Computers and Chemical Engineering 94 (2016) 287311

Contents lists available at ScienceDirect

Computers and Chemical Engineering

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

A modeling strategy for integrated batch process development based

on mixed-logic dynamic optimization
Marta Moreno-Benito a, , Kathrin Frankl b , Antonio Espuna
c , Wolfgang Marquardt b
a
Centre for Process Systems Engineering, University College London, Torrington Place, London WC1E 7JE, UK
b
Process Systems Engineering, RWTH Aachen, Turmstrae 46, D-52056 Aachen, Germany
c
Department of Chemical Engineering, Universitat Politcnica de Catalunya, ETSEIB, Av. Diagonal 647, E-08028 Barcelona, Spain

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

Article history: This paper introduces an optimization-based approach for the simultaneous solution of batch process
Received 3 January 2016 synthesis and plant allocation, with decisions like the selection of chemicals, process stages, task-unit
Received in revised form 10 June 2016 assignments, operating modes, and optimal control proles, among others. The modeling strategy is based
Accepted 31 July 2016
on the representation of structural alternatives in a state-equipment network (SEN) and its formulation as
Available online 6 August 2016
a mixed-logic dynamic optimization (MLDO) problem. Particularly, the disjunctive multistage modeling
strategy by Oldenburg and Marquardt (2008) is extended to combine and organize single-stage and mul-
Keywords:
tistage models for representing the sequence of continuous and batch units in each structural alternative
Batch process synthesis
Plant allocation
and for synchronizing dynamic proles in input and output operations with material transference. Two
Dynamic optimization numerical examples illustrate the application of the proposed methodology, showing the enhancement
Generalized disjunctive programming of the adaptability potential of batch plants and the improvement of global process performance thanks
Synchronization to the quantication of interactions between process synthesis and plant allocation decisions.
Multistage modeling 2016 Elsevier Ltd. All rights reserved.

1. Introduction Stephanopoulos and Reklaitis, 2011). Both sub-problems have sev-

Specialty chemical industries are meant to assimilate and
execute many different processes along their lifecycle, as many
specialty goods are subject to quickly changing markets or may
be displaced by rapidly advancing technologies. The fast introduc-
tion of new products and processes into manufacturing facilities
is crucial to be rst in the marketplace. However, such processes
should be also sustainable to ensure a long-term economic via-
bility and maintain their competitiveness in front of stockholders.
The investment of time and resources to design, construct, and vali-
date a new plant each time that a different product is approved for
commercialization is not a viable strategy in most cases. This way,
batch systems are highly appropriate for specialty chemical sec-
tor because their inherent exibility allows their adaptation and
reconguration to incorporate new production lines with none or
only partial plant modications.
After the discovery of a new product and acknowledgment
of its market opportunity, the integrated batch process devel-
opment problem addresses the synthesis of processing schemes
and the allocation of manufacturing facilities (Rippin, 1983;
Corresponding author.
E-mail address: marta.moreno.benito@ucl.ac.uk (M. Moreno-Benito).
eral degrees of freedom closely interrelated, which ultimately
dene the process and the required equipment investment in new
or previously existing manufacturing facilities in grassroots or
retrot scenarios, respectively. For instance, the development of
a batch process stage includes decisions like the operating mode for
determining the process structure i.e. parallel, series or single unit
operation that best suits the process requirements in the batch
system. This decision is closely related to the task-unit assignment
and equipment selection, especially in the case of existing plants
with a given set of units and connecting pipelines, and is crucial to
exploit the adaptability potential of batch systems. Simultaneously,
each equipment unit has to accommodate the assigned tasks using a
batch or semi-continuous procedure, which involves either a chain
of batch operations or an intermittent use of continuously oper-
ated plant elements, respectively. Notably, batch operations are
dened by their duration and dynamic feed-forward trajectories of
control variables, and their feasibility depends on physical con-
straints like the capacity of the selected units. Additionally, it is
necessary to determine the batch size and the number of batches,
as they are responsible for the total demand satisfaction and inter-
relate with the equipment capacity and process conversion. Other
degrees of freedom for full process development include the selec-
tion of process stages, the recirculation of intermediate ows, the
installation of buffer tanks for working with different cycle times or

http://dx.doi.org/10.1016/j.compchemeng.2016.07.030
0098-1354/ 2016 Elsevier Ltd. All rights reserved.
288 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

Notation
Problem sets and indices
Acronyms  set of technological specications indexed by 
AIBN azobisisobutyronitrile j  technological specication of unit j U, |j | = 1, j
AN acrylonitrile U indexed by 
B&B branch-and-bound i set of operating modes in task i PS indexed by
CHR convex-hull relaxation C set of chemical compounds involved into the pro-
CNF conjunctive normal form cess indexed by c
DMF dimethylformamide Cjs C subset of potential reactants, solvents, or catalysts
DO dynamic optimization in unit j U indexed by c
EPC end-point constraints Di, Ui subset of batch units Ui in task i PS whose input
GBD generalized Benders decomposition ow rate is a control variable in conguration
GDP generalized disjunctive programming i . It is dened such that |Di, | = DOFi, |Ui |, where
IFS initial feasible solutions |Ui | represents the degrees of freedom removed by
KPI key performance indicators output ow rates indexed by j
MIDO mixed-integer dynamic optimization Ij Kj subset of input stages for unit j U at Level 1
MILP mixed-integer linear programming indexed by k
MINLP mixed-integer non-linear programming J set of all existing and potential equipment pieces,
MLDO mixed-logic dynamic optimization J = U T Sp Mx indexed by j
MP mathematical programming Ji J subset of equipment pieces within potential task i
NCO necessary conditions of optimality PS indexed by j
NLP non-linear programming Kj set of stages for unit j U at Level 1 indexed by k
OA outer approximation L set of potential stages at Level 0, L = {1, . . ., Lmax }
PC path constraints indexed by k and l
PWC piecewise constant La L subset of active stages at Level 0 indexed by l
RTN resource-task network Lj0 L subset of stages at Level 0 where unit j U can start
SEN state-equipment network
its operation, Lj0 = {1, |Kj | |Oj | + 1 | j =
/ j, j U}
STN state-task network
indexed by l
VA vinyl acetate
Mj set of pipelines for unit j U at Level 1 indexed by
m
General variables
Mjin Mj subset of input pipelines to unit j U at Level 1

z(t) derivatives of differential process variables
z(t) differential process variables indexed by m
y(t) algebraic process variables Mjout Mj subset of output pipelines from unit j U at Level
udyn (t) dynamic control variables 1 indexed by m
ustat time-invariant or static continuous control variables Mx set of existing and potential mixers indexed by j
uint integer control variables N set of pipelines at Level 0 indexed by n
uBool logical or Booleans decisions Ni N subset of pipelines at Level 0 for task i PS indexed
ubin binary decisions by n
 algebraic time-invariant variables Niin Ni subset of input pipelines to process stage i PS
p process parameters indexed by n
Niout Ni subset of output pipelines to process stage i PS
General functions indexed by n
f DAE system Ni,0 N subset of pipelines at Level 0 whose ow rate is
g path constraints restricted to zero in conguration i of task i
ge end-point constraints PS indexed by n
h algebraic equations evaluated at the nal time f
Ni N subset of pipelines at Level 0 for task i PS whose
l set of relations that dene initial conditions f
ow rate is xed by the preceding task, |Ni | = 1
m stage-to-stage continuity between consecutive
except for process stages preceded by a buffer j T
mathematical stages
indexed by n
fd DAE system in disjunctive equations
Nib Ni bypass pipeline for process stage i PS, |Nib | = 1
gd path constraints in disjunctive equations
indexed by n and b
gd,e end-point constraints in disjunctive equations
Nr N subset of pipelines at Level 0 for recirculation
hd algebraic equations evaluated at the nal time in
indexed by n
disjunctive equations
Njin N subset of input pipelines to unit j J at Level 0
ld set of relations that dene initial conditions in dis-
junctive equations indexed by n
md stage-to-stage continuity between consecutive Njout N subset of output pipelines from unit j J at Level 0
mathematical stages in disjunctive equations indexed by n
Bd equations system to dene bypass stages in disjunc- Oj Kj subset of output stages for unit j U at Level 1
tive equations indexed by k
 objective function PC subset of desired products indexed by p
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 289

