See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/279537387

d2

Data · July 2015

CITATIONS READS

0 1,472

5 authors, including:

Robert Brunet Daniel Cortés-Borda

Universitat Rovira i Virgili University of Nantes
16 PUBLICATIONS   116 CITATIONS    11 PUBLICATIONS   83 CITATIONS   

SEE PROFILE SEE PROFILE

Laureano Jiménez Dieter Boer

Universitat Rovira i Virgili Universitat Rovira i Virgili
259 PUBLICATIONS   2,104 CITATIONS    87 PUBLICATIONS   1,183 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Environmental management tools for sewage sludge management View project

Master research: Comparison of the construction materials through the Life Cycle Assessment: Drywall and Masonry (in Portuguese) View
project

All content following this page was uploaded by Daniel Cortés-Borda on 03 July 2015.

The user has requested enhancement of the downloaded file.

 Applied Thermal Engineering 48 (2012) 367e377

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

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

Minimization of the LCA impact of thermodynamic cycles using a combined

simulation-optimization approach
Robert Brunet a, Daniel Cortés a, Gonzalo Guillén-Gosálbez a, *, Laureano Jiménez a, Dieter Boer b
a
Departament d’Enginyeria Quimica, Escola Tecnica Superior d’Enginyeria Quimica, Universitat Rovira i Virgili, Campus Sescelades, Avinguda Paisos Catalans 26,
43007 Tarragona, Spain
b
Departament d’Enginyeria Mecanica, Escola Tecnica Superior d’Enginyeria, Universitat Rovira i Virgili, Campus Sescelades, Avinguda Paisos Catalans 26, 43007,
Tarragona, Spain

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

Article history: This work presents a computational approach for the simultaneous minimization of the total cost and
Received 15 November 2011 environmental impact of thermodynamic cycles. Our method combines process simulation, multi-
Accepted 13 April 2012 objective optimization and life cycle assessment (LCA) within a unified framework that identifies in
Available online 26 April 2012
a systematic manner optimal design and operating conditions according to several economic and LCA
impacts. Our approach takes advantages of the complementary strengths of process simulation (in which
Keywords:
mass, energy balances and thermodynamic calculations are implemented in an easy manner) and
Process simulation
rigorous deterministic optimization tools. We demonstrate the capabilities of this strategy by means of
Optimization
Rankine cycle
two case studies in which we address the design of a 10 MW Rankine cycle modeled in Aspen Hysys, and
Absorption cycle a 90 kW ammonia-water absorption cooling cycle implemented in Aspen Plus. Numerical results show
Cost analysis that it is possible to achieve environmental and cost savings using our rigorous approach.
Life cycle assessment
2012 Elsevier Ltd. All rights reserved.

1. Introduction coefficient of performance (COP) have been reported [3], which

leads to significant savings in primary energy sources [4].
The energetic and economic analysis of industrial processes has A variety of optimization approaches have been applied to
gained wider interest in recent years. This has been motivated by thermodynamic cycles. Some studies in power cycles focus on the
the need to use the resources available nowadays more efficiently. minimization of a single indicator, such as the net present value
In this context, process optimization has emerged as an effective (NPV), total plant cost (TPC) [5,6], and cycle efficiency [7e10]. In
tool for reducing energy consumption and improving efficiency in cooling cycles, some models were devised to optimize the COP and
process industries. Multi-objective optimization (MOO), in partic- cooling load [3]. The application of MOO to thermodynamic cycles,
ular, offers decision makers a suitable framework to identify the however, has been quite scarce. The simultaneous optimization of
set of operating conditions and design variables that simulta- the exergetic efficiency and the TPC in power generation systems
neously improve the economic and environmental performance of was studied by BecerraeLopez and Golding [11] and Dipama et al.
a system [1]. [2]. Pelet et al. [12] optimized a superstructure of energy systems
Thermodynamic cycles are widely used in energy conversion considering the cost and CO2 emissions. In the context of cooling
processes. They are often found in daily life, but have the drawback cycles, Gebreslassie et al. [13,14] proposed a multi-objective non-
of requiring large amounts of energy to operate. By optimizing linear programming (moNLP) problem for the design of an
power generation cycles, (e.g Rankine cycle) it is possible to ammonia-water absorption cycle considering the cost and life cycle
increase their efficiency and reduce the associated global warming assessment (LCA) performance [15,16].
emissions [2]. Cooling cycles can also benefit from the application The overwhelming majority of the works mentioned above
of rigorous optimization tools. Increments of up to 50% in their follow the so called simultaneous approach, which relies on
formulating algebraic optimization models described in an explicit
form. For simplicity, most of these formulations contain short-cut
models that avoid the numerical difficulties associated with
* Corresponding author. Tel.: þ34 977558618. handling non-linear equations. These simplified formulations
E-mail address: gonzalo.guillen@urv.cat (G. Guillén-Gosálbez). provide ”good” approximations when certain assumptions hold,

1359-4311/$ e see front matter
2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2012.04.032
368 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377

