You are on page 1of 13

Energy 47 (2012) 271e283

Contents lists available at SciVerse ScienceDirect

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

Thermoeconomic analysis and optimization of an ammoniaewater power/cooling


cogeneration cycle
V. Zare a, *, S.M.S. Mahmoudi a, M. Yari b, M. Amidpour c
a
Faculty of Mechanical Engineering, University of Tabriz, Tabriz, Iran
b
Department of Mechanical Engineering, Faculty of Engineering, University of Mohaghegh Ardabili, Ardabil, Iran
c
Energy System Engineering Department, Mechanical Engineering Faculty, K.N. Toosi University of Technology, Tehran, Iran

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

Article history: The performance of an ammoniaewater power/cooling cogeneration cycle is investigated and optimized
Received 27 February 2012 paying more attention on the economic point of view. Thermodynamic and thermoeconomic models are
Received in revised form developed in order to investigate the thermodynamic performance of the cycle and assess the unit cost of
23 July 2012
products. A parametric study is carried out and the cycle performance is optimized based on the thermal
Accepted 1 September 2012
Available online 2 October 2012
and exergy efficiencies as well as the sum of the unit costs of the system products. The results show that
the sum of the unit cost of the cycle products obtained through thermoeconomic optimization is less
than by around 18.6% and 25.9% compared to the cases when the cycle is optimized from the viewpoints
Keywords:
Ammoniaewater mixture
of first and second laws of thermodynamics, respectively. It is also concluded that for each increase of $3/
Power/cooling cogeneration ton in unit cost of the steam as the heat source, the unit cost of the output power and cooling is increased
Thermoeconomic by around $7.6/GJ and $15e19/GJ, respectively.
Optimization Ó 2012 Elsevier Ltd. All rights reserved.
Genetic algorithm

1. Introduction a combined power/cooling cogeneration cycle with ammoniae


water mixture as working fluid is proposed by Goswami [4].
Nowadays, the necessity for more efficient utilization of energy The cycle is modified and further studied by him and others
resources has urged investigators to pay more attention on design considering the effect of some important parameters on the cycle
and optimization of energy conversion systems. In this respect, if performance [5e8]. Renewable energies such as solar or
thermal systems are optimized from the viewpoint of thermody- geothermal energy as well as low temperature waste heat can be
namics only (optimization for maximum output power, maximum used in these systems to produce power and cooling. Xu et al. [9]
first or second law efficiencies), the total cost of the system would showed that using flat plate or low concentration solar thermal
be high resulting in a high product cost. To overcome the dilemma, collectors in a proposed ammoniaewater power/cooling cycle
thermal systems should be analyzed from the thermoeconomic can reduce the electricity production cost from $3500/kW to less
perspective. Thermoeconomics (exergoeconomics) combines the than $2000/kW. They also reported that for a turbine inlet
concepts of exergy and economics to evaluate and improve the temperature and pressure of 410 K and 30 bar, a value of 23.54%
performance of energy systems [1]. It provides a procedure for is achieved for the first law efficiency which is comparable with
assessment of the unit cost of products of a system in order to the Carnot cycle efficiency (31.7%) for the same source temper-
perform a cost-effective design and operation. ature and a sink temperature of 280 K. Xu et al. concluded that
Higher overall energy conversion efficiency in thermal the efficiency of their ammoniaewater cycle is much higher than
systems has always been the goal of researchers [2]. In this that of a conventional steam Rankine cycle for the same source
challenging scenario, a recent improvement in thermal power and sink temperatures. Colonna and Gabrielli [10] worked on
cycles is the use of binary mixtures as working fluid. Kalina [3] a tri-generation plant in which a gas turbine or an internal
introduced the use of ammoniaewater mixture as the working combustion engine is combined with an ammoniaewater
fluid in the bottoming cycle of a combined power plant. Recently, absorption refrigeration plant through a heat exchanger recov-
ering the waste heat. Takeshita et al. [11] proposed another
advanced tri-generation cycle in which the top, intermediate and
* Corresponding author. Tel.: þ98 411 3392489; fax: þ98 411 3356026.
bottoming cycles were a gas turbine cycle, a steam Rankine cycle
E-mail address: v_zare@tabrizu.ac.ir (V. Zare). and an ammoniaewater cycle, respectively. The bottoming cycle

0360-5442/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.energy.2012.09.002
272 V. Zare et al. / Energy 47 (2012) 271e283

Nomenclature RefHE refrigeration heat exchanger


s specific entropy (kJ kg1 K1)
C_ cost rate ($ h1) SUCP sum of the unit costs of the products
c cost per exergy unit ($ GJ1) Sup superheater
D diameter (m) T temperature ( C or K)
D destruction TEOD thermal efficiency optimal design
E_ exergy rate (kW) Th.V throttling valve
e specific exergy (kJ kg1) U overall heat transfer coefficient (kW m2 K1)
EEOD exergy efficiency optimal design V_ volumetric flow rate (m3 s1)
F fouling factors v vapor
F fuel w weighting coefficients
GA genetic algorithm X ammonia concentration
h specific enthalpy (kJ kg1)/heat transfer coefficient Z investment cost of components ($)
(kW/m2 K) Z_ investment cost rate of components ($ h1)
HE heat exchanger
HS heat source Subscripts and abbreviations
ir interest rate 0 ambient
i inside A absorber
k thermal conductivity (kW/m K) B boiler
l liquid ch chemical
m _ mass flow rate (kg s1) CI capital investment
MOF multi-objective function COD cost optimal design
MOOD multi-objective optimal design CRF capital recovery factor
Nu Nusselt number
o outside Greek symbols
OM operation and maintenance hth thermal efficiency
P pressure (bar) ε exergy efficiency
PR pump pressure ratio s annual plant operation hours
Pr Prandtl number hp pump isentropic efficiency
ph physical ht turbine isentropic efficiency
Q_ heat transfer rate (kW) G mass flow rate per wetted perimeter (kg/m s)
Re Reynolds number d film thickness (m)
R rectifier r density (kg/m3)
RecHE recovery heat exchanger m absolute viscosity (N s/m2)

was a cogeneration one producing both power and cooling. Their cooling output, effective first law and exergy efficiencies. It was
results revealed that the contribution of ammoniaewater cycle to also found that an internal rectification cooling process could
the overall cycle efficiency is about 7%. Vidal et al. [12] applied produce higher efficiencies. The second law efficiency of an
the exergy analysis in order to evaluate the performance of ammoniaewater cogeneration cycle, for boiler temperatures of
Goswami cycle focusing on the effect of variation in some higher than 127  C, is found to be 60% as reported by Goswami
parameters such as the ambient temperature, turbine efficiency et al. [20]. Pouraghaie et al. [21] employed multi-objective
and rectifier temperature. Zheng et al. [13] proposed a cycle in algorithms for optimization of thermodynamic performance of
which the flash tank in the Kalina cycle is replaced by a rectifier the Goswami cycle. In this study only the thermodynamic
to enhance the separation process and obtain a higher concen- objective functions (turbine work, cooling capacity and thermal
tration of ammonia for cooling purposes. A different configura- efficiency) were considered and no attention was paid to the
tion of ammoniaewater cycle was proposed and analyzed by Liu thermoeconomics point of view. Different methodologies for this
and Zhang to produce power and refrigeration [14]. Their results point of view have been developed and used by researchers
showed that the energy consumption in the proposed cogene- during the last 25 years. These methodologies are: Theory of the
ration system was 18.2% less than that in the separate conven- Exergetic Cost (TEC) [22], Specific Exergy Costing approach
tional power and refrigeration systems. Another ammoniaewater (SPECO) [23,24], Thermoeconomic Functional Analysis (TFA)
power/cooling cogeneration system is proposed by Zhang and [25,26], Structural Theory of Thermoeconomics (STT) [27],
Lior [15] which operates in a parallel combined cycle mode with Modified Productive Structure Approach (MOPSA) [28]. In prin-
an ammoniaewater Rankine cycle and an ammonia refrigeration ciple, all cited methodologies can be applied to the analysis of
cycle. Wang et al. [16] proposed a new configuration of any thermal system [29].
ammoniaewater combined power and cooling cycle with an Based on the above explained literature review, it seems that
optimized exergy efficiency of 43.06%. The performance of the there is a lack of knowledge in thermoeconomic analysis of
cycle was improved by them employing an ejector [17]. Recently, ammoniaewater power/cooling cogeneration systems, in partic-
a parametric study of the Goswami cycle has been carried out by ular in evaluating the produced power and cooling unit costs which
Padilla et al. [18] who concluded that for the heat source is one of the ultimate goals of thermal system designers. The
temperatures of 90e170  C, the maximum effective first law and present work is an effort to address this lack of information. For this
exergy efficiencies are 20% and 72%, respectively. For this cycle, purpose, the theory of exergetic cost is applied to the case of
another investigation was carried out by Demirkaya et al. [19] a power/cooling cogeneration cycle in order to pinpoint its
who showed that the cycle can be optimized for net work, optimum design conditions with respect to a given set of decision
V. Zare et al. / Energy 47 (2012) 271e283 273