PS set of potential process stages or tasks indexed by i xc,n,l (t) composition of compound c C in connection n N
Q set of ordered positions that can be assumed by unit and stage l L at Level 0
procedures of j U indexed by q Yj (*) equipment Boolean for processing or storage unit
Sp set of existing and potential splitters indexed by j j UT
T set of existing and potential storage tanks indexed Zi (*) process stage Boolean for task i PS
by j
Tnr T buffer tank for potential recirculation of ow n
N r , |Tnr | = 1, n N r indexed by j batch sizes in successive process stages, the selection from a set of
U set of existing and potential batch units indexed by alternative technologies, and the selection of chemicals.
j The mathematical formulation and the solution procedure
Ui U subset of existing and potential batch units in pro- for addressing this problem are challenging and computationally
cess stage i PS indexed by j demanding activities for several reasons. The batch nature of the
process involves not only a combinatorial assessment for linking
Problem parameters
the equipment and task networks (Gani and Papaeconomou, 2005)
Demp demand of product p P
but also the adjustment of processing conditions along time in each
DOFi, degrees of freedom with regard to the ow rates
process stage. This way, dynamic models are required for repre-
at Level 0 at process stage i PS, according to each
senting the process performance and dynamic proles of control
conguration i
0
variables are necessary to fully explore the attainable region of
lj,q starting stage of unit j U when the task-unit
the problem (Srinivasan et al., 2003). Additionally, the transition of
0 = qth element of the ascend-
Boolean Wj,q is true, lj,q batch operations in batch process stages entails the mathematical
ing sort of Lj0 of unit j U representation of discrete events (Barton et al., 1998). Besides, qual-
Lmax maximum number of stages at Level 0 itative information has to be included regarding decisions like tasks
pj,c values for the set of process parameters pj in unit selection, sequencing and splitting, equipment assignment, and
j U when potential chemical alternative c Cjs is chemicals selection. Finally, batch integrity has to be guaranteed
selected for every structural processing alternative by synchronizing the
material transference between batch and semi-continuous plant
Problem variables (* Control variables) elements as a function of the selected processing scheme and the
Batchp production size associated with each batch of prod- task performance achieved. In batch systems, the synchronization
uct p P of the material transference between consecutive processing units
j requires establishing a link between the dynamic proles of the
Fm,k (t)(*) ow rate of input or output connection m Mjin
input and output ows and the duration of the operations with
Mjout in stage k Kj of unit j U at Level 1 material transference that occur at the same time.
Fn,l (t) ow rate of connection n N in stage l L at Level Given the complexity described above, the problem of batch
0 process development is usually decomposed in sub-problems and
j
int k (t)(*) internal control variable in stage k Kj of unit j U solved independently. Most of the literature follows the nat-
at Level 1 and stage k L of unit j J\U at Level 0 ural decomposition into process synthesis and plant allocation
NBp (*) number of batches of product p P sequence for simplifying the problem. This way, some studies
Rn (*) recirculation Boolean for intermediate ow recycle address the synthesis of batch processes by disregarding the sub-
n Nr sequent assignment of equipment items (Ahmad, 1997; Ali, 1999;
j
Sc (*) compound Boolean for reactant, solvent, or catalyst Barton et al., 1999; Linninger et al., 1995; Papaeconomou, 2005;
c Cjs in unit j U Sharif et al., 2000), while some contributions include allocation
Shortfallp unaccomplished demand of product p P decisions through sequential decision-making procedures (Cavin
Sizej (*) discrete capacity of unit j U T et al., 2004; Iribarren, 1985; Iribarren et al., 1994; Mosat et al.,
tend nal time at Level 0 2008). Other works focus on the optimal operation of individual
Tf total time at Level 0 units through the application of tools like dynamic optimization
tj,end nal time of unit j U at Level 1 (DO). DO or optimal control is a well-received research eld in the
Tj,f total time of unit j U model at Level 1 last decades that addresses operating decisions like the dynamic
tl (*) duration of stage l L at Level 0 proles of control variables (Binder et al., 2001; Srinivasan et al.,
tj,s starting time of unit j U at Level 1 2003), duration of batch operations (Vassiliadis et al., 1994), and
j
tk duration of stage k Kj of unit j U at Level 1 structural decisions that determine the conguration of processing
ts starting time at Level 0 units (Jain et al., 2013; Oldenburg et al., 2003). The design of
jk (t) material volume in stage k Kj of unit j U t Level batch plants is a further area of extensive research (Reklaitis, 1989;
1 and stage k L of unit j J\U at Level 0 Barbosa-Pvoa, 2007) where the allocation of manufacturing facil-
j ities and plant sizing are usually solved using predened recipes
V (*) technology Boolean for specication  j in
processing unit j U or approximated models e.g. the xed time and size factor mod-
Wj,q (*) assignment Boolean for processing order q Q in els to simplify the process stage representation (Robinson and
unit j U Loonkar, 1972). In this case, laboratory procedures used during the
X i (*) conguration Boolean for alternative i in pro- product development phase are scaled-up to pre-specify the infor-
mation related to the batch process synthesis, like the chemicals,
cess stage i PS
j operating conditions or processing times involved in the process.
xc,m,k (t) composition of compound c C in input or output
However, divide and conquer strategies jeopardize the efciency
connection m Mjin Mjout and stage k Kj of unit of the resulting process for several reasons. First, the sequential
j U at Level 1 solution of batch process development involves losing a signi-
cant part of the interaction among the degrees of freedom, thus
290 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
restricting the evaluation of the system trade-offs according to
industrial-scale objectives (Moreno-Benito et al., 2014). The poten-
tial of decisions like temperature proles or task durations for
improving global processing times, processing conditions, and
tasks sequences is only relevant if production-scale restrictions and
objectives are taken into account in the decision-making. Besides,
the targets in bench-scale experiments e.g. reaction yield fre-
quently differ from those of industrial-scale manufacture, which
are usually global objectives that include all the different parts of
the process e.g. the total cost and environmental impact. Addi-
tionally, the direct implementation of xed recipes in existing
equipment may force its operation in extreme and even infeasible
conditions (Grossmann et al., 1987). This way, it is crucial to take
into account the physical restrictions of the given plant to ensure
the feasible execution of the new processes. Overall, to fully opti-
mize the processes and guarantee the best utilization of available
equipment and new investments, process synthesis and plant allo-
cation should be solved simultaneously.
Signicant efforts have been made for combining synthesis and
allocation sub-problems with different degrees of integration over
the last fteen years. For instance, some studies address this prob-
lem using DO and mixed-integer dynamic optimization (MIDO)
formulations to optimize the reference trajectories of control vari-
ables (Charalambides et al., 1995; Sharif et al., 1999), although
they involve the pre-specication of the sequence of process stages
and allocation decisions like the selection of the operating mode
and the task-unit assignment. Other approaches simplify the prob-
lem through the approximation of the batch process behavior
using algebraic models e.g. the named screening models (Allgor
et al., 1999) and posynomial functions (Iribarren et al., 2004). These
approaches have been complemented with the optimization of
the dynamic feed-forward trajectories of control variables (Allgor
and Barton, 1999), albeit using an iterative evaluation of struc-
tural and performance decisions instead of a simultaneous solution
procedure. In contrast, other works opt for constant set-points
for dening the processing conditions while solving the syn-
thesis and allocation problems simultaneously (Chakraborty and
Linninger, 2002; Linninger and Chakraborty, 1999). Additionally,
several authors seek the sustainable design of processing schemes
in retrot scenarios taking into account the previously existing
equipment, while they use heuristic approaches based on path
ow decomposition and indicator-based assessment (Banimostafa
et al., 2012; Bumann et al., 2011; Carvalho et al., 2009; Halim et al.,
2011; Simon et al., 2008). Besides, the representation of the pro-
cess performance using dynamic models in allocation problems has
received a strong impulse in the last two decades in the context
of short-term scheduling (Bhatia and Biegler, 1996; Capn-Garca
et al., 2013; Chu and You, 2014; Frankl et al., 2012; Nie et al.,
2012). However, these contributions do not cover other synthesis
and design decisions such as the selection of process technologies,
chemicals, or equipment sizing, among others. Finally, the syn-
chronization of process stages for linking operations with material
transference in consecutive processing units is addressed in the
context of scheduling and design of batch plants (Ferrer-Nadal et al.,
2008; Furman et al., 2007), bearing in mind that the sequence of
processing units is not known beforehand. The principal objective
of the synchronization in these works is to elude physically incon-
sistent allocation solutions; however, material and energy balances
are evaluated in a unique time point in each operation with mate-
rial transference, instead of considering the dynamic proles of the
input and output ows over a time interval, which characterize
batch processes.
This paper presents a novel modeling strategy to optimize the
integrated batch process development problem using a compre-
hensive mixed-logic dynamic optimization (MLDO) model that
evaluates all the interactions between synthesis and allocation
decisions simultaneously. Our goal is the systematic optimization
of batch processes according to global objectives that embrace the
process as a whole, by taking the maximum advantage of new and
previously existing equipment in grassroots and retrot scenarios.
This way, we provide a tool for the agile development, improve-
ment, or adaptation of chemical processes in rapidly changing and
competitive markets. The integrated batch process development
problem studied here (Section 2) comprises process synthesis deci-
sions like the determination of process stages or tasks and their
splitting into subtasks, the selection of technologies and chemicals,
the denition of feed-forward control trajectories, duration of batch
operations, and intermediate ow recycles, and the synchroniza-
tion of operations with material transference. Simultaneously, the
problem includes plant allocation decisions, namely the equipment
selection, the task-unit assignment, the denition of the operating
mode in single, series, or parallel operation, and the batch sizing.
Finally, equipment capacities may represent either model con-
straints in existing processing items or further decisions on plant
investments. The combination of all these degrees of freedom in an
integrated optimization problem is solved in this paper for the rst
time without simplications.
Essentially, the modeling strategy (Section 3) relies on the
representation of processing alternatives in a state-equipment net-
work (SEN) superstructure (Smith and Pantelides, 1995) and its
mathematical formulation as a MLDO problem (Section 4). In par-
ticular, the combination of hybrid discrete/continuous modeling
and generalized disjunctive programming proposed by Oldenburg
and Marquardt (2008) is extended herein to incorporate:
(i) The combination of single-stage and multistage models to rep-
resent continuous and batch processing elements;
(ii) The denition of multiple representation levels to link the
operations of the different batch units that happen at the same
time, whose ordering is not known beforehand; and
(iii) The synchronization of the ow rates and compositions over
time and the duration of operations with material transfer-
ence, as a function of the selected process structure and tasks
performance.
To the authors knowledge, the combination of these modeling
elements to handle simultaneously all the decisions previously dis-
cussed has not hitherto been applied in the context of batch process
design. In consequence, the major novelty of this work is the order-
ing and combination of the coexisting single-stage and multistage
dynamic models that represent the multiple processing units that
compose alternative processing schemes.
Previous contributions on process optimization have not
addressed the problem solved in this work. For instance, some
authors only study one single process stage (e.g. Oldenburg et al.,
2003) or do not use dynamic variables in the material transference
between equipment units (e.g. Ferrer-Nadal et al., 2008). In other
works, physical constraints of existing equipment units are not
considered, what implies that particular equipment constraints do
not restrict the operation of the selected tasks (e.g. Charalambides
et al., 1995), or each batch process stage is composed of a single
batch operation instead of a chain of operations (e.g. Corsano
et al., 2004). With this last simplication, the whole process is
represented by a unique multistage model by associating one
modeling stage with each task. In our work, all these assumptions
are eliminated by addressing the synchronization of coexisting
single-stage and multistage models. The technical difculty to
perform such a synchronization is the need of linking dynamic
proles of the input and output ows some of them being control
variables, like the input/output ow rates over the duration of the
operations with material transference. This is opposite to linking a
single time point and requires the denition of the corresponding
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
time interval in the time axis of the semi-continuous elements.
Additionally, the sequence of processing units is not known
beforehand, as it depends on structural decisions like the task
selection or the operating mode. The resulting MLDO model of our
overall approach can be solved through several solution methods; a
direct-simultaneous methodology is used in this work (Section 5).
The benets associated with the proposed optimization-based
strategy are illustrated with two numerical examples, namely the
Denbigh competitive reaction system (Section 6) and an acrylic
ber production process (Section 7).
2. Problem statement
This work solves the batch process synthesis and the plant allo-
cation sub-problems simultaneously in single-product campaigns,
in order to dene the process and the required equipment invest-
ment in new or existing manufacturing facilities, while enhancing
the process efciency and plant exibility compared to the use of
xed recipes. The entire problem is dened as follows.
Given:
Planning data: set of products, intermediates, and raw materials,
expected demand for nal products, and maximum time horizon;
Plant diagram: the SEN superstructure of available and poten-
tially installed equipment units for each process stage or task,
pipelines and connection nodes like mixers and splitters;
Task network: potential and mandatory tasks, alternative chem-
icals involved in each process stage i.e. reactants, solvents, or
catalysts , allowed technologies, and possible reuse of interme-
diates in following batches;
Batch process operation: allowed task-unit assignments, batch
operations within each unit procedure, operation to operation
switching conditions, and set of limiting processing conditions of
each equipment unit;
Process dynamics: DAE systems to represent the process behav-
ior in each unit procedure, initial conditions, and set of process
variables and dynamic or time-invariant controls;
Data related to performance evaluation: decision criteria and
parameters to evaluate the objective function e.g. selling price of
nal product, direct cost of raw materials, capital costs, amortiza-
tion rates, operating expenses in processing units, environmental
impact indicators;
the goal is to determine:
Synthesis of processing schemes: selection of process stages
and splitting into subtasks, technological specication, chemi-
cals involved i.e. reactants, solvents, or catalysts , reference
trajectories of the feed-forward control variables, duration of
batch operations composing each process stage, recirculation of
intermediate ows to be used in the next batch, and material
transfer synchronization between tasks i.e. synchronization of
ow rates, compositions, and starting and nal times;
Allocation of manufacturing facilities: selection of processing
and storage units, task-unit assignment, operating mode i.e.
single, series, or parallel operation , and batch sizes;
Plant design: equipment sizing;
such that the adopted performance metrics are optimized. This
problem can be solved in grassroots and retrot scenarios. In the
second case, the already existing equipment units have a prede-
ned size value unless their capacity expansion is considered. For
the sake of completeness, it is worth noting that there are other
process development decisions which are out of the scope of this
291
study, as it is the case of decisions associated with multiproduct
and multipurpose campaigns, e.g. batch order.
3. Modeling strategy
The solution of optimization problems with structural decisions
is usually addressed through the representation of all processing
alternatives in a superstructure, which is later formulated into
a mathematical model, to be nally optimized (Grossmann and
Guilln-Goslbez, 2010). In this procedure, the modeling approach
is a key issue for the successful solution of the problem. This
work presents a novel modeling strategy for representing all the
degrees of freedom and interactions that characterize the inte-
grated batch process development in a holistic optimization model.
The approach is based on a bi-level SEN representation and its for-
mulation as a MLDO problem. The background of mathematical
tools for understanding such a strategy is presented next.
3.1. Background
3.1.1. Superstructure representation
Batch design and scheduling problems require superstruc-
ture representations that combine plant, process, and material
states. The most popular proposals are the state-task network
(STN) (Kondili et al., 1993), the resource-task network (RTN)
(Pantelides, 1994), and the state-equipment network (SEN) (Smith
and Pantelides, 1995). Among them, the SEN entails the repre-
sentation of equipment items and states of transferred material.
This representation enables the direct formulation of equipment
equations and clearly illustrates the dependence of the proles
of material ows on selected processing paths and conversion
of preceding process stages. Precisely, other contributions in the
literature have applied SEN superstructures to solve scheduling
problems with process dynamics (Nie et al., 2012; Nie, 2014).
3.1.2. Hybrid discrete/continuous models
These are dynamic models that represent batch process per-
formance and batch events by combining discrete and continuous
variables (Barton and Pantelides, 1994). Depending on the type
of discontinuities represented, they are classied as: (i) single-
stage models, without discrete events, (ii) multistage models, with
explicit discontinuities, and (iii) general hybrid discrete/continuous
models, with implicit and explicit discontinuities. Hybrid dis-
crete/continuous models have been widely used in DO problems for
determining the optimal control proles of individual processing
units over time (Binder et al., 2001; Srinivasan et al., 2003). In pre-
dened sequences of process stages, dynamic control trajectories
have also been optimized in the eld of batch process synthesis
(Allgor and Barton, 1999; Charalambides et al., 1995; Sharif et al.,
1999), batch plant design (Bhatia and Biegler, 1996; Corsano et al.,
2007), and short-term scheduling (Capn-Garca et al., 2013; Chu
and You, 2014; Frankl et al., 2012; Nie et al., 2012), among others.
3.1.3. Logic-based modeling
The incorporation of logical variables and equations into math-
ematical models allows the combination of quantitative and
qualitative information. The use of logics in optimization problems
was initially proposed by Balas (1985) in the so-called disjunctive
programming, and rapidly became an alternative to mixed-integer
modeling for representing structural decisions (Grossmann, 1985;
Duran and Grossmann, 1986). These kinds of problems were later
formalized as generalized disjunctive programming (GDP) (Raman
and Grossmann, 1991; Floudas and Grossmann, 1994), where qual-
itative decisions are mathematically represented using Booleans
and available process knowledge and heuristic rules are formu-
lated through logical propositions. Mixed-logic formulations with
292 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

