ARTICLE IN PRESS

Energy 33 (2008) 153162

www.elsevier.com/locate/energy

Exergy-based method for analyzing the composition of the electricity

cost generated in gas-red combined cycle plants
Samuel Jose Sarraf Borellia, Silvio de Oliveira Juniorb,
a

Promon Engenharia Ltda., Av. Presidente Juscelino Kubitschek, 1830, Itaim, CEP:04543-900 Sao Paulo/SP, Brazil
b
Environmental and Thermal Engineering Laboratory, Polytechnic School, University of Sao Paulo,
Av. Prof. Luciano Gualberto, 1289, Cidade Universitaria, CEP:05508-900 Sao Paulo/SP, Brazil
Received 18 December 2006

Abstract
The proposed method to analyze the composition of the cost of electricity is based on the energy conversion processes and the
destruction of the exergy through the several thermodynamic processes that comprise a combined cycle power plant. The method uses
thermoeconomics to evaluate and allocate the cost of exergy throughout the processes, considering costs related to inputs and investment
in equipment. Although the concept may be applied to any combined cycle or cogeneration plant, this work develops only the
mathematical modeling for three-pressure heat recovery steam generator (HRSG) congurations and total condensation of the produced
steam. It is possible to study any n  1 plant conguration (n sets of gas turbine and HRSGs associated to one steam turbine generator
and condenser) with the developed model, assuming that every train operates identically and in steady state. The presented model was
conceived from a complex conguration of a real power plant, over which variations may be applied in order to adapt it to a dened
conguration under study [Borelli SJS. Method for the analysis of the composition of electricity costs in combined cycle thermoelectric
power plants. Master in Energy Dissertation, Interdisciplinary Program of Energy, Institute of Eletro-technical and Energy, University
of Sao Paulo, Sao Paulo, Brazil, 2005 (in Portuguese)]. The variations and adaptations include, for instance, use of reheat, supplementary
ring and partial load operation. It is also possible to undertake sensitivity analysis on geometrical equipment parameters.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Thermoeconomics; Exergy; Power plant; Combined cycle; Natural gas; Electricity production cost

1. Introduction
The objective of this work is to present a methodology
that allows calculating the costs of electricity generated in
combined cycle power plants, in a variety of congurations, technologies and operating conditions, relating the
energy cost of the outputs to the thermodynamic processes
that surround them. The efciency of such processes in
transforming the potential of the natural gas as an energy
source into work, later converted into electricity, is
Abbreviations: COND, condenser; CT, combustion turbine; CT, cooling
tower; DEAR, Deaerator; ECON, economizer; EVAP, evaporator; FH,
fuel heater; HP, high pressure; IP, intermediate pressure; LP, low pressure;
RH, reheater; SHT, superheater; ST, steam turbine; WTHT, water heater
Corresponding author. Tel.: +55 1130919667; fax: +55 1130919681.
E-mail addresses: sborelli@terra.com.br (S.J.S. Borelli),
silvio.oliveira@poli.usp.br (S. de Oliveira Junior).
0360-5442/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2007.06.008

analyzed aiming at to reach optimum situations, from the

economic and technical points of view.
As tools for the study of the processes that comprise a
combined cycle power plant, the proposed method
includes, beyond the fundamentals of thermodynamics
and heat transfer, the concept of exergy and its applications
to costing analysis [2]. The additional costs of operation
and maintenance, transmission, taxes and others are not
included in the scope of this study, since they are not
directly related to the energy conversion processes. It can
be argued that operation and maintenance costs hold close
relation with the thermal conversion processes, however it
is extremely difcult to separate and allocate them to each
individual process, thus potentially contaminating the nal
results. The relevance of those costs for the overall
evaluation of the costs of the electricity is recognized,
however their non-energy nature separates them from the