but can lead to large numerical errors otherwise. Sequential process

simulation models are more difficult to optimize due to the pres-
ence of non-convexities of different types, but have the advantage
of producing more accurate results. Another limitation of the works
mentioned above is that those that account for environmental
concerns restrict the analysis to a single environmental indicator,
neglecting the effects caused in other environmental damages.
This work applies a combined approach for the design of ther-
modynamic cycles that takes advantage of the complementary
strengths of sequential modular process simulators (e.g. Aspen
Hysys and Aspen Plus), optimization tools (e.g. SNOPT and CPLEX)
and LCA. The pivotal idea of our method is to optimize modular
simulation models of thermodynamic cycles using an external
deterministic optimizer that is guaranteed to converge to an
optimal solution. Our approach is inspired by other simulation- Fig. 1. Steam Rankine cycle.
optimization methods used in a variety of chemical engineering
applications, including the design of systems such as: heat
exchangers and chemical reactions [17e19], chemical plants solution heat exchanger (SHX), solution pump (P), and solution
[20,21], distillation columns [22], and biotechnological processes expansion valve (VLV2). It is assumed that the system works under
[23]. An efficient solution method is presented for tackling these steady state conditions. Heat and pressure losses are neglected.
problems based on decomposing them into two sub-levels between Adiabatic valves are considered. The refrigerant leaves the
which an algorithm iterates until a stopping criterion is satisfied. condenser, absorber and bottom of the generator as saturated
This algorithm performs the calculations using both a process liquid.
simulation and an external optimizer.
The final goal of our analysis is to identify the design and 2.3. Problem definition
operating conditions of different thermodynamic cycles that
simultaneously minimize the total annualized cost (TAC) and The problems can be formally stated as follows. In the case of the
environmental impact (EI). We demonstrate the capabilities of this Rankine cycle, we are given the flowsheet arrangement, net power
methodology through its application to the design of two cycles: yield, turbines and pumps efficiencies, overall heat transfer coef-
a steam Rankine cycle and an ammonia-water absorption cycle. The ficients, thermodynamic properties, cost estimation correlations,
optimization of the steam Rankine cycle is formulated as a moNLP economic parameters and environmental indicators. For the
problem, which is optimized with a Successive Quadratic absorption cycle, we need to specify as well the cooling capacity,
Programming (SQP) solver that interacts at each iteration with the and inlet and outlet temperatures of the external fluids.
process simulator of choice. The optimization of the absorption The goal of our study is to identify the optimal design and
cycle gives rise to a multi-objective mixed-integer non-linear operating conditions that simultaneously minimize the TAC and
programming (moMINLP) problem, in which binary variables are the following damage impact indicators: damage to human
employed to model the number of trays in the desorber. health (HH), damage to ecosystem quality (EQ) and depletion of
resources (DR).
2. Problem statement
3. Methodology
As previously mentioned, we will focus herein on two energy
conversion cycles: a steam Rankine cycle for power generation and This section describes the approach proposed to tackle the
an ammonia-water absorption cooling cycle. Note, however, that problems described above. A general mathematical formulation is
the approach presented is general enough to be adapted to any first presented. We then describe how the economic and environ-
other energy system. We provide next a brief description of each of mental objective functions are calculated. The solution procedure
these systems before immersion into a detailed mathematical and the computer implementation are finally discussed.
formulation.

2.1. Rankine cycle

We consider a reheat-regenerative power cycle with one closed

and one open feedwater heater (see Fig. 1). The system contains one
boiler, one turbine, a condenser, two pumps, and two shell-tube
heat exchangers. Water is used as working fluid in the cycle. The
boiler is assumed to operate with natural gas. The combustion
gases behave as air. For the condenser as well as the heat
exchangers, we use shell-tube heat exchangers.

2.2. Absorption cycle

We consider the single effect ammonia-water absorption cool-

ing cycle described by Gebreslassie et al. [13,14] (see Fig. 2). The
absorption cycle provides chilled water at 5  C. The equipment
units are the absorber (A), condenser (C), rectification column (RC),
evaporator (E), subcooler (SC), refrigerant expansion valve (VLV1), Fig. 2. Ammonia-water absorption cycle.
 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377 369

3.1. Mathematical formulation equipment and maintenance cost, which are determined using the
costing correlations described in sections 4.1 and 4.2.
The design of thermodynamic cycles with economic and envi- X
ronmental concerns can be expressed in mathematical terms as CF ¼ Cu (5)
a moMINLP. We solve this model using the ε-constraint method u˛U
[24,25]. This technique is based on calculating a set of single-
objective models in which one objective is kept in the objective
function while the others are transferred to auxiliary constraints 3.2.2. Environmental indicator (damage categories)
and forced to be lower than a set of epsilon parameters: The environmental impact is quantified following LCA princi-
ples, similarly as done before by the authors in other works [23].
min z ¼ ff1 ðx; u; xD Þg
xD Further details on the calculations are provided in the Appendix.
s:t: fo ðx; u; xD Þ  εo o ¼ 2; :::; n
h1 ðx; u; xD Þ ¼ 0 (1)
3.3. Solution procedure
hE ðx; u; xD Þ ¼ 0
gE ðx; u; xD Þ  0 3.3.1. ε-Constraint methodology
Where f1 is the economic objective function, and f2 to fn denote the The solution of the multi-objective model is given by a set of
LCA metrics used to assess the environmental performance of the Pareto points representing the optimal compromise between the
cycle. ε is an auxiliary parameter that bounds the values of the objectives considered. These points are generated combining the
objectives transferred to the auxiliary inequality constraints. ε-constraint method [24,25] with a tailored decomposition algo-
Equations hI are implicit equations implemented in the process rithm that integrates simulation and optimization tools. The solu-
simulator, whereas hE and gE are explicit constraints that ensure tion method proposed is shown in Fig. 3. It comprises two nested
certain process conditions. The form of these equations depends on loops: an outer loop in which epsilon values are defined for the
the system under study. environmental impacts, and an inner loop that solves each single-
The design variables are denoted by xD, while other process objective problem. We provide next details on the inner loop of
variables are represented by x. Finally, u denotes parameters not the algorithm.
modified during the calculations. It is important to note that xD
include only continuous variables in the case of the Rankine cycle, 3.3.2. Simulation-optimization approach
while in the case of the absorption cycle it includes both, contin- The solution strategy for solving each single-objective problem
uous and integers (i.e., number of trays and feed tray in the relies on an outer-approximation [26] scheme that decomposes
absorber). each model into two hierarchical levels: a primal non-linear
programming (NLP) sub-problem and a master mixed-integer

3.2. Objective functions

The model presented, seeks to optimize simultaneously the TAC

and environmental impact. We describe next how these indicators
are calculated.

3.2.1. Economic indicator (total annualized cost)

The TAC of the thermodynamic cycles is given by Eq. (2).

TAC ¼ CO þ CF$crf (2)

Where CO and CF are the operating and fixed costs, and crf is the
capital recovery factor, which is a function of the interest rate
(parameter i) and the lifetime of the cycle (parameter t) expressed
in years (see Eq. (3)).
!
ið1 þ iÞt
crf ¼ (3)
ð1 þ iÞt e1

The operation cost, denoted by CO, accounts for the cost of the
energy and electricity required to operate the cycle.
X
CO ¼ ðQu $cq þ Wu $cwÞ$top (4)
u˛U

In this equation, Qu [MW] is the thermal power supplied to

equipment unit U, Wu [MW] is the electrical power required by
equipment unit, top [h] is the total annual operation time and cq
[V/MWh] and cw [V/MWh] are the unit costs for heat and elec-
tricity respectively. Note that Qu and Wu are provided by the process
simulator.
Eq. (5) determines the total fixed cost (CF) which accounts for
the cost of the main equipment units of the cycle (Cu) including the Fig. 3. Flowchart of the proposed outer-approximation algorithm.
370 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377

linear programming (MILP) sub-problem. The algorithm iterates Dgki;j ¼ Difference between the values of the inequality constraint
between these levels until a termination criterion is satisfied. j for the new rectification column design i and constraint j in the
The master MILP is constructed using information provided by original NLPk problem.
the primal NLP. This primal NLP is solved integrating a determin- DhkEi;j ¼ Difference between the values of the external equality
istic gradient-based method with the process simulator. The binary constraint j in the new rectification column design i and constraint j
variables are thus handled by the MILP, while the NLP provides the in the original NLPk problem.
optimal values of the continuous variables for a fixed set of binaries. The master MILP takes the following form:
This strategy is inspired by previous simulation-optimization !
approaches applied in chemical engineering [17e23]. The main Q P
n P P
advantage of this method is that it ensures convergence to a local
min aþ s1o þ s2 j þ s3 j
o¼2 j˛IEQ j˛EQ
(or global) optimum, as opposed to heuristic-based approaches that   P  vf   
o
are unable to guarantee the optimality of the solutions calculated. s:t: fo xk ; uk ; xkD þ xDn  xkDn
n vxDn xD
n ¼xDn
i