the superheater where it can be superheated. The vapor is then


expanded to a lower pressure (and temperature) in the turbine
producing power. This low temperature ammoniaewater vapor
(stream 8) passes through the refrigeration heat exchanger and
provides cooling by sensible heat transfer to the chilled water
[12,19]. However, very little phase change is detected for the vapor
passing through the refrigeration heat exchanger (depending on
the conditions the quality of turbine exhaust vapor could increase
by up to 4%).
The following assumptions are made in this work [12,18,19]:

1. The system operates in a steady-state condition.


2. The pressure losses due to frictional effects are neglected.
3. The turbine and the pump have isentropic efficiencies.
4. Changes in kinetic and potential energies are neglected.
5. The input energy to the system is provided by the steam at
a temperature of 130  C. The state of steam is assumed to be
saturated, since superheating the vapor at the system inlet
(state point 19), up to a certain degree, has negligible effect on
the cycle performance as shown in Fig. 3.
6. The chilled fluid in the refrigeration heat exchanger is water
entering at the ambient temperature and pressure.
7. The ammoniaewater solution leaving the absorber is saturated
liquid.

3. Thermodynamic analysis

For thermodynamic analysis, each component of the system is


considered as a control volume, taking into account the heat
transfer, work interaction and inlet and outlet streams. The prin-
Fig. 1. Schematic diagram of the power/cooling cogeneration cycle. ciples of mass and energy conservation as well as the second law of
thermodynamics are applied to the components of the system.
variables. A parametric study is performed to identify the effects of Solution to the equations and to the thermodynamic property
decision variables on the thermal efficiency, exergy efficiency and relations in the EES software simulates the cycle performance. The
the specific cost of the system products. The cycle is then optimized second law of thermodynamics is used to evaluate the cycle
by means of genetic algorithm from the viewpoints of both ther- performance based on the exergy perspective. In the absence of
modynamics and economics using the EES (Engineering Equation magnetic, electrical, nuclear and surface tension effects, and
Solver) software [30]. The objectives from thermodynamic and ignoring the kinetic and potential exergies, the total exergy of
thermoeconomic points of view are to maximize the thermal and a stream becomes the sum of physical and chemical exergies [2]:
exergy efficiencies and to minimize the sum of the unit costs of the
system products, respectively. E_ ¼ E_ ph þ E_ ch (1)

The former can be obtained from the following relation [2]:


2. System description and assumptions

A schematic overview of the considered power/cooling cogen-


eration cycle is shown in Fig. 1. The cycle which is referred to as
Goswami cycle in the literature [19], was designed to produce
power and cooling as main and side products, respectively.
As the figure shows, saturated ammoniaewater solution
(stream 1) leaves the absorber and is pumped to a higher pressure
before being split into two streams; stream 2A and stream 2B. The
stream 2A is preheated in the rectifier and the stream 2B passes to
the heat exchanger where it recovers the energy of weak solution
(with lower ammonia concentration) coming from the boiler
(stream 10). The condensed solution exiting the rectifier (stream 5)
is mixed with two other streams, one coming from the recovery
heat exchanger (stream 14) and the other from the rectifier (stream
13), in the mixer and then is fed to the boiler. The mixture stream
receives heat in the boiler, from the heat source and is then sepa-
rated into two streams; weak (stream 10) and strong (stream 4)
solution streams. The former is passed through the heat exchanger
as mentioned before and then throttled back to the absorber. The Fig. 2. Validation of the model developed in the present work by that of Padilla
vapor exiting the boiler is purified in the rectifier before passing to et al. [18].
274 V. Zare et al. / Energy 47 (2012) 271e283

of Ref. [18] and the average relative difference is about 4%. This
difference is due to the different sources of thermodynamic prop-
erties for ammoniaewater mixture used in two works. The ther-
modynamic properties in Ref. [18] are calculated using the
predictive SoaveeRedlicheKwong equation of state and latent-heat
H model employing the Chemcad software while the properties in
the present work, are taken from the EES software where the
correlations are available from [34].

4. Thermoeconomic analysis

The aim of thermoeconomic analysis is to reveal the cost


formation processes and calculate the cost per exergy unit of the
product streams of the system. The unit exergetic cost of the
products obtained from this procedure is used for economic opti-
mization of the cycle. In order to calculate the unit cost of each
Fig. 3. Effect of the steam superheating on the cycle performance.
exergy stream, a cost balance equation along with the required
auxiliary equations are applied to each component of the cycle. For
a system component receiving thermal energy and generating
E_ ph ¼ m½ðh
_  h0 Þ  T0 ðs  s0 Þ (2) power, the cost balance may be written as [2]:

X X
The latter for the ammoniaewater mixture can be calculated as C_ out;k þ C_ w;k ¼ C_ in;k þ C_ q;k þ Z_ k (9)
[31,32]:
     where:
X 1X 0
E_ ch ¼ m
_ e0ch;NH3 þ ech;H2 O (3)
MNH3 MH2 O C_ ¼ cE_ (10)
where, e0ch;NH3 and e0ch;H2 O are the standard chemical exergies of And c is the cost per unit of each exergy stream. The terms C_ w;k
ammonia and water as given by Ahrendts [33]. and C_ q;k are the cost rates associated with the output power from
In order to evaluate the performance of the cycle from the first the component and input thermal energy to the component,
law perspective, the cascade analogy suggested by Vijayaraghavan respectively. Eq. (9) states that the total cost rate of exiting exergy
and Goswami [8] is used. Accordingly, the thermal (first law) effi- streams equals the total cost rate of entering exergy streams plus
ciency for the cycle can be expressed as [18,19]: the total expenditure rate to accomplish the process.
The term Z_ k in Eq. (9) is the total cost rate associated with capital
_ net þ DE_
W cooling =hII;ref investment and operation and maintenance for the kth component:
hth ¼ (4)
Q_ in CI OM
Z_ k ¼ Z_ k þ Z_ k (11)
In this equation, W _ net , DE_ _
cooling and Q in are the net produced
power, the exergy associated with refrigeration and the heat input The annual levelized capital investment for the kth component
to the cycle (via the boiler and the superheater), respectively. Also can be calculated as [2,31]:
the hII,ref term is the second law efficiency of refrigeration which is  
CI CRF
assumed to be 30% in the present work [18,19].
Z_ k ¼ Zk (12)
The second law or exergy efficiency of the cycle can be defined s
as [12]:
where CRF and s are the capital recovery factor and the annual plant
_ net þ E_
W _ net þ DE_
W operation hours, respectively. The capital recovery factor is a func-
cooling cooling
ε ¼ ¼ (5) tion of the interest rate; ir and the number of useful years of the
_E DE_ HS þ DE_ A
in plant operation; n [2]:
Here, DE_ cooling , DE_ HS and DE_ A are the changes in the exergies of
the chilled water in the refrigeration heat exchanger, the steam as ir ð1 þ ir Þn
CRF ¼ (13)
the heat source and the cooling water in the absorber, respectively. ð1 þ ir Þn 1
These are given as:
The calculation of Zk for each component of the system is given
in the Appendix A. The annual levelized operation and mainte-
DE_ cooling ¼ E_ 16  E_ 15 ¼ m
_ 15 ½ðh16  h15 Þ  T0 ðs16  s15 Þ (6)
nance cost for the kth component is calculated as [2,31]:

DE_ HS ¼ E_ 19  E_ 21 ¼ m
OM
_ 19 ½ðh19  h21 Þ  T0 ðs19  s21 Þ (7) Z_ k ¼ gk Zk þ uk E_ P;k þ R_ k (14)

where, gk and uk account for the fixed and variable operation and
DE_ A ¼ E_ 17  E_ 18 ¼ m
_ 17 ½ðh17  h18 Þ  T0 ðs17  s18 Þ (8)
maintenance costs, respectively, associated with the kth compo-
For validation purposes, the net work output of the cycle is nent and R_ k includes all the other operation and maintenance costs
compared with the available data in literature reported by Padilla which are independent of investment cost and product exergy.
et al. [18]. Fig. 2 shows this comparison which indicates a good Since the last two terms on the right side of the equation are small
agreement between the results. However, at all the pressure ratios, compared to the first one, these terms may be neglected [2] as is
the obtained results for the present work is slightly less than those done by some other researchers [29,31,35].
V. Zare et al. / Energy 47 (2012) 271e283 275

The formulation of cost balance and required auxiliary equa-


C_ 11 C_
tions for each component of the cycle leads to the following system ¼ 12 or c11 ¼ c12 (29)
_E E_ 12
of equations: 11
Boiler: Turbine:

C_ 8 þ C_ 22 ¼ C_ 7 þ Z_ t (30)
C_ 4 þ C_ 10 þ C_ 21 ¼ C_ 3 þ C_ 20 þ Z_ B (15)

C_ 7 C_
C_ 4  C_ 3 C_  C_ 3 ¼ 8 or c7 ¼ c8 (31)
¼ 10 (16) E_ 7 E_ 8
_E  E_ E_ 10  E_ 3
4 3
Pump:

C_ 20 C_ C_ 2 ¼ C_ 23 þ C_ 1 þ Z_ p (32)
¼ 21 or c20 ¼ c21 (17)
E_ 20 E_ 21
Mixer:
Rectifier:
C_ 3 ¼ C_ 5 þ C_ 13 þ C_ 14 þ Z_ M (33)
C_ 5 þ C_ 6 þ C_ 13 ¼ C_ 4 þ C_ 2A þ Z_ R (18)
Since the cost of the mixer is small compared to the other
component costs, it can be neglected [29].
C_ 6  C_ 4 C_  C_ 4 The following supplementary equations are necessary to
¼ 5 (19)
E_ 6  E_ 4 E_ 5  E_ 4 complete the economic analysis of the cycle:

C_ 2 ¼ C_ 2A þ C_ 2B (34)
C_ 13 C_
¼ 2A or c13 ¼ c2A (20)
_E E_ 2A
13
C_ 2A C_
¼ 2B or c2A ¼ c2B (35)
Superheater: E_ 2A E_ 2B

C_ 7 þ C_ 20 ¼ C_ 19 þ C_ 6 þ Z_ Sup (21) C_ 23 C_
¼ 22 or c22 ¼ c23 (36)
_E E_ 22
23
C_ 20 C_ The linear system of Eqs. (15)e(36) include 25 unknown vari-
¼ 19 or c20 ¼ c19 (22)
E_ 20 E_ 19 ables; ½X ¼ fC_ 1 ; C_ 2 ; C_ 2A ; C_ 2B ; C_ 3 ; .; C_ 23 g. Assuming a known value
for the unit exergetic cost of the heat source (c19) and considering
Recovery heat exchanger:
the fact that the unit exergetic cost of the cooling water can be
neglected [37], i.e.: c15 ¼ 0 and c17 ¼ 0, then it is possible to obtain
C_ 14 þ C_ 11 ¼ C_ 10 þ C_ 2B þ Z_ RecHE (23) the unit exergetic cost of all exergy streams of the system by solving
the system of 25 equations and 25 unknowns.
C_ 10 C_
¼ 11 or c10 ¼ c11 (24)
_E E_ 11 5. Results and discussion
10

Refrigeration heat exchanger: A parametric study is carried out in order to investigate the
effect of decision variables, namely, the pump pressure ratio (PR),
C_ 9 þ C_ 16 ¼ C_ 8 þ C_ 15 þ Z_ RefHE (25) the ammonia concentration at the rectifier exit (X6), the degree of
ammonia superheat in the superheater (DTSup) and the minimum
temperature difference in heat exchangers (DTpinch), on the values
C_ 8 C_ of the objectives; thermal and exergy efficiencies and sum of the
¼ 9 or c8 ¼ c9 (26)
_E E_ 9 unit cost of the products. The first three decision variables are the
8
same as those assumed by Padilla et al. [18] for parametric study of
Absorber and throttling valve:
the considered cycle from the thermodynamic perspective.
Since the throttling valve is a component for which a product
However, the minimum temperature difference in heat exchangers
cannot be readily defined, this component and the absorber which
is an important factor as an increase in this parameter decreases the
is served by the throttling valve are considered as a single
heat exchanger cost and increases the irreversibilities.
component [2,31,35]. Therefore, the cost balance equation may be
The basic assumptions and input parameters used in the study
written as follows:
are given in Table 1. As mentioned in Section 2, the input energy to
the system is provided by means of saturated steam. The selection
C_ 1 þ C_ 18 ¼ C_ 9 þ C_ 11 þ C_ 17 þ Z_ A þ Z_ Th:V (27)
of saturated state for the steam is justified by considering the
As the contribution of the expansion valve to the total system results in Fig. 3 which reveals that superheating the steam at the
cost is rather small, its cost is usually neglected [36,37]. For the system inlet (stream 19) has a negligible negative effect on the cycle
absorber and throttling valve the following auxiliary equations can performance. The absorber temperature is assumed to be 10  C in
be developed [31,35]: order to evaluate the maximum theoretical performance of the
cycle. This absorber temperature is also beneficial for comparing
C_ 1 C_ þ C_ 12 our results with those previously published in literature [9,18].
¼ 9 (28) However, since a 10  C absorber temperature may not be practical
E_ 1 E_ 9 þ E_ 12
in everywhere and all the times [18], the optimization procedure is
276 V. Zare et al. / Energy 47 (2012) 271e283

Table 1
The input data assumed in the simulation.