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

154

b
specic exergy
B
exergy rate or ow rate
Binlet,outlet exergy ow rate associated to an inlet or
outlet stream
Bm,n
exergy ow rate associated to an inlet or outlet
of equipment m or n
DB
exergy ow rate difference between the inlet
and outlet of the HRSG

basic premise of this work. The capital investments,

however, are directly related to power generation, given
that the sole purpose of building a power plant is to
ultimately generate a certain amount of electricity for a
number of operation hours. Unlike the previously mentioned operation and maintenance costs, capital investments can be separated and allocated individually
according to each thermodynamic process.
2. Exergy and thermoeconomics
2.1. Introduction
The exergy content of a ow changes during the
occurrence of determined processes, being decreased by
the generation of irreversibilities. Therefore, value can be
attributed to a stream, based on its exergy content. A cost
balance shows the variation in the quality of energy,
attributing to it an exergy-based cost that can be altered
according to the origin of the stream, its interactions
during thermodynamic processes and to the cost of the
equipment in which these processes occur.
In order to develop the cost balances, the systems under
study are divided into control volumes or sub-regions,
dened by inlet and outlet streams. Each control volume is
examined and modeled in such a way that a system of
linear equations is formed and a single solution that
satises it can be obtained. The sub-regions may contain
several equipment and processes inside its boundaries;
however, for this study each sub-region will contain single
equipment and in some cases equipment are divided into
smaller control volumes in order to study the cost
distribution through the processes until the electricity
generation is reached.
For a given sub-region that contains equipment X
(considered adiabatic), with i inlet streams and j outlet
streams, the cost balance is shown in
i1
X
m1

m
m
_
Bm
inlet cinlet W inlet cinlet Z X

exergy ow rate difference, between the inlet

and outlet of the HRSG modules
cinlet,outlet exergy-based cost of inlet or outlet streams
cW
exergy-based cost of work
c
exergy-based cost of exhaust gases
Winlet,outlet power consumed or produced
Z
equipment investment rate

DBi,n

Nomenclature

j1
X

Bnoutlet cnoutlet

n1

W outlet cW
outlet

The exergy-based specic costs of streams are in most of

the cases the variables of the system. The application of

cost balances shown in Eq. (1) for a control volume with n

sub-regions results in
_ n.
_ nn  cn Z
B

(2)

However, the cost balances are dened for each subregion, and so a control volume with n sub-regions results
in a system with n variables but not necessarily n equations.
In order to nd a single solution to Eq. (2), it is necessary
to dene criteria, which establish relation among certain
variables.
The extraction criterion considers that power generation
is the ultimate objective of a turbine [3]. Therefore, all
other capital and irreversibility costs are allocated in the
product (i.e., the electrical/mechanical power), in such a
way that the inlet and outlet streams are assigned the same
specic exergy-based cost. The equality criterion considers,
besides generated power, the stream that leaves the turbine
is also a product; consequently both products share the
same exergy-based specic cost [4,5].
Other criteria will be applied to determined equipment,
with the objective of detailing the composition of the
electricity cost and providing a manner to allocate costs
according to the exergy content of each stream.
Once the criteria are applied, any computational tool can
be used to obtain the solution to the system of linear
equations shown in Eq. (2).
2.2. Method description
The model presented in this work was conceived from a
complex power plant conguration depicted in Fig. 1, over
which several variations can be applied to adapt it to a
determined conguration to be studied.
The variations and adaptations include, for instance,
reheating, supplementary ring and partial load operating.
It is also possible to undertake sensitivity analysis of
equipment geometry.
The rst step is to collect information on the power plant
conguration under study, detailing equipment and its
capacity, transport properties of each process stream, and
the plants consumed and generated power. The data that
dene each stream (i.e. mass owrate, temperature,
pressure, enthalpy, entropy and, consequently, exergy)
may be collected using many different sources. One usual

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

Fig. 1. Power plant model.

155

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

156

source is a computerized process simulator. Nevertheless,

data from the digital control system of operating plants
and information from engineered heat balances also
provide the necessary information.
The next step comprises using the equipment capacity to
estimate costs, using a parametric method, which calculates
an estimated cost based on a reference biome of cost and
capacity [68].
Finally, the system of linear equations is obtained, and
from the solution of Eq. (2), results a vector containing
exergy-based costs for each stream. From this point on,
several plant characteristics may be altered, which will
ultimately recalculate matrix B and vector Z of Eq. (2),
and therefore, resulting in a new exergy-based cost
vector. Improving the power plant performance depends
on dening which values from this vector are to be
optimized and thus seeking alterations that will produce the desired effect. Fig. 2 illustrates the proposed
method.