3.3.2.1. Primal NLP sub-problem. This level optimizes the contin- X

uous decision variables of the NLP sub-problem for fixed values of þ yi $Dobjki;o  a o ¼ 1
i˛Tk
the binary variables predicted by the master sub-problem (Eq. (6)).
This procedure is repeated iteratively for different values of the   X  vf   
o
binary variables until a termination criterion is met. The NLP sub-
fo xk ; uk ; xkD þ xDn  xkDn
vxDn xD
n ¼xDn
i
n
problems are solved using a gradient-based SQP solver that iter-
X
ates with the simulation package in order to obtain information on þ yi $Dobjki;o  εo þ s1o o ¼ 2; :::; n
the derivatives of the decision variables with respect to the objec- i˛Tk
tive function and constraints.   X  vg   
j
Slack variables are used to relax the external equality and gj xk ; uk ; xkD þ xDn  xkDn
inequality constraints, which avoids unconvergencies in the slave n vxDn xD ¼xi
n Dn

problem. Potential intermediate unfeasible points are thus handled X

externally by the optimization algorithm. These slacks are penal- þ yi $Dgi;j  s2j cj˛IEQ
k

i˛Tk
ized in the objective function. This approach avoids unfeasible
simulation runs, preventing the algorithm from ending prema-     X  vhE   
sign lj hEj xk ; uk ; xkD þ
k
turely. The modified objective function is expressed as follows.
j
xDn  xkDn
n vxDn xD ¼xi
n Dn
Q X
min z ¼ f1 ðx; u; xD Þ þ ðs1 þ s2 þ s3 þ s4 Þ þ yi $DhEi;j
k
xD
i˛Tk
s:t: fo ðx; u; xD Þ  εo þ s1 o ¼ 2; :::; n
ε0  εo  εo o ¼ 2; :::; n  s3 j cj˛EQ
hI ðx; u; xD Þ ¼ 0 (6) k ¼ 1; 2; 3; :::; K
hE ðx; u; xD Þ þ s2 es3 ¼ 0 2 3
1o  0 s2j  0
sP s3j  0
gE ðx; u; xD Þ  s4 6 yi ¼ 1 7
4 5 (7)
i˛T
s1  0; s2  0; s3  0; s4  0;
yi ˛f0; 1g
Q
Where is a penalty parameter vector, and s1, s2, s3 and s4 are
The objective function of the MILP contains an auxiliary variable
vectors of positive slack variables. Q
(a) and a penalty value for constraint violation ( ) that multiplies
3.3.2.2. Master MILP sub-problem. The master sub-problem the slack variables. The first constraint is formed by three terms: (i)
provides new values for the binary variables that are expected to the objective function value at iteration k of the algorithm, (ii) the
yield better results than previous solutions. Note that this master linearization performed on the design variables, and (iii) the
MILP is only required in the case of the absorption cycle, in which contribution of changing the current distillation column charac-
the number of trays of the desorber must be decided. In contrast, teristics, by either adding or removing stages in the column or
the optimization of the Rankine cycle can be solved as an NLP. changing the feed stage. This last term is the product of the binary
To construct the master MILP, we use the derivatives of the variable yi (that is 1 if topological modification i is implemented
objective function and constraints of the NLP sub-problem at the and 0 otherwise) with the parameter Dobjki;o. The latter accounts for
optimal NLP solution of the previous iteration. Due to the presence the change in the objective function value when topology i is
of non-convexities in the NLP, the master MILP is not guaranteed to implemented. Fig. 4 provides an illustrative example on how these
provide a rigorous lower bound on the global optimum. The terms are defined.
following notation is defined in the MILP at iteration k of the External inequality (IEQ) and equality (EQ) constraints are
handled following a similar procedure. signðlj Þ refers to the sign of
k
algorithm:
T ¼ {iji is a potential column configuration} the Lagrange multiplier of constraint j at iteration k. This value is
Tk ¼ {iji is a rectification column configuration, entailing a given used to correctly relax equalities into inequalities [27]. Note that
number of trays and a specific feed stage, which can be obtained linear constraints are accumulated in the master MILP, so at itera-
performing one single modification on the design calculated at tion k, the problem includes constraints from current and previous
iteration k} iterations.
EQ ¼ {jjj is an external (explicit) equality constraint} After determining the new set of values for the binary variables,
IEQ ¼ {jjj is an external (explicit) inequality constraint} the primal problem is solved again, and the overall procedure is
Dobjki;o ¼ Difference between the objective function o at itera- repeated until the termination criterion is satisfied. Integer cuts can
tion k of the NLP and the objective function associated with the new be added to the master MILP in order to avoid repetition of solu-
rectification column design i tions explored so far in previous iterations. Implicit constraints are
 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377 371

Fig. 4. Details on the definition of binary variables in the MILP (inspired in the work by Caballero et al. [22]).

handled by the process simulator and their derivatives are obtained As NLP solver, we used SNOPT [34], which was accessed via the
by finite differences. Tomlab [35] modeling system supported by Matlab. This solver is
Note that the complexity of the overall solution procedure particularly suited for non-linear problems whose functions and
grows rapidly with the number of environmental objectives. In gradients are expensive to evaluate [36]. The master MILP sub-
cases with a large number of objectives, we might be interested in problem was solved using the MIP solver CPLEX [37], accessed via
applying dimensionality reduction methods to keep the problem in Tomlab. Fig. 5 outlines the computer architecture of the solution
a manageable size [28e30]. algorithm proposed.