Parameters Value
To ( C) 25
Po (bar) 1
PR 6e18
X6 ðkgNH3 =kgsolution Þ 0.965e0.995
DTSup ( C) 0e15
DTpinch ( C) 5e15
T1 ( C) 10a
X1 ðkgNH3 =kgsolution Þ 0.524
T19 ( C) 130a
m_ 19 ðkg=sÞ 1
hp (%) 85
ht (%) 85
s (h/year) 7000
ir (%) 12
n (year) 20
a
Ref. [18].
Fig. 4. Effect of pressure ratio on the thermal and exergy efficiencies.

also performed for an absorber temperature of 30  C which is


a more practical value. temperature difference in the heat exchangers. At all values of the
The unit cost of heat source (unit steam cost) is an important DTpinch the thermal efficiency decreases as ammonia concentration
factor in determining the cost of the products. Since the unit steam increases. This is mainly due to the fact that, for a given pressure,
cost may vary in a wide range, different values of it are considered increasing X6 results in a lower rectifier exit (and hence turbine
in the thermoeconomic optimization. However, for the base case inlet) temperature which leads to a reduction in net power output.
and parametric study, the value of $9.6/ton is assumed [38]. However, increasing X6 results in a slight increase of DE_ cooling which
Table 2 summarizes the calculated thermodynamic properties is very low as compared with the decrease of W _ net . Fig. 6 also
along with the cost flow rates and unit costs at various state points indicates lower efficiency values with higher minimum tempera-
of the system for the base case operating conditions. ture difference in the heat exchangers, due to the lower power
Fig. 4 shows the effect of pump pressure ratio on the thermal output at higher DTpinch.
and exergy efficiencies of the cycle. The figure indicates that there is Fig. 7 shows the variation of exergy efficiency with ammonia
an optimum value of pressure ratio at which the efficiencies are concentration at rectifier exit for different minimum temperature
maximized. difference in the heat exchangers. As the figure indicates, increasing
The variation of the sum of the unit cost of the system products the ammonia concentration results in a decrease of exergy effi-
(SUCP) with pressure ratio is shown in Fig. 5. The figure depicts that ciency. This is due to the fact that, at higher X6 the net power output
the SUCP can be minimized at a specific value of pump pressure of the cycle is lower while DE_ A is higher. The figure also shows that
ratio. with increasing DTpinch, the exergy efficiency decreases as expected.
The effect of ammonia concentration at rectifier exit on the The influence of ammonia concentration at rectifier exit on the
thermal efficiency is shown in Fig. 6 at different minimum SUCP at different values of minimum temperature difference in the

Table 2
Thermodynamic properties and costs of the streams for the base case.a

Stream Temperature ( C) Pressure (bar) Mass flow rate (kg/s) Ammonia concentration (%) Exergy rate (kW) Costs
_
Cð$=hÞ c ($/GJ)
1 10 1.998 4.353 0.524 45,271 3918 24.040
2 10.17 19.98 4.353 0.524 45,281 3919 24.043
3 93.95 19.98 4.531 0.5227 47,215 4096 24.096
4 125 19.98 1.4 0.9176 26,048 2255 24.047
5 94.97 19.98 0.1787 0.4909 1745 158.5 25.221
6 94.97 19.98 1.221 0.98 24,204 2096 24.054
7 94.97 19.98 1.221 0.98 24,204 2096 24.054
8 4.56 1.998 1.221 0.98 23,815 2062 24.054
9 20 1.998 1.221 0.98 23,807 2062 24.054
10 125 19.98 3.131 0.3461 21,653 1873 24.025
11 15.17 19.98 3.131 0.3461 21,460 1856 24.025
12 15.58 1.998 3.131 0.3461 21,453 1855 24.025
13 54.32 19.98 2.176 0.524 22,650 1960 24.043
14 109.5 19.98 2.176 0.524 22,850 1977 24.032
15 25 1 2.973 e 0 0 0.000
16 15 1 2.973 e 2.134 0.8461 110.147
17 5 1 94.68 e 283.2 0 0.000
18 10 1 94.68 e 159.3 1.225 2.136
19 130 2.7 1 e 629.8 34.55 15.240
20 130 2.7 1 e 629.6 34.54 15.240
21 130 2.7 1 e 63.6 3.489 15.240
22 e e e e 324.8 33.72 28.831
23 e e e e 11.13 1.156 28.831
a
PR ¼ 10, X6 ¼ 0.98, DTSup ¼ 0, DTpinch ¼ 5  C.
V. Zare et al. / Energy 47 (2012) 271e283 277

Fig. 5. Effect of pressure ratio on the sum of the unit costs of the system products. Fig. 7. Effect of ammonia concentration at rectifier exit and DTpinch on exergy
efficiency.
heat exchangers is presented in Fig. 8. The figure reveals that the cycle
can be optimized for the minimum product costs with respect to X6.
considered as optimization objectives while the pump pressure
Another implication of this figure is the increase of the SUCP with
ratio, the ammonia concentration at the rectifier exit, the minimum
increasing DTpinch. This is due to the fact that, increasing DTpinch leads
temperature difference in heat exchangers and the degree of
to a decrease of output cooling and turbine power production.
ammonia superheat at turbine inlet are considered as decision
The effect of vapor superheating degree at turbine inlet on the
variables.
thermal and exergy efficiencies of the cycle is presented in Fig. 9.
The figure shows that as the DTSup increases, both the thermal and
exergy efficiencies are increased. These are mainly due to the 6.1. Optimization method
increase in power output with increasing DTSup as depicted in
Fig. 10. This figure also indicates that the decrease of DE_ A is lower Several optimization methods are available in the EES software.
than the increase of W _ net , so that the overall effect is the We applied the genetic algorithm for optimization purposes since it
enhancement of exergy efficiency as Fig. 9 shows. Another impli- is the most robust; nevertheless, it is the slowest of the available
cation of Fig. 10 is a considerable reduction in output cooling which methods. Also, unlike the direct search and variable metric
is due to an increase of T7 (and hence T8) as DTSup is raised. methods, the genetic method is not affected by the guess values of
Another factor which affects the SUCP is the degree of superheat the independent variables [39]. Moreover, in some researches, it is
at turbine inlet. Fig. 11 reveals this effect. It indicates that an evident that the GA is more effective than the conventional
optimum value of DTSup can be found at which the SUCP is mathematical approach [40] and direct search method [41].
minimized. Baghernejad and Yaghoubi [40] showed that, using genetic algo-
rithm, the minimum value of the electricity cost is found to be
5.35% lower than the corresponding value obtained through using
6. Thermoeconomic optimization the conventional mathematical approach. However, in the present
work, it was concluded that by assigning appropriate guess values
Three output parameters, namely, thermal efficiency, exergy for the parameters, the GA and direct search methods yield the
efficiency and the sum of the unit costs of the system products are same optimization results.

Fig. 6. Effect of ammonia concentration at rectifier exit and DTpinch on thermal Fig. 8. Effect of ammonia concentration at rectifier exit and DTpinch on the sum of the
efficiency. unit costs of the system products.
278 V. Zare et al. / Energy 47 (2012) 271e283

Fig. 9. Effect of degree of superheat on the thermal and exergy efficiencies.


