the results are indicated by the subscript 2.
_ E;2 =W
Fig. 8 represents the ratio W _ MAX;2 versus the ratio TH/TL. The increase in TH/TL leads to a rise in the ratio W
_ E;2 =W
_ MAX;2
_ E;2 =W
with a significant influence. The increase in N decreases W _ MAX;2 , but its influence is not as great as caused by TH/TL.
6842 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

The graph also indicates that W _ E;2 is very close to W


_ MAX;2 , as the power output at maximum ecological function varies
between 85% to 95% of the maximum power. Thus, the ecological optimization suffers only a small loss in power when
compared to the power optimization.
In Fig. 9 is shown the ratio gE,2/gW,2 versus the ratio TH/TL. It is noted that gE,2/gW,2 is always greater than unity, lying
within the range of 1.07 and 1.15. This already indicates one of the advantages of ecological optimization, since it presents
a remarkable increase in efficiency when compared to power optimization, with values than can be 10% higher than gW,2.

1.16
N = 3
N = 4
1.15 N = 5
N = 6
1.14

1.13
ηE2 ηW2

1.12

1.11

1.1

1.09

1.08

1.07
4 5 6 7 8 9 10
TH /T L
Fig. 9. Ratio gE,2/gW,2 versus ratio TH/TL for different numbers of transfer units N.

0.76

0.74

0.72
Ṡ gE2Ṡ gW2

0.7

0.68

0.66

0.64 N = 3
N = 4
N = 5
N = 6
0.62
4 5 6 7 8 9 10
TH /T L
Fig. 10. Ratio S_ gE;2 =S_ gW;2 versus ratio TH/TL for different numbers of transfer units N.
S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844 6843

The increase in ratio TH/TL leads to a smaller difference between the values for both thermal efficiencies while the increase in
N causes a raise in the ratio gE,2/gW,2.
Fig. 10 illustrates the variation of S_ gE;2 =S_ gW;2 with the ratio TH/TL and it can be seen that the rise in N also leads to smaller
values for S_ gE;2 =S_ gW;2 . Another feature to be noticed is that the increase in TH/TL causes the value of S_ gE;2 =S_ gW;2 grows. The val-
ues for this ratio lies within the range of 0.64 and 0.76, which indicates that ecological optimization leads to an entropy
generation rate considerably lower than for the power optimization.
Thus, besides the advantages of presenting greater thermal efficiencies, ecological optimization also conducts to smaller
entropy generation rates of about 70% of the values obtained when power output is maximized.