Start
Reference
Plant

Definition of Power Plant

Configuration

Thermoeconomic
Balance

Process data
acquisition

Process
Data

Equipment
Capacities

Parametric cost
Estimate

Matrix [B]

Vector [Z]

Energy-based Costs

Exhaust gases from the combustion turbine have

monetary value because the steam that turns other turbines
is generated from them. However, in the exhaust stack, the
same gases have no longer use as far as the power plant is
concerned, and for that reason cannot be assigned a value.
Assuming otherwise, which means applying the extraction
criterion, would leave certain costs unallocated and thus
producing an articially lower cost for the generated
electricity [9]. On the other hand, using the extraction
criteria through the heat recovery steam generator (HRSG)
and assigning a zero-value to the exhaust stream,
discharges all costs in the boiler feed water, penalizing
the steam generation in a non-uniform basis [5].
As there is use for the exhaust gases from the combustion
turbines and from the supplementary ring modules and
hence these streams are assigned a cost (and the boiler
stack exhaust must contain a null monetary value), it is
necessary to propose a criterion to distribute costs through
the several modules of the HRSG, taking into consideration the exergy of each stream and its variation.
Consider a HRSG with n components (for instance,
economizer, superheater, vaporizer, reheater), in which
the reduction of the exergy owrate of the combustion
gases when owing through a module i is DBi. The overall
exergy reduction in the HRSG is DB. If it is considered that
the cost reduction of the combustion gases is a linear
function of the exergy owrate reduction in every
component of the HRSG, then one can write for every
module
ci ai ci1 b

(3)

where ci is the specic exergy cost of the combustion gases

at the exit of module i and ci1 is the specic exergy cost of
the combustion gases at the inlet of module i. The proposed
relation does not imply that the combustion gases cost will
have a linear reduction through the HRSG, but only inside
each module.
In order to assure that the variation of the cost in each
component be proportional to the ow exergy variation, ai
is dened as [1]


DBi
ai 1 
(4)
DB
moreover, knowing that at the outlet of the last module the
value (cost) of the gases is zero


DBn
 1 cn1 .
(5)
b
DB

Vector [C]

no

2.3. Cost allocation criteria for the heat recovery steam

generator

Thus, the proposed variation law is shown in





DBi
DBn
 1 cn1 .
ci 1 
ci1
DB
DB

yes
Optimized
End

Fig. 2. Flowchart depicting the proposed method [1].

(6)

This relation provides a useful procedure to determine

the variation of the cost of combustion gases (exergy

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

157

that the installation cost may sometimes compensate a

lower efciency.
A deeper analysis of Table 1 allows the study of the
composition of the average cost as a function of the costs
calculated in each generator. The lower value is related to
the combustion turbine generator for all cases. Hence,
including more combustion turbine in 2  1 or 3  1
congurations results in lower average costs (see Fig. 5).
Table 2 shows results obtained by varying the number of
ABB GT24 machines in the power plant.
By analyzing Table 2, it can be observed that adding
more combustion turbines reduces the average cost.
The installation costs also decreased, indicating a scale
gain. Fig. 6 depicts that the average costs are much
inuenced by the low-pressure steam turbine in the 1  1
conguration. The other congurations show more
equalized costs, however the electricity generation in
this section is the highest considering all steam turbine

transferred and destroyed) in each module of the HRSG.

This procedure is shown in Fig. 3.
3. Results
Table 1 shows results of comparisons made using 1  1
power plants using several combustion turbines technology, as a rst application of the method. The conguration
of the heat recovery boiler is equal for every simulation,
and comprising of three pressures and three modules in
each pressure level, and no reheating. It was allowed to the
simulator [10] to adjust the modules within a range of
geometrical characteristics to obtain best results for each
conguration.
By analyzing Table 1 and Fig. 4, one can notice the
relation between exergetic efciency and the average cost of
electricity. The comparison between a plant based on ABB
GT24 machine and another based on a SWPC 501G shows

Bi-1, ci-1

Bi, ci
Bn 0

n modules

cn = 0

ci =

Bi
B

ci 1 +

Bn
B

cn-1

Fig. 3. Steam cost evaluation in the heat recovery steam generator.