Fig. 11. Effect of degree of superheat on the sum of the unit costs of the system
products.
The genetic algorithm is a stochastic global search method that
simulates natural biological evolution. Based on the Darwinian
survival of fittest principle, the genetic algorithm operates on than this does not result in an improvement of the optimal solution.
a population of potential solutions to produce better and better This upper bound is determined after some test runs.
approximations to the optimal solution. The genetic algorithm
differs from more traditional optimization techniques because it 6.2. Results of single-objective optimization
involves a search from a population of solutions and not from
a single point [16]. The genetic algorithm implemented in EES The optimization in this section is performed for the maximum
software is derived from the public domain PIKAIA optimization thermal efficiency, the maximum exergy efficiency and the
program. PIKAIA is a general purpose optimization subroutine minimum SUCP separately, considering the restrictions as follows:
based on the genetic algorithm written in FORTRAN 77. The details  
of program are provided in Refs. [42,43]. For the present study, the Optimize hth or ε or SUCP PR; X6 ; DTSup ; DTpinch (37)
parameters of GA in the optimization procedure are given in Table 3.
The first three parameters in Table 3 are the most responsible Subject to:
ones to identify the optimum solution and can be specified by the
8  PR  18
EES user. Other parameters and functions of the genetic algorithm 0:965  X6  0:995
(e.g. the crossover and mutation functions, minimum mutation rate (38)
0  DTSup ð CÞ  15
and etc.) have been set to the default values suggested in the PIKAIA 5  DTpinch ð CÞ  15
program and are not changeable within the EES [39,43]. The PIKAIA
subroutine incorporates stochastic sampling mechanism based on Table 4 gives the values of decision variables and objective
the roulette wheel algorithm, reproduction plan based on the full functions for the thermal efficiency optimal (TEOD), exergy effi-
generational replacement, one-point crossover operator and one- ciency optimal (EEOD) and cost optimal (COD) designs. For
point mutation operator which allows the mutation rate to vary comparison purposes, the values for the base case are also pre-
dynamically in the course of the evolutionary run. The optimization sented in this Table.
algorithm stops when the given number of generations is reached. As the Table indicates, the cycle optimization leads to a consid-
The number of generations is selected to be 64 as values higher erable enhancement in the cycle performance from the thermo-
dynamic and economic viewpoints. It is seen from the Table that,
the thermal and exergy efficiencies for the TEOD and EEOD cases
are around 12.7% and 14.2% higher than those for the base case.
However, the SUCP for the COD case is around 14.5% lower than that
for the base case. Also the SUCP associated with the thermody-
namically optimal designs is considerably higher than that for the
cost optimal design. The Table indicates that, the SUCP is reduced
by around 18.6% and 25.9% when the optimization is based on the
cost of the products instead of the thermal and exergy efficiencies,
respectively. In addition, it is observed from Table 4 that the net
power output in the COD case is comparable to those of the TEOD
and EEOD cases. Moreover, the cooling output in the COD case is
around 2 times higher than those in the TEOD and EEOD cases. The
significant reduction in the output cooling for the TEOD and EEOD
cases is mainly due to the fact that a higher degree of superheat at
the turbine inlet results in an increase of turbine exit temperature.
The higher turbine exit temperature brings about lower cooling
output as the RefHE exit temperature (T9) is constant.
The COD case is a promising reference in designing energy
Fig. 10. Effect of degree of superheat on Q_ cooling , W
_ net and DE_ .
A conversion systems as cost effectiveness is the most important
V. Zare et al. / Energy 47 (2012) 271e283 279

Table 3 Table 6
GA parameters. Optimum values of the decision variables and objective functions for an absorber
temperature of 30  C.
Parameter Value
Number of individuals in the population 32 Decision variable/objective function Optimal cases
Number of generations 64 TEOD case EEOD case COD case
Maximum mutation rate 0.25
Pump pressure ratio 10.07 9.96 10.50
Initial mutation rate 0.005
Ammonia concentration at the rectifier exit 0.965 0.965 0.965
Minimum mutation rate 0.0005
Minimum temperature difference in HEs ( C) 5 5 5
Crossover probability 0.85
Degree of superheat ( C) 15 15 0
Net output power (kW) 283.3 277.3 260.7
Output cooling (kW) 2.136 1.039 39.77
Thermal efficiency (%) 13.06 12.75 11.89
Table 4
Exergy efficiency (%) 46.34 46.44 43.69
Optimum values of the decision variables and objective functions.
SUCP ($/GJ) 235.0 330.9 160.6
Decision variable/objective function Base case Optimal cases

TEOD case EEOD case COD case


Pump pressure ratio 10 14.93 14.28 14.51
Ammonia concentration 0.980 0.9650 0.9650 0.9651 largest contribution on exergy destruction and investment cost.
at the rectifier exit The second highest value of exergy destruction rate belongs to the
Minimum temperature 5 5 5 5
recovery heat exchanger. Also it can be seen that, for all the three
difference in HEs ( C)
Degree of superheat ( C) 0 15 15 0.0 optimal cases, the highest cost of exergy destruction corresponds to
Net output power (kW) 313.7 358.1 355.9 349.6 the recovery heat exchanger due to its relatively high exergy
Output cooling (kW) 124.4 41.42 36.86 78.82 destruction and cost of the fuel. For the EEOD case the investment
Power to cooling output ratio 2.52 8.65 9.66 4.44 cost associated with the pump is the lowest one as the pressure
Thermal efficiency (%) 14.76 16.63 16.37 16.29
ratio in this case is the lowest one. For the overall system, Table 5
Exergy efficiency (%) 45.76 52.14 52.27 51.24
SUCP ($/GJ) 139.0 145.9 160.4 118.8 indicates the lowest total exergy destruction and investment cost
rates are associated with the EEOD and COD cases, respectively, as
expected.
In order to analyze the effects of some other operating condi-
aspect of these systems [2]. The results in Table 4 indicate that the tions on the performance of the cycle, the optimization is also
considered cycle, at the conditions specified in Table 1, is a cost carried out for different absorber and heat source temperatures as
effective option if the required power to cooling ratio is around well as different turbine efficiencies. The optimization procedure is
4.44. Therefore, for other required power to cooling ratios care repeated for the following cases:
must be taken in designing the Goswami cycle. However, in Section
6.3, some weighting coefficients are considered for the three a) Absorber temperature of 30  C
objectives and a multi-objective optimization is performed. The b) Different turbine efficiencies
results show that, considering different weighting coefficients for c) Different heat source temperatures
the objectives bring about different optimum power to cooling
ratios.
The contribution of each system component in exergy destruc- 6.2.1. Optimization for an absorber temperature of 30  C
tion rate ðE_ D;k Þ, cost of exergy destruction ðC_ D;k Þ and investment In order to keep the state of solution at the absorber exit as
cost is summarized in Table 5 for the TEOD, EEOD and COD cases. saturated liquid, changing the absorber temperature from 10  C to
The cost of exergy destruction for component k is calculated as 30  C requires a change of ammonia mass fraction at the absorber
follows [2,31]: exit from 0.524 to 0.387. Table 6 gives the optimum values of the
decision variables and objective functions for T1 ¼ 30  C. The results
C_ D;k ¼ cF;k E_ D;k (39) indicate that the COD case, with the lowest SUCP, is the best choice
where, cF,k and E_ D;k are the average unit cost of the fuel supplied to among the three cases as the power output in this case is compa-
the component and the exergy destruction rate within the rable with those in the TEOD and EEOD cases and that, the output
component, respectively. cooling is much higher than those in the TEOD and EEOD cases. The
The results shown in Table 5 indicate that, for all the three higher SUCP in EEOD case is due to the very low value of cooling
optimal cases, the assembly of absorber and throttling valve has the output for this case.

Table 5
Exergoeconomic factors for three optimal cases.

Components TEOD case EEOD case COD case

E_ D;k ðkWÞ C_ D;k ð$=hÞ Z_ k ð$=hÞ E_ D;k ðkWÞ C_ D;k ð$=hÞ Z_ k ð$=hÞ E_ D;k ðkWÞ C_ D;k ð$=hÞ Z_ k ð$=hÞ
Rectifier 25.36 2.114 0.0950 23.5 1.997 0.0907 27.52 2.322 0.107
Boiler 51.17 2.770 1.030 52.37 2.873 1.027 56.76 3.114 1.033
Mixer 15.56 0.00467 0 15.35 0.00466 0 9.908 0.00297 0
Superheater 0.0577 0.003121 0.199 0.056 0.00308 0.186 0.0 0.0 0.0
RecHE 73.89 6.165 0.659 72.36 6.094 0.641 73.61 6.215 0.597
Turbine 58.22 4.842 0.0154 57.51 4.829 0.0154 71.11 5.990 0.0155
Absorber and Th.V 86.27 0.528 1.535 86.02 0.557 1.54 90.09 0.508 1.562
RefHE 0.462 0.0384 0.208 0.348 0.0296 0.222 1.780 0.151 0.186
Pump 4.298 0.425 0.0515 4.041 0.403 0.0502 3.706 0.372 0.0512
Overall system 315.3 16.89 3.793 311.6 16.79 3.772 334.5 18.67 3.552
280 V. Zare et al. / Energy 47 (2012) 271e283

Table 7 Table 8
Optimum values of the decision variables and objective functions for different Optimum values of the decision variables and objective functions for different heat
turbine efficiencies. source temperatures.