4. Case studies
3.4. Computational implementation
Two thermodynamic cycles were studied, a steam Rankine cycle
We use the process simulators Aspen Hysys [31] and Aspen Plus and an ammonia-water absorption cycle. Both systems were
[32] to simulate the thermodynamic cycles. These software pack- simulated using standard commercial process simulators, thereby
ages allow an easy modeling of the cycles, as they implement avoiding the definition of the thermodynamic equations in an
thermodynamic correlations, built-in models for a variety of unit explicit form.
operations and mass and energy balances. These process simulators
were connected with Matlab [33], in which the main code of the 4.1. Case study I: steam Rankine cycle
algorithm was implemented. This software gets the values of the
dependent variables (e.g., temperature, pressure, mass and energy 4.1.1. System description
flows) from the process simulators at each iteration of the The first case study addresses the design of a 10 MW steam
algorithm. Rankine cycle (see Fig. 6) taken from Moran and Shapiro [38]. The

Fig. 5. Main steps of the solution algorithm proposed.

372 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377

Fig. 6. Steam Rankine cycle simulated in Aspen Hysys.

cycle was simulated in Aspen Hysys under steady state conditions. condenser, boiler and reboiler). The mass of steel from tubes,
Heat and pressure losses were neglected. Adiabatic efficiencies in pumps, valves and other equipments in the cycle were neglected.
turbines and pumps were set to 75% [39]. An adiabatic expansion The amount of stainless steel contained in the heat exchangers was
valve was considered in the calculations. determined from the exchange area assuming a thickness of 1/4
inches. The weight of the turbine was assumed to be equal to 10
4.1.2. System modeling tons (typical weight of a 10 MW turbine [42]).
The properties of water, selected as the working fluid of the
cycle, were calculated using the ASME steam tables. The boiler and
reboiler operate with natural gas. The composition of the 4.2. Case study II: absorption cooling cycle
combustion gases in the boiler and reboiler is unknown, but we
assume that they behave as air, which was modeled using UNI- 4.2.1. System description
QUAC. For the condenser, heat exchanger, boiler and reboiler The second example studies a 90 kW single effect ammonia-
simulation, we considered shell-tube heat exchangers, which were water absorption cooling cycle (see Fig. 7). This cycle is discussed
modeled using the weighted model built-in Aspen Hysys. The in detail in Gebreslassie et al. [13,14]. The absorption cycle provides
boiler and reboiler were simulated as separated heat exchangers. water at 5  C.
The same approach was applied to the turbine. The mixer was
modeled as an open flow heat exchanger that mixes streams at 4.2.2. System modeling
different temperatures. The RedlicheKwongeSoave equation of state was selected to
model the ammonia-water mixture in vapor phase [43]. For the
4.1.3. Objective functions simulation of the liquid mixture, the Non-Random Two Liquid
The heat cost was set to 25 V/MWh, and the operation time to model was employed. The absorber, condenser, evaporator, sub-
4000 h per year. The energy flows in the boiler and reboiler were cooler and solution heat exchanger were simulated using the
retrieved from Aspen Hysys. The cost of the expansion valves and MheatX model. The desorber was simulated with a rigorous tray-
mixer were neglected. Table 1 shows the cost estimation correla- by-tray distillation column model.
tions used for the remaining equipment units [39e41].
The environmental impact of the operation phase was deter- 4.2.3. Objective functions
mined from the energy flows imported from Aspen Hysys. To The operational costs were calculated with Eq. (4), assuming an
calculate the environmental impact of the construction phase, we electricity cost of 100 V/MWh, a heat cost of 25 V/MWh, and an
considered only the turbine and heat exchangers (heat exchanger, operation time of 4000 h per year. The energy flows (electricity and
heat) in the pump and desorber were retrieved from Aspen Plus.
The cost correlations are given in Table 2 [44e46].
Table 1 The energy flows were retrieved from the process simulator (in
Cost correlations used in the Rankine cycle.
this case Aspen Plus). The mass of steel contained in the pipes,
Equipment Correlation Reference valves and other equipments in the cycle were neglected. The mass
Boiler and reboiler CB ¼ fmð1 þ fd þ fpÞ$ðQB Þ0:86 Walas [40] of steel contained in the heat exchangers was calculated following
Condenser and CHX ¼ fd$fm$fp$Cb Evans et al. [41] the same approach as in case study 1. The mass of steel from the
heat exchanger desorber was determined by approximating the distillation column
Turbine CT ¼ 4750$ðWT Þ0:75 Nafey [39] by a cylinder. The dimensions of the desorber were imported from
Pumps CP ¼ 3500$ðWP Þ0:47 Nafey [39]
the process simulator.
 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377 373

Fig. 7. Ammonia-water absorption cycle simulated in Aspen plus.

5. Results and discussion Three bi-criteria Pareto sets were generated optimizing the TAC
against each single damage impact category separately (see Fig. 8).
The design problem aims to determine the optimal operating Fig. 8 represents the Pareto solutions of the three bi-objective
conditions of the cycle (fluid flow rates, equipment sizing and optimization problem TAC vs HH, TAC vs EQ and TAC vs DR. As
system pressures and temperatures) that minimize simultaneously observed, the impact in damage category HH was reduced by 2.40%
the economic indicator (TAC) and different impact categories (HH, (334.89 kPoints vs 342.93 kPoints) along the Pareto curve. This was
EQ and DR) given a fixed energy capacity of the cycle. accomplished by increasing the heat exchanger areas, thereby
We generated in both cases a set of Pareto solutions that we reducing the natural gas consumption. This led in turn to an
obtained for simplicity minimizing the TAC versus each individual increase of 3.65% in the TAC (3491 MV/yr vs. 3619 MV/yr). In
damage category separately. addition, the EQ was reduced by 2.38% (336.52 kPoints vs 344.55
kPoints) along the Pareto curve at the expense of increasing the TAC
by 3.84% (3491 MV/yr vs. 3625 MV/yr). Note that in both cases,
5.1. Case study I: steam Rankine cycle
solutions with lower TAC entail larger natural gas consumption
rates and smaller equipments. Finally we analyze the trade-off
We first studied a 10 MW Rankine cycle. The problem was
solutions between TAC vs DR. Here, the DR was decreased by
solved as a moNLP with the following 11 design continuous
2.22% (2873.06 kPoints vs 2941.19 kPoints) while the TAC was
variables: mass flow passing through the cycle (mass flow 1),
increased by 4.44% (3491 MV/yr vs 3646 MV/yr). Further inspec-
temperatures of streams 1 and 4, pressure of stream 1, outlet
tion of the results reveals similar insights, regarding operating
pressure of the turbines (pressures 2, 3, 5 and 6) and outlet
conditions and design characteristics, as in the previous cases.
temperature of the heat exchangers (temperatures B1 and B2). In
addition, the model includes 5 non-linear inequality constraints:
power equal or higher than 10 MW, and a minimum temper-
ature difference of 10  C in the heat exchangers. The remaining 5
x 10 x10
6
3.48 2.975
process variables and constraints are defined in an implicit form
using the process simulator (Aspen Hysys). The algorithm takes
Damage to Ecosystem Quality (Points)