Table 1
1  1 Power plant study
Combustion turbine (CT)
Model
Power (MW)
Exergetic efciency

SWPC 501G
212.7
0.36

ABB GT24
167.6
0.35

SWPC 501FD
168.1
0.35

GE PG7421FA
157.5
0.34

SWPC 501DA
109.9
0.32

GE PG9171E
112.8
0.31

Steam turbine power

HP (MW)
IP (MW)
LP (MW)

29.02
27.69
41.66

23.94
22.54
33.11

23.91
22.97
34.52

24.25
23.14
34.72

14.82
16.27
25.55

15.82
17.12
26.75

Power plant
Costsa (US$/kW)
Power (MW)
Exergetic efciency

324.67
311.11
0.53

314.05
247.21
0.52

336.39
249.47
0.52

348.70
239.63
0.51

385.16
166.49
0.49

382.63
172.52
0.48

Electricity cost
CT (US$/MWh)
HPST (US$/MWh)
IPST (US$/MWh)
LPST (US$/MWh)
Average (US$/MWh)

17.60
41.86
45.17
61.47
28.20

17.18
42.26
45.42
61.50
28.12

17.72
42.68
45.88
62.52
28.90

17.59
42.58
45.82
62.48
29.35

18.23
47.61
50.84
66.45
31.43

18.51
47.07
50.23
66.01
31.64

Installation costs presented are referred to power island only [11].

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

158

Electricity Average Cost (US$/MWh)

32.00
31.50
31.00
30.50
30.00
29.50
29.00
28.50
28.00
27.50
300

340

320

360

400

380

Installation Cost (US$/kW)

Fig. 4. Installation cost study.

28.50
1x1
Electricity Average Cost (US$/MWh)

28.00
27.50
27.00
26.50
26.00
2x1
25.50
3x1
25.00
24.50
200

220

240

260

280

300

320

Installation costs (US$/kW)

Fig. 5. n  1 conguration study.

sections. To solve this problem it is necessary either

to decrease its production or to increase production
in other sections, in such a way that the average cost
will be reduced. Table 3 presents results obtained with
three optimization studies of a 2  1 ABB GT24 power
plant.
The rst study shows results for the same conguration
studied before. In Study 2, reheat modules were added,
along with other high and intermediate pressure modules.
The efciency increase can be observed, but the addition of
a reheat section rearranged the steam and power generation inside, producing new results. The average cost was
reduced; however, the low-pressure section continues to
push its value upwards.

Study 3 was obtained taking the condenser pressure

to a lower value. As a result, the destruction of exergy
in this equipment was reduced, which reduced also
the exergy-based cost of the condensate and, consequently, of the boiler feed water. Hence, the steam
was generated at a lower cost, which pulled the overall costs downwards, especially in the low-pressure
section, which had its power production increased. In this
sense, the combined result is a lower average electricity
cost.
Fig. 7 depicts the effect of the proposed law of variation
over the exergy-based cost of the exhaust gases (see
also Fig. 1). It can be observed that the stack exhaust
carried no monetary value, and the cost is distributed

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

159

to each module according to the exergy it captures.