dynamic equations have contributed to the solution of several

problems such as model predictive control (Bemporad and Morari,
1999), simultaneous design and control (Bansal et al., 2002), and
scheduling of continuous processes with grade transitions (Prata
et al., 2008). In the context of batch process development, Linninger
and Chakraborty (1999) and Chakraborty and Linninger (2002)
addressed the problem of batch process synthesis and plant allo-
cation including logical rules in the superstructure formulation,
but simplifying the problem through approximated process per-
formance models. In contrast, Oldenburg and Marquardt (2008)
and Oldenburg et al. (2003) presented a MLDO formulation that
combined hybrid discrete/continuous models and GDP for solving
the optimal control proles, conguration, sequencing, and opera-
tion of batch equipment items with structural decisions. However,
the formulation only considered individual units rather than task
sequences, so the ordering and synchronization of processing units
were not approached.

3.1.4. Coexisting single-stage and multistage models for

synchronization purposes
Single-stage and multistage models can be used to repre-
sent batch and semi-continuous procedures in the superstructure,
respectively. Since it is necessary to guarantee the lot integrity,
these kinds of models have to be synchronized in operations with
material transference as a function of the selected sequence of
processing units. This synchronization involves establishing a link
between the dynamic proles of the input and output ows and the
duration of the operations with material transference. However,
currently, there are neither modeling frameworks nor software
tools which can handle coexisting multistage and single-stage
models. Previous works on optimization of sequences of processing
elements with dynamic proles have not addressed the order-
ing and combination of several multistage models associated with
batch units. This is either because the physical constraints of exist-
ing equipment units were not considered (Charalambides et al.,
1995), or because batch process stages were simplied as a single
batch operation and the whole process could be represented by a
unique multistage model i.e. by associating one modeling stage
with each task (Corsano et al., 2004).
3.2. Bi-level SEN representation and MLDO modeling strategy
The proposed modeling strategy is based on the combination of
processing alternatives in a SEN superstructure, which is divided
into two representation levels and formulated mathematically as a
MLDO problem. A basic SEN superstructure illustrates the inter-
connection between the existing and potential equipment units
and the states of material ows according to the plant diagram,
e.g. Fig. 1a. This superstructure representation enables the formu-
lation of physical plant restrictions and procedures of processing
elements as well as the denition of the proles of material ows
as a function of the selected units, their order, and their process
conversion.
The mathematical formulation of the superstructure and all the
associated decisions into a MLDO model comprises disjunctions
and logical propositions, which dene the structural decisions,
as well as hybrid discrete/continuous models referred to as
multistage models in this paper , which determine the pro-
cess performance in each unit. This formulation is inspired by
the approach by Oldenburg and Marquardt (2008) for solving
the optimization of individual units. Here, the MLDO formulation
is extended to combine, organize, and synchronize the multi-
ple equipment units in the SEN superstructure according to the
processing alternative that is selected. With this purpose, each unit
is represented by either a single-stage or a multistage model asso-
ciated with a semi-continuous or batch procedure, respectively.
Fig. 1. Basic SEN superstructure (a) and bi-level superstructure (b) of the motivating
example 1.
Therein, each mathematical stage corresponds to a batch operation.
Moreover, in order to perform the coordination and timing of con-
secutive units, these multistage and single-stage models are nested
in their entirety inside specic terms of the disjunctive equations,
which determine their usage and their order.
For visualizing this concept, the plant elements and their mod-
els are distributed into two representation levels by dividing the
SEN superstructure into Level 1, containing batch units, and Level
0, with connection nodes, storage tanks, and semi-continuous
processing units, as illustrated in Fig. 1b for the previous exam-
ple. Independently of their representation level, all the models are
optimized at the same time. Overall, this modular bi-level repre-
sentation allows the ordering of processing elements as a function
of decisions like the task-unit assignment because, this way, the
time coordinates of batch units within the integrated optimiza-
tion problem can be moved in relation to the time horizon of other
batch units and the complete process. The novel coexistence and
synchronization of single-stage and multistage models are illus-
trated in Fig. 2a. The time links across the different models are
established through disjunctive equations and logical propositions
in the formulation.
Additionally, for enabling the relations across the mathematical
stages of the different models using standard optimization tools,
the equations that compose the optimization problem are refor-
mulated as following detailed:
1. Transformation of single-stage models at Level 0 into mul-
tistage models. The connection between the models of the
different processing units requires the conversion of single-stage
models at Level 0 into multistage models that include all the
possible batch operations at Level 1, as represented in Fig. 2b.
With this purpose, the sets of equations of the original single-
stage models are replicated for every possible new stage, and the
continuity of process variables is guaranteed by incorporating
stage-to-stage relations.
2. Normalization of mathematical stages. Since the several
processing elements at Levels 0 and 1 may be composed of a
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 293

Fig. 2. Multistage and single-stage models of the motivating example 1 to represent

batch and semi-continuous plant elements respectively.

diverse number of mathematical stages with different duration

(see Fig. 2b) which should be solved simultaneously, it is nec-
essary to normalize their time axis. This way, all the stages of
every model can be solved at the same time, as shown in Fig. 2c.
The normalization of time intervals is carried out by multiplying
the differential equations by the stage duration and redening
their derivation interval from 0 to 1. Except for the actual rst
stage, the initial conditions of all other stages should be treated
as degrees of freedom whose values are subject to the fulllment Fig. 3. Synchronization of batch stages and ow rates at Levels 0 and 1 for congu-
of stage-to-stage continuity equations. rations , , , and of the motivating example 1, with |Kj | = 3, j {U1 , U2 }.