3.46 2.95
Depletion Natural Resources (Points)

around 600 to 1000 CPU seconds to generate 10 Pareto solutions

Damage to Human Health (Points)

of each 2-dimensional Pareto set on a computer AMD Phe- 3.44 2.925

nomTM 8600B, with a Triple-Core Processor 2.29 GHz and
3.23 GB of RAM. 3.42 2.9

3.4 2.875
Table 2
Cost correlations used in the absorption cooling cycle.
3.38 2.85
Equipment Correlation Reference
  2.825
Desorber MS Guthrie [44] 3.36 HH
CRC ¼ ð101:9Diam1:066 H 0:802 Þ EQ
280
DR 2.82
 ð2:18 þ 2Fc Þdeuro 3.34
P 3.48 3.5 3.52 3.54 3.56 3.58 3.6 3.62 3.64
Heat CHX2 ¼ ðc1 Amu˛HX þ c2 Þ Kizilkan et al. [45] 6
exchangers u˛HX Total Annualized Cost (euro/yr) x 10
Pump CP ¼ c3 WP0:4 Siddiqui [46]
Fig. 8. Total annualized cost vs impact damage categories. Case study I: Rankine cycle.
374 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377

100 drop in the turbine is lower in the minimum cost solution, which
Percentage with respect to the maximum value (%)

99.5
leads to smaller turbines and investment costs. Table 3 also displays
the heat exchangers areas and the energy consumption (heat and
99
electricity) of the extreme solutions. As observed, the heat
98.5 exchangers area in the economic optimum is between 5 and 11%
98
smaller than in the minimum environmental impact solutions.
Regarding energy consumption, the use of heat and electricity in
97.5 the minimum impact designs is between 1 and 3.5% smaller than in
97 the economic optimum.
The objective function values of the extreme designs are
96.5
compared in Table 4. Note that impacts HH, EQ and DR can be
96 decreased by up to 2.40%, 2.38% and 2.22% respectively.
95.5
TAC HH EQ DR 5.2. Case study II: absorption refrigeration cycle
95
1 2 3 4
Objective functions A moMINLP model of the 90 kW absorption cycle was devel-
Fig. 9. Parallel coordinates plot. Case study I: Rankine Cycle. Objective functions: Total
oped. This formulation featured 10 design variables, 8 continuous
annualized cost (TAC), damage to human health (HH), damage to ecosystem quality and 2 discrete, and 4 non-linear inequality constraints. The
(EQ), and depletion of natural resources (DR). continuous variables denote the reboiler duty in the desorber, the
high and low pressure of the system, the mass flow and mass
fraction of stream 1, the temperature at the outlet of the hot side of
Fig. 9 depicts the Pareto solutions in a parallel coordinates plot, unit SHX (temperature 5), the temperature at the outlet of the hot
which is a useful graphical tool to display data sets of large side of the SC unit (temperature 9), and the reflux ratio in the
dimension. The figure shows in the x axis the set of objective desorber. Discrete variables model the number of trays and the feed
functions (TAC, HH, EQ and DR) and in the y axis the normalized tray in the desorber. Inequality constraints impose a minimum
value attained by each solution in every criterion. The normaliza- cooling capacity and minimum temperature difference between
tion was performed by dividing each objective function value by its the inlet and outlet external flows. The remaining process variables
maximum over the entire set. Note that each line in the plot and constraints were implemented in the process simulator, in this
represents a different Pareto solution, entailing a set of operating case Aspen Plus. The algorithm took around 2500 to 3000 CPU
conditions. As observed, all environmental impacts seem equiva- seconds to generate 10 Pareto solutions on the same computer as
lent, since they tend to behave similarly. Moreover, all of the before.
impacts are conflictive with the TAC of the cycle. This is because Fig. 10 depicts the Pareto solutions of the three bi-objective
reductions in the environmental impact are achieved at the optimization problems: TAC vs HH, TAC vs EQ and TAC vs DR. The
expenses of increasing the cost. HH index is reduced by 5.84% (2734 points vs 2584 points) along
Table 3 shows the details of the corresponding extreme points the Pareto curve. This is accomplished by reducing the steam
(i.e., minimum TAC and minimum environmental damage). We provided to the cycle. On the other hand, the TAC is increased by
present first the decision variables values in the extreme solutions, 4.66% (21,917 V/yr vs. 22,940 V/yr). The steam consumption is
which differ mainly in the mass flow rate and temperature of reduced by increasing the heat exchanger areas, which leads to
stream 1, and the pressure in the turbines. The mass flow rate of larger capital investments. Concerning the EQ, this is reduced by
stream 1 in the minimum cost solution is greater than in the 6.82% (2740 points vs 2565 points) along the Pareto curve, whereas
minimum environmental impact. This is because larger mass flow the TAC is increased by 4.71% (21,917 V/yr vs. 22,951 V/yr). Finally,
rates require more natural gas to evaporate water in the boiler and impact DR is decreased by 7.03% (10,497 points vs 11,228 points),
reboiler. The temperature of stream 1 in the economic optimum is while the TAC is increased by 4.73% (21,917 V/yr vs 22,954 V/yr).
lower than in the environmental optimum. Moreover, the pressure Note that all the Pareto solutions involve the same configuration of
the rectification column (1 single stage).
Table 3
The Pareto solutions obtained in the bi-criteria problems were
Details of the extreme solutions. Case study I: Rankine cycle. plotted in a parallel coordinates plot (see Fig. 11). Similar conclu-
sions as in the Rankine cycle are obtained in this second case.
Variable Min TAC Min HH Min EQ Min DR
As observed, the environmental impacts are equivalent since
Mass flow stream 1 [kg/s] 9.33 8.78 8.82 8.90
when one is minimized the others are also decreased. This is
Temperature of stream 1 [ C] 516.43 587.30 577.85 571.55
Pressure of stream 1 [kPa] 8550.08 8878.49 8860.25 8805.51 because all the damages are highly dependent on the steam
Pressure of stream 2 [kPa] 2300.00 2269.70 2274.19 2276.43 consumption. Further, they are all conflictive with the cost as their
Pressure of stream 3 [kPa] 737.29 698.34 703.54 707.00 minimization increases the cost of the cycle.
Pressure of stream 5 [kPa] 329.92 301.20 304.74 307.58
Table 5 summarizes the extreme points (i.e., minimum TAC and
Pressure of stream 6 [kPa] 8.80 7.00 7.23 7.40
Temperature of stream 4 [ C] 496.09 493.99 494.39 494.46
minimum environmental damage indicators). As in the previous
Temperature of stream 11 [ C] 205.19 215.15 214.66 212.94 case, the minimum TAC design differs considerably from the
Temperature of combustion 250.00 249.90 249.91 249.92
gases B1 [ C]
Temperature of combustion 278.69 300.00 295.56 295.26 Table 4
gases B2 [ C] Extreme solutions. Case study I: Rankine cycle.
Area of the boiler and 160.82 165.47 165.73 166.59
reboiler [m2] Objective function Min TAC Min HH Min EQ Min DR
Area of the condenser [m2] 133.14 149.11 149.96 152.63 TAC [V/yr] 3,491,584 3,619,084 3,625,842 3,646,903
Area of the Heat Exchangers [m2] 60.41 63.44 63.60 64.17 HH [Points] 342,931 334,887 334,952 334,987
Steam [tones] 3.96.108 3.89.108 3.89.108 3.88.108 EQ [Points] 344,555 336,547 336,518 336,586
Electricity [MJ] 1.48.106 1.43.106 1.43.106 1.42.106 DR [Points] 2,941,189 2,877,357 2,876,312 2,873,056
 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377 375