Hence the high-pressure sections, which capture more of
the exergy of the gas exhaust stream, received larger
portion of costs.
Aiming to demonstrate the applicability of the method
to sensitivity analysis, Figs. 8 and 9 are presented. Both
based on a 1  1 SWPC 501FD power plant.
Fig. 8 shows the response of the electricity exergy-based
cost and efciency to a variation in the condenser pressure.
As a lower condenser pressure implies in larger heat
exchanger areas, pumps and cooling towers, the method

allows the capital costs to vary to capture these changes,

and produce results according to the heat balance
modications. Fig. 9 depicts the behavior of the exergybased cost according to the load condition of the power
plant. Partial load conditions produce naturally less
efcient results, a result which can be captured by the
method when allocating cost according to exergy destruction, thus resulting in higher costs.
The last set of results comprises the study on supplementary ring. The base power plant model for this study
is a 2  1 SPWC 501F, which generates 590 MW using its

Table 2
N  1 congurations study

Table 3
Cycle optimization

Conguration

11

21

31

Study

Combustion turbine (CT)

Power (MW)
Exergetic efciency

167.61
0.35

167.61
0.35

167.61
0.35

Combustion turbine
Power (MW)
Exergetic efciency

167.61
0.35

167.48
0.36

167.48
0.36

Steam turbine power

HP (MW)
IP (MW)
LP (MW)

23.94
22.54
33.11

49.28
44.98
66.05

74.63
67.42
98.99

Steam turbine power

HP (MW)
IP (MW)
LP (MW)

49.28
44.98
66.05

41.15
60.91
73.79

41.15
60.95
79.90

Power plant
Costsa (US$/kW)
Power (MW)
Exergetic efciency

314.05
247.21
0.52

257.17
490.12
0.52

221.92
567.96
0.52

Power plant
Costsa (US$/kW)
Power (MW)
Exergetic efciency

257.17
490.12
0.52

264.18
505.81
0.54

260.98
512.01
0.55

Electricity cost
CT (US$/MWh)
HPST (US$/MWh)
IPST (US$/MWh)
LPST (US$/MWh)
Average (US$/MWh)

17.18
42.26
45.42
61.50
28.12

17.17
38.84
42.69
47.22
25.65

17.17
37.51
41.65
43.11
24.88

Electricity cost
CT (US$/MWh)
HPST (US$/MWh)
IPST (US$/MWh)
LPST (US$/MWh)
Average (US$/MWh)

17.17
38.84
42.69
47.22
25.65

17.28
45.37
30.97
45.73
25.21

17.28
40.69
28.45
36.01
23.41

Installation costs presented are referred to power island only [11].

Installation costs presented are referred to power island only [11].

70
60
Electricity Cost (US$/MWh)

50
40
30
20
10
0
1x1

2x1

High Pressure Steam Turbine

Combustion Turbine

Low Presure Steam Turbine

3x1
Average Cost

Intermediate Pressure Steam Turbine.

Fig. 6. Electricity cost composition.

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

12.00
10.00
8.00
6.00
4.00
2.00

T
St
ac
k

AP

TH
W

EV

LP

PE

O
N

O
N
C
H

IP
E

P
SH
T

VA

LP

IP
E

O
2

H
PE

IP
S

PE RH
VA
P

T1
H

TH

PS
H

IP
S

PS
H

Exergy-Based Cost of Exhaust Gas (US$/MWh)

160

Heat recovery steam generator Modules

Fig. 7. Effect of the combustion gases cost law variation.

Efficiency

30.5

54.8%

30.4

54.6%
54.4%

30.3

54.2%
30.2
54.0%
30.1

Efficiency

Exergy-based cost

Energy Based Cost

53.8%
30

53.6%

29.9
29.8
0.03

53.4%

0.05

0.07

0.09

53.2%
0.11

Condenser Pressure (bar)

Fig. 8. Sensitivity analysis: condenser pressure.

full supplementary ring capacity, that corresponds to its

design condition. The fuel burnt in the heat recovery
boilers is decreased until it is completely turned off, in
which case the power plant produces approximately
500 MW. The installation cost is kept constant throughout
the study in order to more accurately reect the load
variations and investment in excess capacity. Figs. 10
and 11 present the variation of the exergy-based cost of
electricity and according both to power generation and to
plant efciency.
It can be observed that the efciency decreases with the
increase of power generation, which is to be expected since
the supplementary ring causes this effect on the overall

cycle efciency. However, the exergy-based cost of

electricity generated nds a local minimum value, which
can be interpreted as the balance between the loss in terms
of efciency and its compensation in capital cost utilization, since operating the supplementary ring in less than
100% capacity implies using less capacity than the
installed.
4. Discussion
The proposed method proved itself capable of analyzing
the composition of the electricity costs in a variety of
combined cycle power plants congurations. The para-

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

Turbine efficiency

Cycle Efficiency

161

Energy-based cost
40

55.0%

39
38

50.0%

Efficiency

37
36

45.0%

35
34

40.0%

33
32

35.0%

31
30

30.0%
50%

60%

70%

80%

90%

100%

110%

Partial Load Operation

Fig. 9. Sensitivity analysis: partial load operation.