3. Explicit denition of stage durations. The normalization of the

time intervals dened in the previous step has a collateral advan-
tage of expressing the stage durations as explicit variables. This
way, it is possible to match the timespans of the mathematical
stages of synchronized operations.
Beyond these reformulations of the model equations, the iden-
tity of each mathematical stage at Levels 0 and 1 is preserved
because it is a key issue for the synchronization of the selected
processing units.
4. Mathematical formulation
The variables and equations that compose the optimization
model for integrated batch process development are detailed in
this section according to the modeling strategy mentioned above
(Section 3). The key sets, parameters and variables required to
understand the formulation are rst presented.
4.1. Notation
4.1.1. Mathematical stages
Mathematical stages k Kj at Level 1 should be linked with
stages l L at Level 0 to synchronize the duration of batch oper-
ations in units j U. However, the number of total mathematical
divisions that are necessary at Level 0 for this purpose depends on
the selected tasks i PS and operating modes i . For instance,
the utilization of a single batch unit U1 in the superstructure of
Fig. 1 could involve the three operations load, hold, and unload of
unit U1 , whereas two additional stages would be required in series
operation to complete the hold and unload operations of a sec-
ond batch unit U2 , as illustrated in Fig. 3a (conguration ) and d
(conguration ). The maximum number of stages Lmax concerning
all processing alternatives is determined by the structural system
composed of all the batch units in series. The total time Tf at Level
0 is then divided into Lmax intervals with different durations. Out of
294 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

Fig. 5. Petri net representing process and bypass stages k Kj in the model of batch
unit j U at Level 1. Circles symbolize stages and bars represent stage-to-stage
transitions.

Fig. 4. Input/output variables and stages of batch unit j U models, considering

min Mjin input and mout Mjout output ows, and c C chemicals. these variables refer specically to equipment units or mathemat-
ical stages, they may be written with subindices j or k respectively.

the resulting set of stages L = {1, . . ., Lmax }, only the subset La com- 4.2. Batch procedures at Level 1
posed by the so-called active stages is in force in each structural
alternative. The connection between Levels 0 and 1 is especially 4.2.1. Batch units
relevant to stages that represent operations with material trans- Let us consider discontinuous units j U located at Level 1 of
ference. Particularly, these are termed input and output stages and the superstructure. Their operation and process performance are
are denoted by Ij Kj and Oj Kj respectively. represented by a |Kj |-stage model. For each unit, a two-term dis-
junction controlled by Boolean Yj is introduced. If Yj is true, this
4.1.2. Input and output ows processing item is selected to allocate one task, and the |Kj |-stage
Additionally, input and output ows of batch units are repre- model in the corresponding disjunctive term is activated. On the
sented at both Level 0 (denoted by n Njin Njout ) and Level 1 contrary, if Yj is false, the unit is not selected and a bypass strat-
egy (Oldenburg and Marquardt, 2008) is applied in order to dene
(denoted by m Mjin Mjout ) and are linked between the two lev-
an equivalent number of |Kj | mathematical stages, termed bypass
els in the formulation. Essentially, these ows are characterized by stages, where no process takes place. Bypass stages are represented
j j
ow rates Fm,k (t) and compositions xc,m,k (t), which are referred to in the Petri net of Fig. 5. They are characterized by stage durations
as input and output variables. As illustrated in Fig. 4, the inows j
tk equal to zero and a set of equations that override the dynamic
and outows of batch units should be restricted to zero in those model of the batch process.
stages that are not input Kj \Ij or output Kj \Oj stages respectively, Overall, the disjunctive equation for batch unit j U is dened
to guarantee the absence of material transmission. by the following form, taking into account that the time interval
has been normalized (see Fig. 2c):
4.1.3. Control variables Yj
There are four types of decision or control variables:
fj,kd (z j,k (t), zj,k (t), yj,k (t), udyn (t), ustat , uint , pj ), t [0, 1], k Kj ,
j,k

1. Dynamic control variables udyn (t): ow rates in input and out- ljd (z j,1 (0), zj,1 (0)),
j d
put connections Fm,k (t), m Mjin Mjout and internal control gj,k (zj,k (t), yj,k (t), udyn (t), u stat
, uint
, p ) 0, t [0, 1], k K ,
j,k j j

j
variables int k (t) over operations k Kj in batch units j U; 
dyn
gj,k (zj,k (1), yj,k (1), uj,k (1), u , u , pj ) 0, k Kj ,
d,e stat int

2. Static or time-invariant decision variables ustat : duration of
zj,k+1 (0) mj,k (zj,k (1)) = 0, k {1, . . ., |Kj | 1},
d
batch operations tl represented by mathematical stages l at Level

0; dyn
j = hdj (zj,|Kj | (1), yj,|Kj | (1), uj,|K | (1), ustat , uint , pj ),
3. Integer decision variables uint : number of batches NBp of prod- 
j
(1)
j
tk , t j,end = t j,s + T j,f

uct p P and discrete sizing Sizej of equipment units j U; T j,f =
4. Boolean decision variables uBool : selection of tasks Zi , i k Kj

PS, processing and storage units Yj , j U T, operating modes

Yj
X i ,  in process stages i PS, processing orders Wj,q , q

j
Q, technological specications V ,  , and chemical com- Bd (z j,k (t), zj,k (t), yj,k (t), udyn (t), ustat , uint , j , pj ) = 0, t [0, 1],
j j,k
j
pounds Sc , c C of processing units j U, and recirculation of , j U,
j
tk = 0, k Kj ,
intermediate ows Rn , n Nr .
j,s
T j,f = 0, t j,end = 0, tj = 0

The consideration of all these decisions may lead to over-

specied problems in some cases. Therefore, it is possible to
improve the efciency of the search procedure and reduce con-
vergence issues, by removing the extra degrees of freedom, which
are mostly connected with logical decisions and dynamic control
variables in operations with material transference. The reduction
of the number of control variables is analyzed at the end of this
section based on the optimization model.
4.1.4. Other variables and parameters
The problem comprises other types of variables, namely differ-
ential variables z(t) and their derivatives z(t), algebraic variables
y(t), time-invariant or static variables , and parameters p. When
d , g d , g d,e , and md are the DAE system, path constraints
where fj,k
j,k j,k j,k
(PC), end-point constraints (EPC), and stage-to-stage continuity in
stage k Kj of unit j U. ljd are the relations to determine the ini-
tial conditions, hdj is the set of equations to calculate time-invariant
variables  j evaluated at the nal time (t = 1) of the last stage
|Kj |, and Bjd are the equations that dene the system in bypass
stages. The dynamic model of unit j U comprises time-dependent
derivatives z j,k (t) and differential zj,k (t), algebraic yj,k (t), and con-
dyn j
trol uj,k (t) variables including input/output ow rates Fm,k (t) and
j j
compositions xc,m,k (t)
and internal control variables int k (t) in
stages k Kj . It is also characterized by process parameters pj and
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 295

Fig. 6. Basic SEN superstructure of the motivating example 2.

time-invariant variables  j , necessary for the objective function and with alternative c Cjs according to the chemical selection Boolean
other key performance indicators (KPIs), like product selectivity or j
Sc . This constraint is dened as follows:
process conversion in j U.

To complete batch unit models at Level 1, input/output variables j

Sc
should be dismissed in those stages which are not input/output  , j U. (5)
c Cs pj = pj,c
stages respectively, as illustrated in Fig. 4, by restricting their value j

to zero. With this purpose, the following equation is dened:

4.3. Synchronization
Yj
j
F in (t) = 0, min Mjin , k Kj \Ij , , j U. (2)
4.3.1. Processing order of batch units
m ,k The time axis for the models of batch units j U that are selected
j
Fmout ,k (t) = 0, mout Mjout , k Kj \Oj (Eqs. (1) and (2) with Yj = true) should be moved over the time axis
of the entire process at Level 0 according to their processing order,
j
Additionally, the volume k (t) at unit j U cannot surpass the as illustrated in Fig. 3. Then, each batch unit model is synchronized
equipment capacity Sizej in any stage k Kj , either Sizej is a free with the models of other plant elements by relating the set of stages
decision variable to design a newly installed unit or an expansion or Kj at Level 1 to specic stages l L at Level 0. Assignment Booleans
it is a xed variable associated with an existing item. This constraint Wj,q are dened to lead the synchronization by indicating the order
is formulated by: q Q of unit j U with reference to other batch units. Once a unit is
selected, this should be assigned one and only one processing order
jk (t) Sizej , t [0, 1], k Kj , j U. (3) q Q:

Yj  Wj,q , j U, (6)
4.2.2. Technological specications qQ
In some cases, different processing technologies  can be used
0 L0 . This way, the rela-
and its operation should start in stage lj,q
in the same unitary operation. The selection between alternative j
technologies requires the representation of different equipment tion between stage k Kj at Level 1 and its corresponding stage l L
arrangements that determine the particular physicochemical equa- at Level 0 can be expressed as a function of the processing order
0 1 L. On this
q Q of unit j U through the function l = k + lj,q
tions, parameters, and batch operations sequences for each case.
Hence, every alternative has to be associated with a specic basis, the variables of models at Levels 0 and 1 are connected as
processing item j , |j | = 1, j U, in order to characterize follows:
its DAE system through Eq. (1). For instance, Fig. 6 shows a SEN
with two technological alternatives A and B in equipment units The starting time tj,s of unit j U (Level 1) is calculated from the
U1,A and U1,B respectively. The technology  j associated with 0
general starting time ts and duration tl of stages that precede lj,q
j
equipment unit j U is represented by Boolean V , which indicates (Level 0) by:
j
whether such specication is selected (V
j
= true) or not (V = false)
Wj,q
and is related to the equipment Boolean Yj as follows:
 0 1 ,
lj,q j U; (7)
j qQ t j,s = t s + tl
Yj V ,  j , j U. (4) l=1