4 Table 6
x10
2850 1.14 Extreme solutions. Case study II: Absorption cycle.

Objective function Min TAC Min HH Min EQ Min DR

Damage to Ecosystem Quality (Points)

Depletion Natural Resources (Points)

2800 1.12
TAC [V/yr] 21,917 22,940 22,951 22,954
Damage to Human Health (Points)

HH [points] 2734 2584 2593 2593

2750 1.10 EQ [points] 2740 2582 2565 2578
DR [points] 11,228 10,568 10,552 10,497

2700 1.08

rectification column and the temperatures of stream 5 and 9. With

2650 1.06
regard to the discrete variables, all of the designs lead to a rectifi-
cation column with one single stage. In the minimum TAC, the
2600 1.04
EQ energy consumption rate in the reboiler and reflux ratio in the
HH
DR
rectification column are larger than in the minimum impact one.
2550 1.02
2.18 2.2 2.22 2.24 2.26 2.28 2.3 2.32 2.34 2.36
As observed, solutions with minimum impact show larger COP
Total Annualized Cost (euro/yr) x 10
4 values and greater exchanger areas. The exchange area in these
solutions is approximately 31% greater than in the minimum TAC
Fig. 10. Total annualized cost vs impact damage categories. Case study II: Absorption design. This is due to the fact that the contribution of the mass of
cycle. steel to the total impact is rather small. Regarding the use of energy,
the minimum TAC solution consumes approximately 8% more
steam and 4% more electricity than the minimum environmental
Percentage with respect to the maximum value (%)

100
impact one. Hence, the impact caused during the operation phase is
99
more significant than that associated with the construction phase.
98 Particularly, the construction of the equipment units contributes
97 around 4% to the total EI. As observed in Table 6, the TAC in the
minimum cost solution is 4.67%, 4.71% and 4.73% lower than in the
96
optimal HH, EQ and DR solutions, respectively. Moreover, HH, EQ
95 and NR can be reduced by up to 5.80%, 6.82%, and 6.96%, respec-
94 tively, compared to the minimum TAC solution.
93
6. Conclusions
92

91 This work has introduced a computational approach for the

TAC HH EQ DR
90 optimal design of thermodynamic cycles considering economic and
1 2 3 4 environmental concerns. Our approach combines simulation
Objective functions
packages with rigorous deterministic mathematical programming
Fig. 11. Parallel coordinates plot. Case study II: Absorption Cycle. Objective functions: tools and LCA analysis. The capabilities of this approach were tested
Total annualized cost (TAC), damage to human health (HH), damage to ecosystem in two thermodynamic cycles: a steam power cycle and an
quality (EQ), and depletion of natural resources (DR). ammonia-water absorption cooling cycle, for which we minimized
the total annualized cost and a set of environmental impacts
minimum environmental impact alternatives. The main difference measured in three LCA damage categories.
concerns the duty provided to the system (140 kW vs. 131 kW). Numerical results showed that the environmental performance
These results are consistent with the fact that the environmental of thermodynamic cycles can be improved by compromising their
impacts are highly dependent on the steam supplied to the reboiler. economic performance. We also found that the main contribution
The extreme designs differ also in the reflux ratio of the to the total impact is the operation phase. The optimization of the
individual damage categories produces similar results, indicating
Table 5 redundancies between them.
Details of the extreme solutions. Case study II: Absorption cycle.

Variable Min TAC Min HH Min EQ Min DR Acknowledgements

Reboiler duty [kW] 140.39 131.79 131.05 131.11
High pressure [bar] 13.28 12.95 12.96 12.95
The authors wish to acknowledge support from the Spanish
Low pressure [bar] 4.48 4.82 4.97 4.93
Mass flow of stream 0.32 0.33 0.33 0.33 Ministry of Education and Science (projects ENE 2011-28269-CO3-
1 [kg/s] 03) and the Spanish Ministry of External Affairs (projects PHB
Ammonia fraction 0.51 0.53 0.54 0.54 2008-0090-PC).
of stream 1
Temperature of 41.15 38.27 38.20 38.16
stream 5 [ C] Appendix A. Life cycle assessment of the thermodynamic
Temperature of 20.20 21.45 22.80 22.08 cycles
stream 9 [ C]
Reflux ratio (mass) 0.055 0.043 0.042 0.042
The environmental impact is quantified following LCA princi-
Number of trays 1 1 1 1
Feed tray 1 1 1 1 ples, in a similar manner as done before by the authors in other
COP 0.63 0.67 0.68 0.68 works [47e50]. The LCA comprises four phases [51]:
Total exchange 89.13 119.77 120.26 120.34
Area [m2] 1. Goal and scope definition. This phase defines the system
Steam [kg] 904,972 838,389 837,074 836,897
Electricity [MJ] 10,895 10,454 10,544 10,542
boundaries, functional unit, assumptions made and type of
impact assessed. The system boundaries correspond to the
376 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377