Decision variable/ Optimal cases for ht ¼ 1 Optimal cases for Decision variable/ Optimal cases for Optimal cases for
objective function ht ¼ 0.773 objective function T19 ¼ 120  C T19 ¼ 140  C

TEOD EEOD COD TEOD EEOD COD TEOD EEOD COD TEOD EEOD COD
case case case case case case case case case case case case
Pump pressure ratio 14.84 14.43 13.03 14.41 14.11 14.15 Pump pressure ratio 12.95 12.20 11.95 16.81 16.46 17.34
Ammonia concentration 0.965 0.965 0.965 0.965 0.965 0.971 Ammonia concentration 0.965 0.965 0.976 0.965 0.965 0.968
at the rectifier exit at the rectifier exit
Minimum temperature 5 5 5 5 5 5 Minimum temperature 5 5 5 5 5 5
difference in HEs ( C) difference in HEs ( C)
Degree of superheat ( C) 15 15 0 15 15 0 Degree of superheat ( C) 15 15 15 15 15 15
Net output power (kW) 426.1 423.3 418.6 323.4 321.4 311.4 Net output power (kW) 345.5 343.1 336.0 367.6 365.9 362.1
Output cooling (kW) 108.6 105.1 92.02 2.874 0.841 77.9 Output cooling (kW) 25.64 19.03 75.59 54.56 52.29 76.13
Thermal efficiency (%) 19.36 19.20 18.62 14.91 14.78 14.14 Thermal efficiency (%) 15.66 15.53 15.02 17.04 16.92 16.69
Exergy efficiency (%) 62.18 62.33 61.66 47.01 47.14 45.50 Exergy efficiency (%) 52.70 52.87 51.62 51.42 51.58 50.97
SUCP ($/GJ) 116.5 115.5 114.4 238.5 342.7 124.4 SUCP ($/GJ) 249.8 235.3 121.0 130.7 131.9 116.7

TEOD and EEOD cases at T19 ¼ 120  C justifies the higher SUCP in
A comparison of the results indicated in Tables 4 and 6 shows
these cases compared to that in the COD case. For COD case,
that increasing the absorber temperature results in a decrease of
a comparison between the results in Tables 4 and 8 reveals that as
power and cooling output as well as the thermal and exergy effi-
the heat source temperature increases, the output power increases
ciencies. This is due to the fact that as the absorber temperature
and the SUCP decreases, both can be accounted as advantages.
increases, the vapor mass flow rate passing through the turbine and
In order to show the influence of cost of heat source which may
refrigeration heat exchanger decreases. The comparison also
vary in a very wide range, different values of $3/ton, $6/ton, $9/ton
reveals that increasing the absorber temperature from 10  C to
and $12/ton are assumed in the economic optimization [38] and the
30  C has a negative influence on the SUCP. As it can be seen, for
results are outlined in Table 9.
instance, the SUCP for the COD case is increased by around 35%.
The results reveal that for each $3/ton increase in the unit cost of
Finally, it should be pointed out that, for all the three optimal cases
the heat source, the unit costs of the output power and cooling are
presented in Table 6, the quality at turbine exit, which is an
increased by around $7.6/GJ and $15e19/GJ, respectively. However,
important parameter [18], is always higher than 0.92.
when a low or medium temperature waste heat is available as
a heat source, the analyzed power/cooling cycle will become even
6.2.2. Optimization for different turbine efficiencies
more cost-effective.
In order to investigate the effects of irreversibility level in the
Regarding the unit cost of the products, a comparison can be
expansion process on the cycle output parameters and SUCP, the
made between the obtained results and those of Balli et al. [44] in
optimization is repeated for two different values of turbine effi-
a tri-generation system which combines a gas-diesel engine with
ciency and the results are outlined in Table 7. The results show that,
a heat recovery steam generator and an absorption chiller. The
the COD case seems to be the best choice among the optimal cases,
comparison shows that the unit cost of the net power and cooling
since for this case a relatively high power output with low SUCP is
output for the cycle considered in this study is considerably lower
achieved. However, a comparison among the results in this Table-
than those for the mentioned tri-generation system.
with those in Table 4 reveals that the turbine efficiency is an
important parameter from the viewpoint of thermodynamics and
economics. For COD case the output power and cooling of the cycle
6.3. Results of multi-objective optimization
decrease by around 25.6% and 15.3% as the turbine efficiency
decreases from the ideal case (100%) to 77.3%. The associated
The majority of energy system optimization cases will require
increase in SUCP is around 8.7%. Therefore employing a turbine
the consideration of multiple objectives. Often, the objectives will
with highest possible efficiency is preferred.
be conflicting; a system with low efficiency is usually a cheap one,
It is worth mentioning that, as reported in the literature and
whereas the cost of a more efficient system is high [45]. The Pareto
concluded in this work as well, the turbine efficiency is crucial for
approach is not the only possible way to deal with multi-objective
cooling production since a decrease in turbine efficiency brings
optimization problems. The most straightforward approach to
about a reduction in cooling output [12,18]. On the other hand, for
these problems is to weight each function and add them together.
a given pressure ratio, a decrease in X6 and/or an increase in DTSup
results in an increase of turbine exit temperature and consequently
in a decrease of cooling output. The efficiency value of ht ¼ 0.773 is Table 9
the lowest value with which some cooling output can be achieved, Optimum values of the sum of the unit costs of the products for different unit steam
at the decision variables ofX6 ¼ 0.965, DTSup ¼ 15  C, DTpinch ¼ 5  C cost.

and PR  14. Decision variable/unit cost of products Unit steam cost ($/ton)

3 6 9 12
6.2.3. Optimization for different heat source temperatures
Pump pressure ratio 14.81 14.23 14.34 14.40
The optimum values of decision variables and objective func- Ammonia concentration at the rectifier exit 0.970 0.968 0.966 0.965
tions for heat source temperatures of 120  C and 140  C are shown 
Minimum temperature difference in HEs ( C) 5 5 5 5
in Table 8. The results indicate that, for three optimal cases (TEOD, Degree of superheat ( C) 0.0 0.0 0.0 0.86
EEOD and COD) the variations in power output, cooling output and Unit cost of output power ($/GJ) 10.51 18.08 25.69 33.30
Unit cost of output cooling ($/GJ) 53.95 72.32 88.52 103.58
SUCP are similar to those when heat source temperature is 130  C. SUCP ($/GJ) 64.46 90.4 114.2 136.9
However, a comparatively lower value of cooling output for the
V. Zare et al. / Energy 47 (2012) 271e283 281

Fig. 12. A comparison between single and multi-objective optimization.

The different optimal solutions on the Pareto front can then be beside a relatively high cooling production with low product costs,
obtained by varying the weight coefficients [46,47]. the MOOD3 case seems to be the most promising one among those
For the ammoniaewater cycle considered in this paper, the studied. As Fig. 12 indicates, for MOOD3 case the power output is
combined objective can be constructed by summing the three just 1.8% lower than the highest value (obtained for TEOD case).
before mentioned objectives with some appropriate weights, as However, the SUCP in MOOD3 case is 16.9% lower than that in TEOD
follows [46]: case and just 2.1% higher than the minimum SUCP which is ob-
tained for COD case.