4.2.3. Chemical compound alternatives The duration t j of stage k Kj of unit j U (Level 1) has to be
k
The selection of chemicals like reactants, solvents, or catalysts equal to the duration tl of the corresponding stage l L (Level 0)
c Cjs that are involved in the process of unit j U is represented by by:
j
Boolean Sc , which indicates whether a chemical compound is used Wj,q
j j
(Sc = true) or not (Sc = false). The SEN of Fig. 6 illustrates an example  j
, j U; (8)
with two possible compounds c1s and c2s in unit U2 . Unlike the case of qQ tl = t , l {lj,q
0 , . . ., |K | + l0 1}
j
ll0 +1 j,q
j,q
alternative technologies  j , the choice of a chemical compound
c Cjs does not necessarily affect the complete set of equations in The proles of input and output variables of unit j U (Level
j j
the model of unit j U (Eq. (1)) but only the set of parameters pj . In 1) i.e. Fm,k (t) and xc,m,k (t), c C, m Mjin Mjout have
this case, such parameters are dened by the values pj,c associated to be equal to the proles of the analogous variables at Level
296 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

0 i.e. Fn,l (t) and xc,n,l (t), c C, n Njin Njout . Additionally,

the latter ones have to be xed to zero in the remaining non-
synchronized stages, as well as in the case that unit j U is not
selected. The resulting equation for every pair of equivalent ows
at Levels 0 and 1 (n, m) (Njin , Mjin ) (Njout , Mjout ) reads as:

Fig. 7. Petri net representing active stages l La = {1, . . ., |La |} and bypass stages
Wj,q l L\La = {|La | + 1, . . ., Lmax } for semi-continuous elements at Level 0. Gray elements
Fn,l (t) = F j j represent feasible paths for other locations of stage |La |.
m,ll0 +1
(t), xc,n,l (t) = x
c,m,ll0 +1
(t), t [0, 1],

j,q j,q

qQ
l {lj,q , . . ., |Kj | + lj,q
0 0
1},

Fn,l (t) = 0, xc,n,l (t) = 0, t [0, 1],
l L\{lj,q
0
, . . ., |Kj | + lj,q
0
1} Xi
 ,
Y i Fn,l (t) = 0, t [0, 1], n Ni,0 , l L
j

 Fn,l (t) = 0,
,
j U. (9) i PS. (16)
xc,n,l (t) = 0, t [0, 1], l L

4.5. Plant elements at Level 0

4.4. Process stages
4.4.1. Process stages
Logical variable Zi represents the selection of task i PS and,
therefore, controls whether the input ow n Niin is processed or
whether it is bypassed through pipeline b Nib to the following
task. In the problem superstructure of Fig. 6, this decision is related
to splitter Sp3 . The two exclusive alternatives are dened through
the following disjunction:
The bypass strategy, previously applied to batch units j U (see
Fig. 5), is also used at Level 0 to dene the set of extra stages
L\La , out of the total number of stages Lmax , that should be dis-
regarded in each structural alternative. The reason is that the set
of active stages La representing the sequence of batch operations
in the complete process depend on the selected tasks Zi and their
conguration X i . As shown in Fig. 7, the set of stages l L\La that
are not required at Level 0 are overridden by forcing stage durations
tl to zero and withdrawing dynamic equations and constraints of

Zi Zi Level 0 elements.
 , t [0, 1],
Fb,l (t) = 0 Fb,l (t) = Fn,l (t) (10) 4.5.1. Semi-continuous elements
The time variables at Level 0 are related as follows:
l L, n Niin , b Nib , i PS. 
Tf = tl ,
lL (17)
4.4.2. Operating mode
Out of the set of allowed operating modes or congurations t end = t s + T f ,
i in a process stage i PS, with the corresponding Boolean
where Tf is the total batch processing time which starts at initial
variables X i , only one can be selected. This is dened through the
time ts , ends at nal time tend , and is divided in Lmax intervals with
following proposition:
duration tl . Following the bypass strategy, the stage duration is zero
in extra stages l L\La , whereas it is conned between lower tL and
Zi  Xi , (11)
i upper tU bounds in active ones l La :

The principal purpose of conguration Booleans is to govern the t L tl t U , l {1, . . ., |La |},
selection of equipment items j U, represented by variable Yj , and (18)
bypass stages:tl = 0, l {|La | + 1, . . ., Lmax }.
their processing order q Q, represented by variable Wj,q . A logical
proposition is used to dene each conguration. For example, con-
Additionally, material balances in connecting nodes at Level 0
gurations i = {, , , } which correspond to single operation
relate ow rates Fn,l (t) and compositions xc,n,l (t), c C of connec-
in U1 , single operation in U2 , parallel operation in U1 and U2 , and
tions n N in every stage l L. In particular, mixers j Mx are
series operation in U1 followed by U2 in the example of Fig. 1, are
characterized by several input connections nin Njin , |Njin | > 1, and
dened by:
a single output nout Njout , |Njout | = 1, and are described by:
Xi WU1 ,1 YU2 , (12)

Xi WU2 ,1 YU1 , (13) Fnin ,l (t) = Fnout ,l (t), t [0, 1],
nin N in
j
Xi WU1 ,1 WU2 ,1 , (14)
nout Njout , l L, j Mx, (19)
X i WU1 ,1 WU2 ,2 . (15)

Additionally, conguration Booleans X i enforce specic ow Fnin ,l (t)xc,nin ,l (t) = Fnout ,l (t)xc,nout ,l (t), t [0, 1],
distributions with each operating mode. To do so, the ow rate of nin N in
j
a specic set of connections n Ni,0 in each conguration i
is restricted to zero, according to: c C, nout Njout , l L, j Mx, (20)
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 297

whereas splitters j Sp are characterized by a single input connec- 4.5.2. Recirculation of intermediate ows
tion nin Njin , |Njin | = 1, and several outputs nout Njout , |Njout | > The recirculation of an intermediate material to be used in a fol-
1, and are described by: lowing batch is represented by connection n Nr . In particular, this
decision is controlled by Boolean Rn , which indicates if the inter-
 mediate ow is recirculated (Rn = true) or not (Rn = false). As a result,
Fnin ,l (t) = Fnout ,l (t), t [0, 1], the ow rate can either acquire a value between the lower FnL and
nout N out upper FnU bounds or be canceled:
j

n in
Njin , l L, j Sp, (21) Rn

FnL Fn,l (t) FnU , t [0, 1], l L

xc,nin ,l (t) = xc,nout ,l (t), t [0, 1],

Rn
c C, nin Njin , nout Njout , l L, j Sp. (22) , n Nr . (25)
Fn,l (t) = 0, t [0, 1], l L

It is not necessary to remove these equations in bypass stages The recirculation of intermediate ows additionally requires the
because all ow rates would be forced to zero through previous installation of buffer tanks j Tnr in the superstructure e.g. tank
Eqs. (9), (19) and (21). r in the example of Fig. 6 in order to store the intermediate
T11
Storage units j T are also dened at Level 0 by |L|-stage mod- material that is later supplied in a subsequent batch. The logical
d , ld , g d , g d,e , hd ,
els. Such models comprise the general functions fj,l j j,l j,l j proposition reads as:
and mdj,l for storage tanks j T and mathematical stages l La . Rn Yj , j Tnr , n N r . (26)
Since storage tanks operate continuously, all the equations are
d = fd
repeated in active stages as dened by fj,l d = gd
, gj,l , and In order to dene mathematically the sequence of storage and
j,l+1 j,l+1
d,e d,e usage of intermediate material, initial conditions zj,l (t = 0) in the
gj,l =
gj,l+1 , l {1, . . ., |La | 1}.
Following, the bypass strategy is rst stage l = 1 of unit Tnr are set to be equal to the nal conditions
applied to two situations: First, if storage unit j T is not selected, zj,l (t = 1) in the last stage l = Lmax , what reads as:
its complete model is deactivated like in the case of batch units

j U. Second, the bypass method is required to differentiate active Rn

La and bypass L\La stages according to Fig. 7, provided that this unit , n Nr . (27)
zj,1 (0) = zj,|Lmax | (1), j Tnr
is selected. The resulting formulation of storage tanks j T reads
as:
Yj 4.6. Batching
d
fj,l
dyn
(z j,l (t), zj,l (t), yj,l (t), uj,l (t), ustat , uint , pj ), t [0, 1], l {1, . . ., |L |},
a
4.6.1. Number of batches

ljd (z j,1 (0), zj,1 (0)), The batching problem consists in the division of the total prod-

uct demand into a number of batches of a specic production size.

dyn
d
gj,l (zj,l (t), yj,l (t), uj,l (t), ustat , uint , pj ) 0, t [0, 1], l {1, . . ., |La |},
Common approaches for solving scheduling problems rst address
d,e
gj,l
dyn
(zj,l (1), yj,l (1), uj,l (1), ustat , uint , pj ) 0, l {1, . . ., |La |}, the batching activity and following solve the allocation, timing, and

zj,l+1 (0) mdj,l (zj,l (1)) = 0, l {1, . . ., |La | 1}, task sequencing sub-problems. In this work, the simultaneous solu-

(23) tion of batch process development incorporates the optimization
dyn
j = hdj (zj,|La | (1), yj,|La | (1), uj,|La | (1), ustat , uint , pj ),
of the number of batches by means of the following equation:
bypass stages:
dyn
Bjd (z j,l (t), zj,l (t), yj,l (t), uj,l (t), ustat , uint , j , pj ) = 0, t [0, 1], NBp Batchp + Shortfallp Demandp , p P, (28)
l {|La | + 1, . . ., Lmax }
where Demandp is the total demand of product p P, NBp repre-
Yj
sents the number of batches, Batchp stands for the batch production
size, and Shortfallp is the unaccomplished demand.
 Bjd (z j,l (t), zj,l (t), yj,l (t), udyn
j,l
(t), ustat , uint , j , pj ) = 0, , j T,
t [0, 1], l L 4.7. Objective function and degrees of freedom

dyn The optimization problem presented above is solved to mini-

where z j,l (t), zj,l (t), yj,l (t), and uj,l (t) are the time dependent
mize the objective function :
derivatives and differential, algebraic, and control variables in
mathematical stage l L, ustat are time-invariant control variables,
namely stage durations, uint are integer decisions such as the stor-
age tank size, pj are process parameters, and  j are time-invariant
variables which may contribute to the evaluation of the objective
function or KPIs in unit j T. Finally, a constraint to limit the max-
imum volume should be also formulated, which reads as:
jl (t) Sizej , t [0, 1], l L, j T,
where jl (t)
is the inventory volume of unit j T in stage l L and
Sizej is the maximum capacity of the tank. The decision on the selec-
tion of intermediate tanks is crucial for allowing the recirculation
of intermediate ows as well as for dening more than one cycle
time or batch size for different parts of the process.