limits of the energy system. The functional unit is defined as [16] PRé consultants, Damage Oriented Method for Life Cycle Assessment Meth-
odology Report and Manual for Designers, Technical Report, PRé Consultants,
a given amount of power/cool generated. We quantify the
Amersfoort, The Netherlands, 2000.
impact in three categories: damage to human health (HH), [17] U.M. Diwekar, I.E. Grossmann, E.S. Rubin, An MINLP process synthesizer for
damage to ecosystem quality (EQ), and damage due to deple- a sequential modular simulator, Industrial & Egineering Chemistry Research
tion of natural resources (DR). 31 (1) (1992) 313e322.
[18] J.M. Reneaume, B. Koehret, X.L. Joulia, Optimal process synthesis in a modular
2. Life cycle inventory analysis (LCI). This phase quantifies the simulator environment: new formulation of the mixed-integer nonlinear
input and output flows associated with the operation and programming problem, Industrial & Egineering Chemistry Research 34 (12)
construction of the cycles. The damage during the operation (1995) 4378e4394.
[19] Z. Kravanja, I.E. Grossmann, Computational approach for the modelling/
phase is given by the natural gas and electricity consumption decomposition strategy in the MINLP optimization of process flowsheets with
rates, which are retrieved from the process simulation. This implicit models, Industrial & Egineering Chemistry Research 35 (6) (1996)
information is translated into the corresponding LCI using 2065e2070.
[20] M.S. Díaz, J.A. Bandoni, A mixed integer optimization strategy for a large
environmental databases. The LCI of the construction phase is chemical plant in operation, Computer and Chemical Engineering 20 (5)
approximated by the LCI of the mass of steel contained in the (1996) 531e545.
process units. [21] H. Kim, I.H. Kim, E.S. Yoon, Multiobjective design of calorific value adjustment
process using process simulators, Industrial & Egineering Chemistry Research
3. Life Cycle Impact Assessment. This phase translates the LCI into
49 (6) (2010) 2841e2848.
the corresponding environmental damages (denoted by the [22] J.A. Caballero, D. Milan-Yañez, I.E. Grossmann, Rigorous design of distilla-
continuous variable DAMd) using damage factors (dfb,d) avail- tion columns: integration of disjunctive programming and process simu-
lators, Industrial & Egineering Chemistry Research 44 (17) (2005)
able in the literature [16].
6760e6775.
[23] R. Brunet, G. Guillen-Gosalbez, J.R. Pérez-Correa, J.A. Caballero, L. Jimenez,
X Hybrid simulation-optimization based approach for the optimal design of
DAMd ¼ df b;d $LCIb (8) single-product biotechnological processes, Computers and Chemical Engi-
neering 37 (1) (2012) 125e135.
b˛B
[24] Y.Y. Haimes, Integrated Systems Identification and Optimization in Control
and Dynamic Systems: Advances in Theory and Applications, vol. 10,
4. Life cycle interpretation. In this phase, the LCA results are Academic Press, Inc., 1973.
analyzed and a set of conclusions and recommendations are [25] G. Mavrotas, Effective implementation of the ε-constraint method in multi-
objective mathematical programming problems, Applied Mathematics and
formulated. In this work, this step is carried out in the post Computation 213 (2) (2009) 455e465.
optimal analysis of the optimal solutions found. [26] M.A. Duran, I.E. Grossmann, An outer-approximation algorithm for a class of
mixed-integer nonlinear programs, Mathematical Progress 36 (3) (1986)
307e339.
[27] G.R. Kocis, I.E. Grossmann, Relaxation strategy for the structural optimization
of process flow sheets, Industrial & Egineering Chemistry Research 26 (9)
References (1987) 1869e1880.
[28] R. Brunet, G. Guillen-Gosalbez, L. Jimenez, Cleaner design of single-product
[1] I.E. Grossmann, G. Guillen-Gosalbez, Scope for the application of mathematical biotechnological facilities through the integration of process simulation,
programming techniques in the synthesis and planning of sustainable multiobjective optimization, life cycle assessment, and principal component
processes, Computers and Chemical Engineering 34 (9) (2010) 1365e1376. analysis, Industrial & Egineering Chemistry Research 51 (1) (2012) 410e424.
[2] J. Dipama, A. Teyssedou, F. Aubé, L. Lizon-Lugrin, A grid based multi-objective [29] G. Guillen-Gosalbez, A novel MILP-based objective reduction method for
evolutionary algorithm for the optimization of power plants, Applied Thermal multi-objective optimization: application to environmental problems,
Engineering 30 (8e9) (2010) 807e816. Computers and Chemical Engineering 35 (8) (2012) 1469e1477.
[3] A.T. Bulgan, Optimization of the thermodynamic model of aqua-ammonia [30] C. Pozo, R. Ruíz-Femenia, J.A. Caballero, G. Guillen-Gosalbez, L. Jiménez, On the
absorption refrigerations systems, Energy Conversion and Management 36 use of principal component analysis for reducing the number of environ-
(2) (1995) 135e143. mental objectives in multi-objective optimization: application to the design of
[4] J.C. Bruno, J. Miquel, F. Castells, Modeling of ammonia absorption chillers chemical supply chains, Chemical Engineering Science 69 (1) (2012) 146e158.
integration in energy systems of process plants, Applied Thermal Engineering [31] Aspen HYSYS Version 7.1 (23.0.4507). Aspen Technology, Inc., Cambridge,
19 (12) (1999) 1297e1328. www.aspentech.com.
[5] C. Wu, L. Chen, F. Sun, Optimization of solar absorptions refrigerator, Applied [32] Aspen Plus, version 7.1 (0.0.7119). Aspen Technology, Inc., Cambridge, www.
Thermal Engineering 17 (2) (1997) 203e208. aspentech.com.
[6] G.Q. Zhang, L. Wang, L. Liu, Z. Wang, Thermoeconomic optimization of small [33] Matlab 2011b, The MathWorks, Software (2011). Available from: www.
size central air conditioner, Applied Thermal Engineering 24 (4) (2004) mathworks.com.
471e485. [34] K. Holmström, A.O. Göran, M.M. Edvall, User’s Guide for Tomlab/SNOPT v12.1
[7] J.C. Bruno, A. Valero, A. Coronas, Performance analysis of combined microgas (2009).
turbines and gas fired water/LiBr absorption chillers with post-combustion, [35] TOMLAB, Optimization (2009) Avaliable from, www.tomopt.com.
Applied Thermal Engineering 25 (1) (2005) 87e89. [36] P.E. Gill, W. Murray, M.A. Saunder, SNOPT: an SQP algorithm for large-scale
[8] M. Pouraghaie, K. Atashkari, S.M. Besarati, N. Nariman-Zadeh, Thermodynamic constrained optimization, SIAM Journal on Optimization 12 (4) (2002)
performance optimization of a combined power/cooling cycle, Energy 979e1006.
Conversion and Management 51 (1) (2010) 204e211. [37] K. Holmström, A.O. Göran, M.M. Edvall, User’s Guide for Tomlab/CPLEX v12.1
[9] A. Huicochea, W. Rivera, G. Gutiérrez-Urueta, J.C. Bruno, A. Coronoas, Ther- (2009).
modynamic analysis of a trigeneration system consisting of a micro gas [38] M.J. Moran, H.N. Shapiro, Fundamentals of Engineering Thermodynamics SI
turbine and a double effect absorption chiller, Applied Thermal Engineering Version, second ed. Wiley, New Jersey, United States, 1993.
31 (6) (2011) 3347e3353. [39] A.S. Nafey, M.A. Sharif, Combined solar organic rankine cycle with reverse
[10] A. Sapienza, I.S. Glaznev, S. Santamaria, A. Freni, Y.I. Aristov, Adsorption osmosis desalination process: energy, exergy and cost evaluations, Renewable
chilling driven by low temperature heat: new adsorbent and cycle optimi- Energy 35 (11) (2010) 2571e2580.
zation, Applied Thermal Engineering 32 (1) (2012) 141e146. [40] S.M. Walas, Chemical Process Equipment Selection and Design, Butterworth-
[11] H. Becerra-Lopez, P. Golding, Multi-objective optimization for capacity Heinemann, Washington, United States, 1990.
expansion of regional power-generation systems: case study of far west [41] A. Mulet, L.B. Evans, A.B. Corripio, K.S. Chretien, Estimate costs of pressure
Texas, Energy Conversion and Management 49 (6) (2008) 1433e1445. vessels via correlations, Chemical Engineering 88 (20) (1981) 148e150.
[12] X. Pelet, D. Fevrat, G. Leyland, Multiobjective optimisation of integrated energy [42] Mitsubishi AT64C. Catalogue Steam Turbine Generator. Mitsubishi Heavy
systems for remote communities considering economics and CO2 emissions, Industries, LTD.
International Journal of Thermal Sciences 44 (12) (2005) 1180e1189. [43] A. Vidal, R. Best, R. Rivero, J. Cervantes, Analysis of a combined power and
[13] B.H. Gebreslassie, G. Guillén-Gosálbez, L. Jiménez, D. Boer, Design of environ- refrigeration cycle by the exergy method, Energy 31 (15) (2006) 3401e3414.
mentally conscious absorption refrigeration systems via multi-objective opti- [44] K.M. Guthrie, Data and techniques for preliminary capital cost estimating,
mization and life cycle assessment, Applied Energy 86 (9) (2009) 1712e1722. Chemical Engineering 24 (1969) 114e142.
[14] B.H. Gebreslassie, G. Guillén-Gosálbez, L. Jiménez, D. Boer, Economic perfor- [45] O. Kizilkan, A. Sencan, S.A. Kalogirou, Thermoeconomic optimization of a LiBr
mance optimization of an absorption cooling system under uncertainty, absorption refrigeration system, Chemical & Engineering Processing 46 (12)
Applied Thermal Engineering 29 (17e18) (2009) 3491e3500. (2007) 1376e1384.
[15] Eco-Indicator 99. The Netherlands, The Eco-Indicator 99, Manual for [46] M.A. Siddiqui, Economic analyses of absorption systems: part A-design and
Designers, The Hague, The Netherlands, 2000. cost evaluation, Energy Conversion Management 38 (9) (1997) 889e904.
 R. Brunet et al. / Applied Thermal Engineering 48 (2012) 367e377 377