MaxðMOF ¼ w1  hth þ w2  ε þ w3  ð1  SUCP=c19 ÞÞ


0  w1 ; w2 ; w3  1 7. Conclusions
w1 þ w2 þ w3 ¼ 1
(40) Thermoeconomic analysis and optimization of a power/cooling
cogeneration cycle is carried out in this paper. For a practical range
where, c19 is the unit cost of the fuel (steam) supplied to the overall of the decision variables, the performance of the cycle is assessed
system and w1, w2 and w3 are the weighting coefficients for the to find out their effects on the thermal and exergy efficiencies as
thermal efficiency, exergy efficiency and SUCP, respectively. The well as on the unit cost of the cycle products. It is found that, the
decision maker can choose any set of optimal solutions, by selecting economic optimization leads to a considerable reduction in the
desired values of weighting coefficients. For instance, assuming product costs as compared to those obtained from the thermo-
w1 ¼ 1 leads to the TEOD case. If an agreement exists among dynamic optimization. The results show that, the sum of the unit
decision makers concerning preference (weight) of each objective, costs of the products for the cost optimal design is reduced by
final decision can be made through solving Eq. (39). around 18.6% and 25.9% as compared to that of the thermal effi-
For some values of the weighting coefficients, a comparison is ciency and exergy efficiency optimal designs, respectively. For the
performed between the single and multi-objective optimizations operating conditions considered in this work it can be concluded
for the considered cycle as shown in Fig. 12. that the cost optimal design case is the most promising one as the
In the first case of multi-objective optimal design (MOOD1), the product cost for this case is very lower than those for the other
three objectives are assumed to be of the same importance. For the optimal cases. Furthermore, for the cost optimal case the amount
MOOD2 and MOOD3 cases, it is considered that the cost is the most of produced power is close to that obtained for thermal and exergy
important objective. However, in MOOD2 case exergy efficiency is efficiency optimal design cases. It is also found that for each $3/ton
preferred to thermal efficiency while in MOOD3 case this trend is increase in unit cost of the heat source, the unit cost of output
reversed. The results reveal that, if a designer aims to attain a high power and cooling is increased by around $7.6/GJ and $15e19/GJ,
power output, which is the main product of the Goswami cycle [19], respectively.
282 V. Zare et al. / Energy 47 (2012) 271e283

Appendix A. Investment costs of the system equipments For laminar flow inside the circular tubes, the Nusselt number
is; Nu ¼ (hD)/k ¼ 3.66. The hydraulic diameter for the annulus is the
For a thermoeconomic analysis, the investment costs of equip- difference between the inside diameter of the external tube and the
ments must be evaluated. For the case of ammoniaewater cycle outside diameter of the internal tube [48].
considered in this work, the boiler, the rectifier, the RefHE, the For a vertical tier of N horizontal tubes, the average convection
RecHE, the absorber and the superheater are considered as simple coefficient (over the N tubes) may be expressed as [50]:
heat exchangers [31,35]. The investment costs of these components
are calculated based on the weighted area using the following " #14
power law relation [2,31]: grl ðrl  rv Þk3l h0fg
h ¼ 0:729 (A.5)
N mðTsat  Ts ÞD
 0:6
Ak
Zk ¼ ZR;k (A.1)
AR where, for this case, the modified latent heat is:

where, subscript k corresponds to a heat exchanger and subscript R


h0fg ¼ hfg þ 0:68cp;l ðTsat  Ts Þ (A.6)
refers to the reference component of a particular type and size.
For falling film flow, outside the vertical tubes, Patnaik suggests
Using the logarithmic mean temperature difference ðDTklm Þ and
the Wilke’s correlation as follows [48]:
the overall heat transfer coefficient (Uk), the heat transfer process in
the heat exchangers is modeled as follows:
k 
h ¼ 0:029Re0:53 Pr 0:344
Qk ¼ Uk Ak DTklm (A.2) d
 13
For each of the components, reference costs for AR ¼ 100 m2, in
3mG
d ¼
the year 2000, are given in Table A.1. r2 g
(A.7)
4G
Re ¼
Table A.1
m
Reference costs and overall heat transfer coefficient for heat m_
exchangers. G ¼
pD
Component Reference cost ($)a
Boiler 17,500 For vapor condensation, inside horizontal tubes, Dobson and
Superheater 16,000 Chato recommended an expression of the following form [50]:
RefHE 16,000
RecHE 12,000
" #14
Absorber 16,500 grl ðrl  rv Þk3l h0fg
Rectifier 17,000 h ¼ 0:555 (A.8)
a
ml ðTsat  Ts ÞD
Ref. [35].

where, for this case, the modified latent heat is:


Neglecting the wall resistance, the overall heat transfer coefficient,
U, can be defined as [48]: h0fg ¼ hfg þ 0:375cp;l ðTsat  Ts Þ (A.9)

1 The investment cost for the pump can be written as [31,35]:


U ¼ (A.3)
ðDo =Di Þð1=hi Þ þ ðDo =Di ÞFi þ Fo þ ð1=ho Þ
!mp !np
In order to calculate the overall heat transfer coefficient, phys- W _ p 1  hp
Zp ¼ ZR;p (A.10)
ical information of the heat exchanger configuration and the _
W hp
R;p
characteristics of the streams are required. For all the heat
exchangers, the value of the fouling factors (F) at the water side and where,
ammoniaewater mixture side are assumed to be 0.09 and 0.2 m2 K/ ZR;p ¼ 2100$ ðyear 2000Þ; W _ R;p ¼ 10kW; mp ¼ 0:26; np ¼ 0:5
kW, respectively [48,49]. and hp is the isentropic pump efficiency.
To evaluate the heat transfer coefficients (hi and ho), the types of For the turbine, the investment cost is as follows [51]:
exchangers should be specified. The boiler and rectifier are
assumed to be single pass horizontal tube heat exchangers. In the  
boiler, steam flows inside the horizontal tubes. The recovery heat Zt ¼ 1:5 225 þ 170V_ (A.11)
exchanger and the superheater are single pass annulus heat
exchangers in which the cold stream flows inside. The absorber and where, V_ is the volumetric flow rate of the turbine inlet stream and
the refrigeration heat exchanger are single pass vertical tube heat Zt is in V (A coefficient of 1.3 is used to convert V to $).
exchangers in which the water flows inside the tubes and the The cost of the system components which are available in an
ammoniaewater mixture condenses (or evaporates) on the outside original year is converted from that original time to a same refer-
surface of the tubes [48]. ence year (year 2010 for present work) using Marshall & Swift
For fully developed turbulent flow in smooth circular tubes, the equipment cost indices:
Nusselt number can be obtained from the DittuseBoelter equation
[50]: Cost at reference year ¼ Original cost
Cost index for the reference year
Nu ¼ 0:023Re0:8 Pr n (A.4) 
Cost index for the original year
where, n is 0.4 for heating and 0.3 for cooling. (A.12)
V. Zare et al. / Energy 47 (2012) 271e283 283