minimize  z(t), z(t), y(t), udyn (t), ustat , uint , uBool , , p , (29)
udyn (t), ustat
uint , uBool
dened by derivatives z(t), differential variables z(t), algebraic
variables y(t), time-invariant variables  and parameters p,
j j
according to the control variables udyn (t) = {Fm,k (t), int k (t)},
(24)
j
ustat = {tl }, uint = {NBp , Sizej } and uBool = {Zi , Yj , X i , Wj,q , V ,
j
Sc , Rn }.
The degrees of freedom of the optimization model are the num-
ber of decision variables remaining after subtracting the number
of equations and predened values. This parameter determines
the number of practical control variables in the reality. In the
298 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
j j
case of Booleans uBool = {Zi , Yj , X i , Wj,q , V , Sc , Rn }, the number of
degrees of freedom is limited by logical propositions (Eqs. (4), (6),
(9), (11), (12)(15), and (26)). As for the dynamic control proles of
j convert the time-continuous optimization problem into a nite-
ow rates in batch units Fm,k (t) udyn (t), these proles are interre-
lated through global balances in mixers and splitters (Eqs. (19) and
(21)). Therefore, the evaluation of degrees of freedom of dynamic
proles between consecutive units is especially relevant because
the input and output ow rates of different units j U that are syn-
chronized cannot behave as decision variables simultaneously. In
fact, the system over-specication deteriorates the performance of
the solution search procedure.
The degrees of freedom related to ow rates are denoted by
DOFi, and differ for each conguration i and process stage
i PS because they depend on the total number of ow rates
at Level 0 |Ni |, the number of equations in splitters and mixers
|Spi Mxi |, and the number of predened ow rates, namely the
ow restrictions |Ni,0 | in conguration i and input ows |Ni |
dened in preceding tasks. Overall, the parameter DOFi,
puted as follows:
DOF i, = No. variables No. equations No. xed variables
f
= |Ni | |Spi Mxi | |Ni,0 | |Ni |.
To adjust the number of control variables according to the
degrees of freedom of each structural alternative, outows of batch
units are always considered as free-decision variables:
j
Fmout ,k (t) udyn (t), t [0, 1],
mout Mjout , k Oj , j Ui , i PS,
while only the inows of the subset of batch units Di, Ui are
dened as control variables in conguration i of process stage
i PS. The set Di, is dened such as |Di, | = DOFi, |Ui |, where |Ui |
corresponds to the degrees of freedom assigned to the output ows
in Eq. (31). The resulting equation reads as:
Xi
j
 F in (t) udyn (t), t [0, 1], , i PS.
i m ,k
min Mjin , k Ij , j Di,
5. MLDO solution
Different kinds of solution strategies can be used to tackle
the integrated batch process development formulated as a MLDO
problem (Oldenburg and Marquardt, 2008; Stein et al., 2004). The
majority of deterministic methods guarantee to nd a local opti-
mum and most advanced algorithms are also able to provide global
solutions under fairly general assumptions, e.g. (Beaumont, 1990;
Sahinidis, 1996). In particular, deterministic approaches are based
on the reformulation of the mixed-logic problem into a mixed-
integer one, the reformulation of Booleans as continuous variables,
or the problem solution with logic-based solution techniques. The
methods of the rst group are also known as classical solution
approaches and are the most extended ones. This is because the
problem reformulation allows exploiting the advantages of mixed-
logic modeling e.g. the natural inference of previous knowledge
and rules about the process and their direct conversion into mathe-
matical equations as well as using well established MIDO solution
strategies.
Among the classical methods for solving MIDO problems, we
nd three different groups, namely: (i) Dynamic Programming
(DP) based on HamiltonJacobiBellman formulation to trans-
form the original problem into a system of partial differential
equations (Luus, 1990), (ii) indirect methods through necessary
conditions of optimality (NCO) derived from Pontryagins for-
mulation (Bryson and Ho, 1975), and (iii) direct methods which
dimensional nonlinear programming problem by discretization.
Notably, the advantage of the direct approaches is that they do not
require an explicit derivation of the necessary conditions of opti-
mality and the adjoined equations and, after the discretization of
time-dependent variables of the original problem, provide a math-
ematical programming (MP) problem. There exist different direct
methods, depending on what parts of the problem are discretized.
In the specic case of the direct-simultaneous strategy (Neuman
and Sen, 1973; Tsang et al., 1975), there is a full discretization of
the dynamic model, as both process and control variable proles
are approximated by a series of nite points along the time horizon.
The resulting problem is a mixed-integer non-linear programming
f
(MINLP) model that can be solved without requiring sensitivities,
second order derivatives computation, and obligatory variable con-
is com- tinuity, as opposed to direct-sequential approaches (Sargent and
Sullivan, 1978). However, their disadvantage is that the model size
can increase exponentially in the discretization step.
(30) This work employs the classical direct-simultaneous strategy
detailed in the next subsections. The reader interested in alter-
native stochastic and hybrid solution strategies is addressed to
Moreno-Benito et al. (2015).
5.1. MLDO reformulation
To address the relaxation of the original MLDO model
(31)
into MIDO one in classical solution methods, several trans-
formations are  required in  the formulation. First, Boolean
variables uBool true, false are replaced by binaries ubin
 
0, 1 . Thus, the vector of logical decision variables uBool =
j j
{Zi , Yj , X i , Wj,q , V , Sc , Rn } is transformed into a vector of binary
j j
variables ubin = {zi , yj , xi , wj,q , v , sc , rn }, whose 0 and 1 values cor-
respond to prior false and true values respectively.
Second, disjunctive equations that enclose other constraints are
(32)
transformed into mixed-integer equations. In the MLDO formula-
tion presented in the previous section, these equations are Eqs.
(1), (2), (5), (7)(10), (16), (23), (25), (27) and (32). There are sev-
eral alternatives for performing this transformation, like the big-M
(Raman and Grossmann, 1991) and the convex-hull reformulation
(CHR) (Trkay and Grossmann, 1998). In this work, we use binary
multiplication. This technique is based on the decomposition of all
the variables that are contained in the disjunctive terms into vari-
able contributions, except for control variables udyn (t), ustat , and
uint . In particular, the variables that are decomposed are of the fol-
lowing types: derivatives z(t), differential z(t), algebraic y(t), and
time-invariant  variables. After their decomposition, each variable
contribution is associated with the constraints inside a disjunctive
term and its corresponding Boolean variable. For instance, let us
consider the following disjunctive equation:
Ad
 , Ad {true, false}, x R, u R, (33)
dD Fd (x, u) 0
where D is the set of disjunctive terms in the equation, Ad are the
Boolean variables that control each of disjunctive terms d D, Fd
are the functions nested in each disjunctive term d D, x are the
dependent variables, and u are the decision variables. Through sub-
stituting the Booleans Ad {true, false} by the binaries ad {0, 1}
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
and applying binary multiplication, the variables x are decomposed
into variable contributions xd and Eq. (33) is reformulated as:
Fd (xd , u) 0,