4. Conclusions

A mathematical model for an irreversible Brayton cycle adding regeneration, inter-cooling and reheating processes was
developed. These modifications have been added to the simple cycle power can be removed through the use of certain speci-
fic values of the model parameters, to allow results to be obtained also for other types of Brayton cycle. The irreversibility are
related to differences in temperature between the working fluid and the thermal reservoirs, the pressure drop in pipes, the
irreversibility present in compressors and turbines and heat loss to the cold reservoir. The optimization was performed for
two types of criteria and the results were compared. The criteria are the output power and ecological function.
The first optimization was done numerically, seeking the optimum temperatures that maximize analyzed criteria. A sec-
ond optimization is also performed numerically in order to improve the results by finding great pressure ratios for the first
stage of the compression process and the first stage of expansion. Results indicate that the irreversibility due to pressure
drops in heat exchangers and heat loss lead to a considerable depreciation in the cycle performance. However, while the heat
loss only has a negative effect on the thermal efficiency and the entropy generation rate, the losses still lead to a reduction in
output power.
Another interesting feature presented the results is due to the improvement caused by the increase in isentropic efficien-
cies of compressors and turbines. Though both the increase in efficiency of the expansion process and the compression sig-
nificantly raise the performance of the cycle, the increase of this parameter for most turbines have a positive effect on the
results, indicating that the investment on the efficiencies of the turbines is more advantageous that investment in
compressors.
The increase in the values of the effectiveness of the regenerator also showed significant influence on improving results
for the thermal efficiency and entropy generation rate. Furthermore, it is seen that the addition of a turbine in a more simple
Brayton cycle leads to a small advantage in output power than if it was added another compressor. However, the addition of
another compressor in place for one more turbine, causing an increased gain in thermal efficiency and a decrease in entropy
generation rate. And when both compressor and turbine are added, performance has increased even more, especially in
power output.
The results presented for entropy generation rate obtained by optimizing the ecological function, about 70% of the values
for the optimization of power also shows the advantage of using this criterion, since it takes power values near the power
maximum, but with a considerably lower entropy generation rate and closer to the values for maximum efficiency.
Therefore, it can be said that its performance is between the cycle operating at maximum power and efficiency at maximum.
For the results to optimize the ecological function was observed that their values are close to those obtained for maximum
efficiency.
It can be concluded that matching between the two extreme maximum efficiency and maximum power is the maximum
operation point of the ecological function. This point combines low rates of entropy generation, which are optimizing the
characteristics of thermal efficiency, with higher values of output power, aspect to nearby points of maximum power.
These attributes comply with the objective criterion to present a best compromise between high output power and low rate
of entropy generation.
The advantage of the method used in this work is the possibility to analyze various design parameters from different con-
figurations of the Brayton cycle. Basic Brayton cycles can be analyzed and Brayton cycles with regeneration, reheating, inter-
cooling and various stages of compression and expansion. For each analysis cycle, different construction and operation
parameters can be analyzed and optimized, seeking maximum power, maximum thermal efficiency, minimum rate of
entropy generation or maximum ecological function. A combination of these criteria may also be obtained by successive
optimizations. The criterion of ecological function can be viewed as an option for optimizing a Brayton cycle aiming at a
lower output power and also a lower rate of generation of entropy, in contrast to maximize power output, it produces a
greater amount of entropy. As closing, it can be said that the choice of the ideal criteria for any project should be determined
according to the intended objectives, even because each criterion seeks to optimize different aspects of the power cycle. For
future work the following topics can be developed:

- The use of different criteria for optimization of the power cycle of this analysis work such as the specific power ratio
between output power and total heat exchange area and a second law efficiency.
- The use of thermal reservoirs with finite heat capacity (variable temperature) in the proposed model.
6844 S. del Rio Oliveira et al. / Applied Mathematical Modelling 39 (2015) 6830–6844

- The modeling and optimization Brayton cycles with more stages of compression and expansion, which would quantify
the influence on plant performance for each stage of compression and expansion.
- The use of the concepts presented here for modeling and optimizing other thermodynamic cycles, such as internal com-
bustion cycles, such as Otto, Diesel and Dual, steam power cycles, refrigeration cycles and heat pump cycles.