Exergy-based electricty cost (US$/MWh)

26.40
26.20
26.00
25.80
25.60
25.40
25.20
25.00
480

500

520

540

560

580

600

Generated Power (MW)

Fig. 10. Cost variation according to generated power.

metric cost estimation allows the method to vary equipment costs accordingly to the capacity of the equipment
used, providing parameters for a cost-benet analysis.
The application of cost balances showed its benets
in terms of process analysis, allowing a detailed study
of each stream that comprises the power plant. The
proposed law of variation on the exergy-based cost of
the exhaust gases aided to identify the costs, which
affect signicantly the composition of the electricity
average cost.

Another benet of the method is providing data for a

cost-benet analysis, combining thermoeconomics and
parametric cost estimation. Investment decisions may be
based on parameters provided by the method, in terms of
equipment sizing and reheat application. As seen in the
supplementary ring cases, similar analysis may provide
data for peaking plant operation.
Another possible outcome of using the proposed method
is evaluating the impact of emissions-trading schemes in
the cost of electricity. Assuming that a monetary value can

ARTICLE IN PRESS
S.J.S. Borelli, S. de Oliveira Junior / Energy 33 (2008) 153162

162

Exergy-based Electricity Cost (US$/MWh)

26.40
26.20
26.00
25.80
25.60
25.40
25.20
25.00
0.510

0.515

0.520

0.525

0.530

Efficiency
Fig. 11. Cost variation according to efciency.

be assigned to the exhaust stack gases, the solution of the

whole system of equations will then re-allocate costs to
accurately reect the costs of emissions.
Finally, it must be pointed out that the proposed
methodology will be also useful in the determination of
the steam production cost in cogeneration plants.
References
[1] Borelli SJS. Method for the analysis of the composition of electricity
costs in combined cycle thermoelectric power plants. Master in
Energy Dissertation, Interdisciplinary Program of Energy, Institute
of Eletro-technical and Energy, University of Sao Paulo, Sao Paulo,
Brazil, 2005 [in Portuguese].
[2] El-Sayed YM. The thermoeconomics of energy conversions. Amsterdam: Elsevier; 2003.
[3] Bejan A, Tsatsaronis G, Moran M. Thermal design and optimization.
New York: Wiley Interscience; 1996.

[4] Kotas TJ. The exergy method of thermal plant analysis. New York:
Krieger; 1995.
[5] Silva MM. Repowering in electricity generation systems in the
siderurgic industry. Master in Engineering Dissertation, Mechanical
Engineering Faculty, State University of Campinas, Campinas,
Brazil, 2004 [in Portuguese].
[6] Corregiari P, Borelli S. Technical and economical viability analysis of
thermoelectrical plant. Undergraduation conclusion project. Sao Paulo,
Brazil: Polytechnic School, University of Sao Paulo; 2000 [in Portuguese].
[7] Means RS. Mechanical cost data2002. RS Means Company, Inc.,
EUA; 2002.
[8] Boehm RF. Design analysis of thermal systems. New York: Wiley; 1987.
[9] Gomes CM. Thermodynamic, exergetic and thermoeconomic analysis of a thermoelectric powerplant with combined cycle and
cogeneration. Master in Engineering Dissertation, Mechanical
Engineering Faculty, State University of Campinas, Campinas,
Brazil, 2001 [in Portuguese].
[10] Gatecycle, v.5.5.1General Electric Power Systems.
[11] Gas turbine world handbook. 20042005, vol. 24. Pequot Publications, 2005.