x= xd ad ,
dD
ad = 1,
dD
ad {0, 1}, x R, u R.
Finally, logical propositions, namely Eqs. (4), (6), (11)(15) and
(26) of the formulation, are expressed as linear constraints. Such a
transformation can be done systematically by formulating the con-
junctive normal form (CNF) of the original logical equations. This
way, it is possible to obtain an expression like C1 C2 . . . CN ,
where Cn , n {1, . . . N} are the clauses that must be true in the prob-
lem, related by and operators (). This procedure involves the
application of a series of pure logical operations, which were rst
formalized by Clocksin and Mellish (1981), as reported by Raman
and Grossmann (1991). The operations are detailed in Appendix
A. Next, the resulting true-clauses Cn can be transformed into
algebraic equality or inequality equations using a number of equiv-
alences between logical and algebraic expressions (Raman and
Grossmann, 1991), which are summarized in Table 5 of Appendix A.
5.2. MIDO solution
The solution of the resulting MIDO problem is addressed
through the full discretization of the model using the orthogonal
collocation method. This technique was originally introduced by
Cuthrell and Biegler (1989) to solve optimal control problems
with discontinuous control proles, and renders attractive stabil-
ity, symmetry, and accuracy properties (Cuthrell and Biegler, 1989,
Section 3.1). The strategy consists of dividing the time axis into sev-
eral intervals referred to as nite elements e {1, . . ., Ne } and
specic time points termed collocation points m {1, . . ., M}
and approximating the state and control variable proles in these
locations, as represented in Fig. 8. The position of the collocation
points is determined by computing the roots of an orthogonal poly-
nomial, like the shifted Legendre polynomial of order M. Then, the
state and control variable proles are approximated using mono-
mial basis representations, such as Lagrange polynomials.
Further details about the calculation of the collocation points
and the construction of the orthogonal polynomials used in the
numerical examples are provided in Appendix B. As illustrated in
Fig. 8, boundary conditions are dened for differential variables
z(t) in each nite element e to enforce their continuity from the
previous element e 1 in time te with the symmetry properties of
the method. For this, it is necessary to establish a larger order Mth
Fig. 8. Finite elements discretization for differential z(t), algebraic y(t), and control udyn (t) variable proles, where continuity is only enforced for the rst ones. Diamonds
represent the approximations of z(t), y(t), and udyn (t) and triangles represent the estimates of the derivatives z(t)
collocation points m {0, 1, . . ., M, M + 1}.
299
in the polynomials that dene the differential variables, compared
to the polynomials of algebraic y(t) and control variables udyn (t),
with order (M 1)th (see Eqs. (B.2) and (B.3) in Appendix B). In
contrast, continuity across neighboring elements is not necessary
for algebraic and control variables. In fact, control variables are
approximated to a piecewise constant (PWC) function.
(34) Overall, the quality of the approximation of the dynamic model
is determined by the number of collocation points M and nite ele-
ments Ne , and so is the complexity of the resulting MINLP problem.
The number of collocation points is a critical parameter because it
denes the order of the resulting polynomials. It is usually selected
considering the trade-off between being large enough for securing
a good approximation of the discretized variables and small enough
for limiting the increase in the computational demand associated
with high-order polynomials. As for the number of nite elements,
this factor can be adjusted according to the length of the time inter-
val to be discretized. In this work, the full discretization approach
is applied to each of the mathematical stages k Kj , j U and k L
that compose the multistage models of the formulation using a pre-
dened number of nite elements and collocation points. Since the
duration of the batch operations is a decision variable, the length
of the nite elements can vary after scaling the time intervals of
the optimal solution according to their nal values. In order to
guarantee an accurate discretization, a large number of nite ele-
ments is selected in this work, taking into account the upper bounds
of the batch times. Alternatively, an adaptive strategy could be
implemented, by adjusting the number of nite elements for the
discretization of each mathematical stage of the formulation iter-
atively according to their specic duration. This way, it could be
possible to reduce the model size.
The originated MINLP problem can be solved through sev-
eral search algorithms available in literature (Grossmann, 2002).
Their strategies are either based on enumeration, such as branch-
and-bound (B&B) methods with non-linear programming (NLP)
sub-problems (Nemhauser and Wolsey, 1988), or on decomposi-
tion, such as outer approximation (OA) (Duran and Grossmann,
1986) or generalized Benders decomposition (GBD) (Geoffrion,
1972). On the one hand, enumeration strategies like the B&B non-
linear global optimization solver BARON (Sahinidis, 1996) can
guarantee global solutions under general assumptions like the
availability of nite lower and upper bounds on the variables and
expressions in the MINLP to be solved. However, these approaches
also have the disadvantage of rendering costlier computational
loads because they operate through the systematic enumeration of
integer combinations and have to address a NLP problem for each
integer solution. On the other hand, the advantage of decomposition
approaches is precisely their higher efciency, since the problem
is iteratively solved through mixed-integer linear programming
(MILP) master problems with linearized equations and NLP primal
in the nite elements e {1, . . ., Ne } and boundary and
300 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
problems with xed integer variables. The limitation, in this case, is
that decomposition approaches only guarantee local optima unless
the nonlinear objective function and constraints are convex.
Many of the MINLP solution algorithms have been implemented
in commercial software and operate through modeling interfaces
like GAMS (Rosenthal, 2012). After comparing the performance of
different solvers for optimizing the numerical examples of Sec-
tions 6 and 7, the OA solver DICOPT has been selected as the most
suitable alternative to be used in this study due to its efciency. The
principal disadvantage met using global enumeration solvers is the
k (t) in stages k {1, 2, 3} of batch units
difculty to nd feasible solutions after more than 50 h of compu-
tation, in front of the attainment of optimal solutions, which are
possibly local but already improve the results of non-integrated
problems, in not more than 3 h with DICOPT.
Global optimality cannot be guaranteed with the OA algorithm
due to the existence of non-convex terms in the model, e.g. bilinear
functions in mixers. Hence, it is desirable to repeat the optimiza-
tion procedure using different initial feasible solutions (IFSs), with
the intention of identifying accidental local optima. In this work, in
order to generate IFSs effectively, the reduced problem is solved in a
preliminary step by xing feasible combinations of structural vari-
ables, namely the selection of process stages Zi , equipment units Yj ,
operating modes X i , task-unit assignments Wj,q , technologies V ,
j of the already existing equipment items, which are incorporated
j
compounds Sc , and recirculation ows Rn . Additionally, constant
proles of the control variables can be used. The combination of fea-
sible structural decisions in the IFSs is chosen randomly, given that
the starting point for the solver does not regulate the optimal con-
guration obtained with the full-space MINLP search. Ultimately,
the objective is to start the search from different search space loca-
tions.
6. Numerical example 1: Denbigh reaction system
The numerical example presented in this section addresses the
introduction of a new chemical into an already existing plant by
k (t) within each mathematical
addressing the integrated batch process development problem. The
optimal solution is compared to a standard predened recipe to
quantify the potential of the proposed approach. Specically, the
process consists of a competitive reaction mechanism, the Denbigh
reaction system (Denbigh, 1958):
(35)
whose kinetic and thermodynamic data are available in Table
S1 (see Supplementary Material). The SEN superstructure of the
example corresponds to the reactor network of Fig. 1, comprising
two batch reactors U1 and U2 with a capacity of 1 m3 . Since the
investment in new equipment units and pipelines is initially not
contemplated, the operating modes that can be selected are: oper-
ation in one single unit U1 (conguration ) or U2 (conguration ),
parallel in-phase operation in U1 and U2 (conguration ), or series
operation in U1 followed by U2 (conguration ). Batch procedures
in the reactors are composed of load, hold, and unload operations
and the control variables are the input and output ow rates and
the reaction temperature.
6.1. Optimization problem
The production target is to fulll a demand of 21 t of product
S minimizing the raw material expenses within a time horizon of
k (t), k {1, 2, 3}, j {U1 , U2 };
144 h. Additionally, a penalty is applied to the product shortfall.
This way, the optimization problem is dened by:
minimize  = Cost A + Penalty, (36)
udyn (t), ustat ,
uint , uBool
j j
where udyn (t) comprise input F1,k (t) and output F2,k (t) ow rates
j
and reaction temperature
j {U1 , U2 }, ustat
is the duration tl of mathematical stages l, uint is
the number of batches NBS , and uBool represent the Booleans for
equipment units Yj , unit ordering Wj,q , and congurations X 1 . The
rst term of the objective function CostA stands for raw material
expenses of the complete production campaign and is computed
in base of a raw material price p A of 4.8 cD /kg (for further details,
see Eqs. (S16) and (S21) in Section S1 of the Supplementary Mate-
rial), whereas the Penalty term represents the economic charge of
unaccomplished demand, calculated considering twice the selling
price p S of product S, with a value of 43.1 cD /kg (see Eqs. (S17) and
(S23)). The complete MLDO model for solving this example is pro-
vided in the Supplementary Material, including the specications
into the optimization problem as operating constraints.
6.2. Problem solution
The MLDO problem is reformulated into a MINLP one using 32
nite elements and three collocation points in normalized Legendre
roots, according to the direct-simultaneous strategy explained in
Section 5. This is implemented in GAMS v.24.7.1 (Rosenthal, 2012)
and solved through the decomposition-based OA solver DICOPT
(Duran and Grossmann, 1986), using CONOPT 3.17A and CPLEX
12.6 to handle the NLP and MILP sub-problems respectively. In this
example, four IFSs are rst generated using constant proles for the
j j j
control variables F1,k (t), F2,k (t), and
stage k {1, 2, 3} of units j {U1 , U2 } and xing the congura-
tion Booleans X 1 = true for each of the four possible congurations
{, , , }. Next, the complete MINLP problem is solved with
the prior IFSs and with no IFS. As seen in Table 1, which summarizes
the features of the resulting MINLP model implemented in GAMS,
the best solution is obtained by solving the full MINLP problem
initialized with IFS 4.
6.3. Results and discussion
The optimal solution of the integrated problem is compared
with a xed recipe that has a predened production size of
300 kg/batch of product S and requires 70 batches and a batch
cycle time of 2.06 h/batch to fulll the demand of 21 t in its entirety
within the time horizon of 144 h. The operating mode is xed to
conguration , setting Boolean decisions X1 , YU2 , and WU2 ,1 with
a true value. Unit U2 is dened to operate with maximum input and
output ow rates (7.7 m3 /h) during load and unload operations and
maximum reaction temperature (110 C) to guarantee that the total
demand is satised. The optimization problem is solved for three
cases with different degrees of freedom to track the effect of the
structural and the performance decisions:
Dynamically optimal recipe: optimization of dynamic control
proles with xed conguration ;
Structurally optimal recipe: optimization of structural decisions
j j
with constant proles of control variables F1,1 (t), F2,3 (t), and
j
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
Table 1
Features of the MINLP models implemented in GAMS in the Denbigh example, obtained using 32 nite elements and three collocation points in the discretization step: IFS
solutions with xed congurations X 1 = true, {, , , } and complete MINLP problem initialized with IFS 4.
No. equations No. continuous variables
IFS 1 (xed )
IFS 2 (xed )
103,209 98,339
IFS 3 (xed )
IFS 4 (xed )
MINLP with IFS 4 102,633 98,321
Fig. 9. Optimal solution in the Denbigh example: operating mode and synchroniza-
tion of unit procedures.
Optimal recipe i: simultaneous optimization of dynamic proles
and structural decisions according to the problem statement.
In all cases, the number of batches and the duration of batch
operations are free decision variables.
6.3.1. Results
The optimal operating mode for both optimizations that include
structural decisions is series conguration , as shown in Fig. 9a.
This conguration activates and synchronizes the 3-stage models
of units U1 and U2 , such that the unload operation of the rst unit
corresponds to the load operation of the second one. For the case
of the optimal recipe i, the synchronization of batch operations is
detailed in Fig. 9b, which illustrates the mathematical stages at
Levels 1 and 0 for the production of one batch. The trajectories
j j j
of dynamic control variables F1,1 (t), F2,3 (t), and
k (t), k {1, 2, 3}, lead to their upper bound most of the load/unload time intervals
j {U1 , U2 } are presented in Fig. 10(b1c1, b2c2), showing the
evolution of the control variable proles range between their lower
and upper bounds. The synchronization of operations with mate-
rial transference is illustrated through the equivalence between the
output ow rate F3,2 U1 from unit U and the input ow rate F U2 to
1 1,1
unit U2 . Additionally, Fig. 10(b3c3, b4c4) shows process variables
like the reaction volume and molar compositions. All of them are
compared to the trajectories of the xed recipe in batch unit U2
(Fig. 10(a1a4)).
6.3.2. Improvement extent
The incentives to improve the process performance through the
proposed modeling strategy are assessed by comparing the raw
material savings obtained with the xed recipe. The results, sum-
marized in Table 2, show a 12% reduction in the consumption of A by
301
No. discrete variables Non-zero elements Non-linear terms Solution time (s)
259
226
6 346,791 159,578
581
276
10 345,639 159,578 4294
optimizing the dynamic proles with a predened conguration ,
whereas the expenditure decreases as far as a 24% when qualitative
decisions are considered as degrees of freedom with constant pro-
les. Going a step further, this improvement becomes a 25% when
all decisions are optimized simultaneously in the optimal recipe i.
The reduction of raw material expenses is related to the increase
in the product selectivity, from 0.448 in the xed recipe to 0.507,
0.593, and 0.597 in the three optimal solutions (see Table 2). Such an
increase is reected in a molar fraction of product S at the nal time
which is higher in the optimal solution (Fig. 10c4) than in the xed
one (Fig. 10a4). At the same time, the consumption of intermediate
R rises signicantly in the optimal recipe, whereas the generation
of byproducts like T is considerably lower.
6.3.3. Key decision variables
Essentially, the improvement of product selectivity is due to the
optimization of the dynamic temperature prole and the arrange-
ment of the two reaction units in series (conguration ). On the
one hand, the temperature prole in the optimal recipe starts at
the lower bound 50 C in the unit procedure of U1 (Fig. 10b1), in
order to enhance the selectivity of R with respect to T by favoring
reaction 1 (Ea,1 = 1000 kcal/kmol, see Table S1 in the Supplemen-
tary Material) with respect to reaction 2 (Ea,2 = 2580 kcal/kmol). The
prole gradually increases to the upper bound of 80 C in U1 , thus
favoring reaction 3 (Ea,3 = 1800 kcal/kmol) with respect to reaction
4 (Ea,4 = 1210 kcal/kmol) to improve the selectivity of the desired
product S instead of U. In the unit procedure of U2 , the tempera-
ture prole rises even more (Fig. 10c1), reaching the upper bound
of 110 C in this unit. On the other hand, production is restricted
in all recipes by the maximum time horizon of 144 h (see Table 2)
and by the available equipment capacities of 1 m3 in U1 and U2
(Fig. 10(a3c3)). Therefore, series conguration favors the pro-
cess performance by increasing the reaction volume.
6.3.4. Dynamic controls in operations with material transference
In contrast, the use of dynamic control variables in opera-
tions with material transference barely affects the solution. The
feed-forward trajectories of ow rates in the optimal recipe are
(Fig. 10(b2c2)). Thus, in this example, load and unload operations
tend to be as rapid as the pumping capacity permits. The dynamic
feed-forward prole of reaction temperature is also kept constant
in operations with material transference (Fig. 10(b1c1)). In con-
trast, the dynamic control proles are exploited in other batch
operations.
6.3.5. Trade-offs among KPIs
Since heating costs are not included in the objective function, the
interaction between the selectivity of product S and the processing
expenses does not affect the optimal temperature proles in this
example. Regardless, processing costs associated with heat con-
sumption also diminish (see Table 2) thanks to the reduction in
energy consumption in side reactions 2 and 4, which are endother-
mic. The rise of S selectivity also goes hand in hand with a reduction
302 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