References

[1] F.L. Curzon, B. Ahlborn, Efficiency of a Carnot engine at maximum power output, Am. J. Phys. 43 (1975) 22–24.
[2] H.S. Leff, Thermal efficiency at maximum work output: new results for old heat engines, Am. J. Phys. 55 (1987) 602–610.
[3] A. Bejan, Theory of heat transfer-irreversible power plants, Int. J. Heat Mass Transf. 31 (1988) 1211–1219.
[4] O.M. Ibrahim, S.A. Klein, J.W. Mitchell, Optimum heat power cycles for specified boundary conditions, ASME J. Eng. Gas Turbines Power 113 (1991)
514–521.
[5] C. Wu, R.L. Kiang, Power performance of a nonisentropic Brayton cycle, ASME J. Eng. Gas Turbines Power 113 (1991) 501–504.
[6] J. Chen, The maximum power output and maximum efficiency of an irreversible Carnot heat engine, J. Phys. D Appl. Phys. 27 (1994) 1144–1149.
[7] X. Dan, C. Lingen, S. Fengrui, W. Chih, Universal ecological performance for endoreversible heat engine cycles, Int. J. Ambient Energy 27 (1) (2006) 15–
20.
[8] F. Angulo-Brown, An ecological optimization criterion for finite-time heat engines, J. Appl. Phys. 69 (1991) 7465–7469.
[9] Z. Yan, Comment on ‘‘An ecological optimization criterion for finite-time heat engines’’ [J. Appl. Phys. 69, 7465 (1991)], J. Appl. Phys. 73 (1993) 3583.
[10] C.Y. Cheng, C.K. Chen, The ecological optimization of an irreversible Carnot heat engine, J. Phys. D Appl. Phys. 30 (1997) 1602–1609.
[11] C.Y. Cheng, C.K. Chen, Ecological optimization of an endoreversible Brayton cycle, Energy Convers. Manage. 39 (1998) 33–44.
[12] C.Y. Cheng, C.K. Chen, Ecological optimization of an irreversible Brayton heat engine, J. Phys. D Appl. Phys. 32 (1999) 350–357.
[13] W. Junhua, C. Lingen, G. Yanlin, S. Fengrui, Ecological performance analysis of an endoreversible modified Brayton cycle, Int. J. Sustain. Energ. 33 (3)
(2014) 619–634.
[14] C. Lingen, W. Wenhua, S. Fengrui, Ecological performance optimization for an open-cycle ICR gas turbine power plant. Part 1: Thermodynamic
modelling, J. Energy Inst. 83 (4) (2010) 235–241.
[15] Y. Ust, A. Safa, B. Sahin, Ecological performance analysis of an endoreversible regenerative Brayton heat-engine, Appl. Energy 80 (2005) 247–260.
[16] Y. Ust, B. Sahin, O.S. Sogut, Performance analysis and optimization of an irreversible dual-cycle based on an ecological coefficient of performance
criterion, Appl. Energy 82 (2005) 23–39.
[17] Y. Ust, B. Sahin, A. Kodal, Performance analysis of an irreversible Brayton heat engine based on ecological coefficient of performance criterion, Int. J.
Therm. Sci. 45 (2006) 94–101.
[18] Y. Ust, B. Sahin, A. Kodal, I.H. Akcay, Ecological coefficient of performance analysis and optimization of an irreversible regenerative-Brayton heat
engine, Appl. Energy 83 (2006) 558–572.
[19] B. Sahin, A. Kodal, Y. Hasbi, Efficiency of a Joule–Brayton engine at maximum power density, J. Phys. D Appl. Phys. 28 (1995) 1309–1313.
[20] L. Chen, J. Zheng, F. Sun, C. Wu, Performance comparison of an endoreversible closed variable temperature heat reservoir Brayton cycle under
maximum power density and maximum power conditions, Energy Convers. Manage. 43 (2002) 33–43.
[21] L. Chen, J. Wang, F. Sun, Power density analysis and optimization of an irreversible closed intercooled regenerated Brayton cycle, Math. Comput. Model.
48 (2008) 527–540.
[22] Y. Haseli, Optimization of a regenerative Brayton cycle by maximization of a newly defined second law efficiency, Energy Convers. Manage. 68 (2013)
133–140.
[23] W. Wang, L. Chen, F. Sun, C. Wu, Power optimization of an endoreversible closed intercooled regenerated Brayton cycle, Int. J. Therm. Sci. 44 (2005) 89–
94.
[24] S.K. Tyagi, G.M. Chen, Q. Wang, S.C. Kaushik, Thermodynamic analysis and parametric study of an irreversible regenerative-intercooled-reheat Brayton
cycle, Int. J. Therm. Sci. 45 (2006) 829–840.
[25] S. Sánchez-Orgaz, A. Medina, A.C. Hernández, Thermodynamic model and optimization of a multi-step irreversible Brayton cycle, Energy Convers.
Manage. 51 (2010) 2134–2143.
[26] Y. Zhao, J. Chen, An irreversible heat engine model including three typical thermodynamic cycles and their optimum performance analysis, Int. J.
Therm. Sci. 46 (2007) 605–613.
[27] J.C. Lin, S.S. Hou, Performance analysis of an air-standard Miller cycle with considerations of heat loss as a percentage of fuel’s energy, friction and
variable specific heats of working fluid, Int. J. Therm. Sci. 47 (2008) 182–191.
[28] F. Huijun, C. Lingen, S. Fengrui, Exergoeconomic optimal performance of an irreversible closed Brayton cycle combined cooling, heating and power
plant, Appl. Math. Model. 35 (9) (2011) 4661–4673.