[47] Towards an integrated framework for supply chain management in the batch moMINLP: Multi-objective mixed-integer non-linear programming
chemical process industry L. Puigjaner, G. Guillén-Gosálbez, Towards an inte- moNLP: Multi-objective non-linear programming
grated framework for supply chain management in the batch chemical process MOO: Multi-objective optimization
industry, Computers and Chemical Engineering 32 (4e5) (2008) 650e670. NLP: Non-linear programming
[48] G. Guillen-Gosalbez, J.A. Caballero, L. Jiménez, Application of life cycle NPV: Net present value
assessment to the structural optimization of process flowsheets, Industrial & P: Pump
Egineering Chemistry Research 47 (3) (2008) 777e789. RC: Rectification column
[49] G. Guillen-Gosalbez, I.E. Grossmann, Optimal design and planning of SC: Refrigerant subcooler
sustainable chemical supply chains under uncertainty, AIChE Journal 55 (1) SHX: Solution heat exchanger
(2009) 99e121. SQP: Successive quadratic programming
[50] G. Guillen-Gosalbez, I.E. Grossmann, A global optimization strategy for the TAC: Total annualized cost
environmentally conscious design of chemical supply chains under uncer- TPC: Total plant cost
tainty in the damage assessment model, Computers and Chemical Engi- VLV1: Refrigerant expansion valve
neering 34 (1) (2010) 42e58. VLV2: Solution expansion valve
[51] ISO 14040, Environmental Management. Life Cycle Assessment: Principles and
Framework, International Organisation for Standardisation(ISO), Geneve, Variables
Switzerland, 2006, 2006. Am: Heat exchanger area of unit m (m2)
CF: Fixed cost (V)
Cb: Exchange area (m2)
Nomenclature Cu: Cost of the unit (V)
CO: Operating cost (V/yr)
DAMd: Environmental damages (Points)
Sets/Indices Diam: Diameter of the rectification column (m)
D: decision variables H: Height of the rectification column (m)
E: equality constrains LCIb: Input and output flows (kg/yr)
I: inequality constraints Qu: Heat transfer of unit u (kW)
i: topology TAC: Total annualized cost (V/year)
j: external equality and inequality constraints Wu: Mechanical power of unit u (kW)
j˛ EQ: external (explicit) equality constraint a
Q: Auxiliary variable
j˛ IEQ: external (explicit) inequality constraint : Penalty value for the constraint violation
k: iteration
o: objective functions Parameters
u˛HX: heat exchangers c1: Cost parameter (V/m2)
u˛U: equipment units c2: Cost parameter (V)
c3: Cost parameter (V/kW)
Abbreviations cq: Unitary cost of steam (V/MJ)
A: Absorber cw: Unitary cost of electricity (V/MWh)
C: Condenser crf: Capital recovery factor
COP: Coefficient of performance dfbd: Damage factors (Points/kg)
DR: Depletion of natural resources deuro: Conversion from dollars to euros (V/$)
E: Evaporator Fc: Cost factor that depends on the type of column
EI: Environmental impact fd: Coefficient of the design type
EQ: Damage to ecosystem quality fp: Coefficient of the design type
HH: Damage to human health fm: Coefficient of the material construction
LCA: Life cycle assessment M & S: Cost factor
LCI: Life cycle inventory analysis top: Operational hours (h/yr)
MILP: Mixed-integer linear programming

View publication stats