References [27] Erlach B, Serra L, Valero A. Structural theory as standard for thermoeconomics.
Energy Conversion and Management 1999;40:1627e49.
[28] Kim SM, Oh SD, Kwon YH, Kwak HY. Exergoeconomic analysis of thermal
[1] Tsatsaronis G, Lin L, Pisa J. Exergy costing in exergoeconomics. Journal of
systems. Energy 1998;23:393e406.
Energy Resources Technology 1993;115:9e16.
[29] Vieira LS, Donatelli JL, Cruz ME. Exergoeconomic improvement of a complex
[2] Bejan A, Tsatsaronis G, Moran M. Thermal design and optimization. New York:
cogeneration system integrated with a professional process simulator. Energy
John Wiley and Sons Inc; 1996.
Conversion and Management 2009;50:1955e67.
[3] Kalina AI. Combined cycle system with novel bottoming cycle. Journal of
[30] Klein SA. Engineering equation solver. Middleton, WI: F-Chart Software; 2008.
Engineering for Gas Turbines and Power 1984;106:737e42.
[31] Misra RD, Sahoo PK, Gupta A. Thermoeconomic evaluation and optimization
[4] Goswami DY. Solar thermal power: status of technologies and opportunities
of an aqua-ammonia vapour-absorption refrigeration system. International
for research. In: Proceedings of the 2nd ISHMT-ASME heat and mass trans-
Journal of Refrigeration 2006;29:47e59.
action conference. New Delhi: Tata McGraw Hill; 1995. p. 57e60.
[32] Rossa JA, Bazzo E. Thermodynamic modeling of an ammoniaewater absorp-
[5] Goswami DY. Solar thermal technology: present status and ideas for the
tion system associated with a microturbine. International Journal of Ther-
future. Energy Sources 1998;20:137e45.
modynamics 2009;12:38e43.
[6] Goswami DY, Xu F. Analysis of a new thermodynamic cycle for combined
[33] Ahrendts J. Reference states. Energy 1980;5:667e77.
power and cooling using low and mid temperature solar collectors. Journal of
[34] Ibrahim OM, Klein SA. Thermodynamic properties of ammoniaewater
Solar Energy Engineering 1999;121:91e7.
mixtures. ASHRAE Transactions 1993;99:1495e502.
[7] Lu S, Goswami DY. Optimization of a novel combined power/refrigeration
[35] Misra RD, Sahoo PK, Sahoo S, Gupta A. Thermoeconomic optimization of
thermodynamic cycle. Journal of Solar Energy Engineering 2003;125:212e7.
a single effect water/LiBr vapour absorption refrigeration system. Interna-
[8] Vijayaraghavan S, Goswami DY. On evaluating efficiency of a combined power
tional Journal of Refrigeration 2003;26:158e69.
and cooling cycle. Journal of Energy Resources Technology 2003;125:221e7.
[36] Al-Otaibi DA, Dincer I, Kalyon M. Thermoeconomic optimization of vapor
[9] Xu F, Goswami DY, Bhagwat SS. A combined power/cooling cycle. Energy
compression refrigeration systems. International Communications in Heat
2000;25:233e46.
and Mass Transfer 2004;31:95e107.
[10] Colonna P, Gabrielli S. Industrial trigeneration using ammoniaewater absorp-
[37] Gebreslassie BH, Guillén-Gosálbez G, Jiménez L, Boer D. Design of
tion refrigeration systems. Applied Thermal Engineering 2003;23:381e96.
environmentally conscious absorption cooling systems via multi-
[11] Takeshita K, Amano Y, Hashizume T. Experimental study of advanced
objective optimization and life cycle assessment. Applied Energy 2009;
cogeneration system with ammoniaewater mixture cycles at bottoming.
86:1712e22.
Energy 2005;30:247e60.
[38] Wang Y, Lior N. Thermoeconomic analysis of a low-temperature multi-effect
[12] Vidal A, Best R, Rivero R, Cervantes J. Analysis of a combined power and
thermal desalination system coupled with an absorption heat pump. Energy
refrigeration cycle by the exergy method. Energy 2006;31:3401e14.
2011;36:3878e87.
[13] Zheng D, Chen B, Qi Y, Jin H. Thermodynamic analysis of a novel absorption
[39] Klein SA. Engineering equation solver user’s manual. Middleton, WI: F-Chart
power/cooling combined-cycle. Applied Energy 2006;83:311e23.
Software; 2008.
[14] Liu M, Zhang N. Proposal and analysis of a novel ammoniaewater cycle for
[40] Baghernejad A, Yaghoubi M. Exergoeconomic analysis and optimization of an
power and refrigeration cogeneration. Energy 2007;32:961e70.
integrated solar combined cycle system (ISCCS) using genetic algorithm.
[15] Zhang N, Lior N. Development of a novel combined absorption cycle for power
Energy Conversion and Management 2011;52:2193e203.
generation and refrigeration. Journal of Energy Resources Technology 2007;
[41] Maschio C, Vidal AC, Schiozer DJ. A framework to integrate history matching
129:254e65.
and geostatistical modeling using genetic algorithm and direct search
[16] Wang J, Dai Y, Gao L. Parametric analysis and optimization for a combined
methods. Journal of Petroleum Science and Engineering 2008;63:34e42.
power and refrigeration cycle. Applied Energy 2008;85:1071e85.
[42] Charbonneau P, Knapp B. A user’s guide to PIKAIA 1.0. NCAR technical note
[17] Wang J, Dai Y, Zhang T, Ma S. Parametric analysis for a new combined power
418þIA. Boulder: National Center for Atmospheric Research; 1996.
and ejector absorption refrigeration cycle. Energy 2009;34:1587e93.
[43] Charbonneau P. Release notes for PIKAIA 1.2. NCAR technical note 451þSTR.
[18] Padilla RV, Demirkaya G, Goswami DY, Stefanakos E, Rahman MM. Analysis of
Boulder: National Center for Atmospheric Research; 2002.
power and cooling cogeneration using ammoniaewater mixture. Energy
[44] Balli O, Aras H, Hepbasli A. Thermodynamic and thermoeconomic analyses of
2010;35:4649e57.
a trigeneration (TRIGEN) system with a gasediesel engine: part II e an
[19] Demirkaya G, Padilla RV, Goswami DY, Stefanakos E, Rahman MM. Analysis of
application. Energy Conversion and Management 2010;51:2260e71.
a combined power and cooling cycle for low-grade heat sources. International
[45] Spelling J, Favrat D, Martin A, Augsburger G. Thermoeconomic optimization of
Journal of Energy Research 2011;35:1145e57.
a combined-cycle solar tower power plant. Energy 2012;41:113e20.
[20] Goswami DY, Tamm G, Vijayaraghavan S. A new combined power and cooling
[46] Sayyaadi H, Saffari A, Mahmoodian A. Various approaches in optimization of
cycle for low temperature heat sources. In. Proceedings of international joint
multi effects distillation desalination systems using a hybrid meta-heuristic
power generation conference. ASME, Atlanta, U.S.A.; June 2003. p. 16e9.
optimization tool. Desalination 2010;254:138e48.
[21] Pouraghaie M, Atashkari K, Besarati SM, Nariman-zadeh N. Thermodynamic
[47] Haupt RL, Haupt SE. Practical genetic algorithms. 2nd ed. New Jersey: John
performance optimization of a combined power/cooling cycle. Energy
Wiley and Sons Inc; 2004.
Conversion and Management 2010;51:204e11.
[48] Florides GA, Kalogirou SA, Tassou SA, Wrobel LC. Design and construction of
[22] Lozano MA, Valero A. Theory of the exergetic cost. Energy 1993;18:939e60.
a LiBrewater absorption machine. Energy Conversion and Management 2003;
[23] Lazzaretto A, Tsatsaronis G. On the calculation of efficiencies and costs in
44:2483e508.
thermal systems. Proceedings of ASME Advanced Energy Systems Division
[49] Kwon K, Jeong S. Effect of vapor flow on the falling-film heat and mass
1999;39:421e30.
transfer of the ammonia/water absorber. International Journal of Refrigeration
[24] Lazzaretto A, Tsatsaronis G. SPECO: a systematic and general methodology for
2004;27:955e64.
calculating efficiencies and costs in thermal systems. Energy 2006;31:1257e89.
[50] Bergman TL, Lavine AS, Incropera FP, DeWitt DP. Fundamentals of heat and
[25] Frangopoulos CA. Thermoeconomical functional analysis: a method for
mass transfer. 7th ed. John Wiley & Sons; 2011.
optimal design or improvement of complex thermal systems. Ph.D. thesis.
[51] Quoilin S, Declaye S, Tchanche BF, Lemort V. Thermo-economic optimization
Atlanta, USA: Georgia Institute of Technology; 1983.
of waste heat recovery organic Rankine cycles. Applied Thermal Engineering
[26] Frangopoulos CA. Thermoeconomic functional analysis and optimization.
2011;31:2885e93.
Energy 1987;12:563e71.

You might also like