Fig. 10. Control and process variable proles in the xed and optimal recipe i in the Denbigh example.

of reactant consumption and to a decrease in the total amount of
reaction mixture required in the system. Consequently, the heat
needed for achieving the reaction temperature is smaller and the
optimal recipe benets from a collateral reduction of heating costs
of a 15%. Besides, since new investments are not considered, null
amortization values are computed in all recipes. Moreover, due to
a signicant increase in occupation costs by using two reactors,
the total prot is higher in the dynamically optimal solution with
xed conguration than in the optimal recipe i with operating
mode . This way, improvements of the 11.9% and 6.5% compared
to the prot of the xed recipe are obtained, respectively. These
results reect the duplication of labor costs, such as unit cleaning
and rental.
6.4. Adaptability potential
In order to evaluate the potential of the proposed strategy
to enhance the batch plant exibility and adaptability, the opti-
mal recipe is solved according to other decision criteria, namely
the prot (case ii) and protability (case iii) maximization. These
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 303

Table 2
KPIs and individual economic weights in the Denbigh example: xed recipe, dynamically optimal recipe in U2 , structurally optimal recipe with constant control proles,
and simultaneous optimal recipes for all degrees of freedom in the problem statement in cases: (i) raw material cost minimization, (ii) prot maximization, (iii) protability
maximization, and (iv) prot maximization with capacity expansion. Items in bold correspond to the objective function.

KPI Fixed Dynamically Structurally Optimal Optimal Optimal Optimal

recipe optimal optimal recipe i recipe ii recipe iii recipe iv

Equipment conguration  
No. batches 70 54 47 47 23 31 13
Batch size [kg/batch] 300 389 447 447 913 677 1615
Total processing time [h] 144.03 144.03 144.03 144.03 144.03 66.49 144.03
Batch processing time [h/batch] 2.06 2.67 6.00 5.98 6.26 2.14 11.08
Batch cycle time [h/batch] 2.06 2.67 3.06 3.06 6.26 2.14 11.08
Total prot [D ] 4,764 5,333 5,080 5,096 5,841 4,861 5,965
Total revenue 9,046 9,046 9,046 9,046 9,046 9,046 9,046
Raw material cost 2,250 1,989 1,699 1,687 1,696 2,286 1,692
Processing cost in U 967 899 827 822 789 955 729
Occupation cost in U 1,065 825 1,440 1,440 720 944 535
Amortization in U 0 0 0 0 0 0 125
Penalty 0 0 0 0 0 0 0
Prot per batch [D /batch] 68 99 108 108 254 157 459
Protability [D /h] 33 37 35 35 41 73 41
Selectivity of S [kmol S/kmol total] 0.448 0.507 0.593 0.597 0.595 0.441 0.596
Total energy consumption [kWh] 38,692 35,956 33,088 32,897 31,558 38,212 29,159

two objectives additionally include product revenue (RevenueS ),

processing cost (Costj,p ), occupational expenses (Costj,o ), and amor-
tization cost (Costj,a ) of unit j, and are dened by the minimization
problems:

minimize  = Profit,
udyn (t), ustat ,

uint , uBool
= RevenueS + Penalty + Cost A (37)

+ (Cost j,p + Cost j,o + Cost j,a ),
j {U1 ,U2 }

and

minimize  = Profitability = Profit/T total ,

udyn (t), ustat , (38)
uint , uBool

considering the economic data detailed in Table S2 (see Supplemen-

tary Material). Moreover, the prot maximization problem is solved
allowing the re-sizing of batch reactors to evaluate the additional
process improvement that can be obtained with plant modica-
tions (case iv). In this last case, equipment capacities larger than
the original 1 m3 volume become degrees of freedom and have
associated amortization cost.
The resulting KPIs in these cases are presented in Table 2.
The most relevant variations between optimal recipes iiv are
the selected operating modes, batch processing times, the num-
ber of lots, and temperature proles, the last ones illustrated in
Fig. 11. Additionally, the optimal recipe with equipment capacity
expansion involves the re-sizing of unit U1 with a new volume of
3.5 m3 . Overall, by including occupation costs into the objective
function, the series conguration is replaced by parallel one 
because the simultaneous operation of U1 and U2 in congura-
tion  involves a that fewer batches are needed e.g. 23 lots in
case ii instead of 47 in case i. This way, the sum of start-ups in
each unit decreases and the occupation costs, which have a term
that depends on the number of start-ups, are reduced signicantly
in optimal recipes ii and iii (see Table 2 and Eqs. (S2) and (S19)
in the Supplementary Material). As for the processing mode with
capacity expansion of U1 , the optimal conguration is with the
single operation of U1 , making the smaller unit U2 available to other
processes.
Fig. 11. Temperature proles in batch units in the Denbigh example for the xed
and optimal recipes in cases: (i) raw material cost minimization, (ii) prot maxi-
mization, (iii) protability maximization, and (iv) prot maximization with capacity
expansion.
Regarding the temperature proles, the prot maximization
(case i) leads to the maximum temperature of 80 C in unit U1 and
a compromise in U2 , operating at 94 C rather than at the upper
bound of 110 C. In contrast, protability maximization reaches the
upper bound in unit U2 , since the processing time reduction is con-
templated indirectly in the objective function (case iii); hence, the
optimal solution takes into account the compromise between the
economic gain and the reduction of the processing time. As a result,
the processing time in optimal recipe iii achieves a lower value
almost three times less compared to optimal recipe ii with prot
maximization alone. Finally, the higher processing capacity in opti-
mal recipe iv permits the functioning at the minimum temperature
in recipes with a much longer processing time and batch size (see
Table 2).
Essentially, an improvement of the 23% in the total prot is noted
in optimal recipe ii compared to the use of the xed recipe. This
304 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
Fig. 12. Process stages in acrylic ber production system: general processing scheme (solid lines) and potential processing alternatives (dashed lines).
increase is related to a reduction of raw material, occupation, and
processing costs, due to a higher S selectivity, lower energy con-
sumption, and reduction in the number of unit start-ups. Moreover,
optimal recipe iii involves a rise of the 121% in the protability
objective with respect to the xed recipe, even though this solu-
tion exhibits the highest total cost and the lowest selectivity of
all the optimal solutions. Besides, the consideration of expand-
ing the capacity of processing units in case iv hardly improves
the total prot compared to the optimal solution without equip-
ment modications. The reason is that the original two-reactor
plant was already equipped with the necessary capacity to ful-
ll the requested product demand, provided that the processing
units are properly arranged. Nevertheless, the plant modication
affects the expense distribution, dropping all the costs except for
the amortization.
7. Numerical example 2: acrylic ber production system
This case study addresses the development of an industrial-size
batch copolymerization process to produce an acrylic ber with
a specic composition and quality in a grassroots scenario. With
this purpose, the presented modeling strategy is applied for con-
structing the integrated optimization model that simultaneously
quanties the interaction between process synthesis and plant allo-
cation decisions like the selection of process stages, equipment
technology, chemicals, and solvent recovery and reuse, among oth-
ers.
The study of the integrated problem solution is motivated by the
complex reaction mechanisms and physicochemical phenomena
for example, the auto-acceleration and the Trommsdorf effect
which determine the compromise between productivity and
polymer quality. Then, the choice of processing conditions, such
as reactor temperature and monomer feed rate trajectories, and
processing times is crucial to determine the resulting polymer
properties and production yield (Giudici, 2000). Additionally,
there exist trade-offs between polymerization and downstream
tasks, which also contribute to the achievements of the over-
all economic and environmental production targets. Particularly,
polymer-solvent separation can be omitted when high conver-
sions are achieved (Bajaj et al., 1996), the magnitude of the
environmental impact of organic and aqueous solvents depends
on their molecular composition (Goldfein and Zyubin, 1990),
and cleaning technologies inuence the polymerization processes
(Capn-Garca et al., 2011).
The process for acrylic ber production comprises a primary
stage for bulk copolymer production and a secondary stage for
cool,k (t)) in the solu-
transforming the copolymer into spun format, following the gen-
eral process stages of the block diagram of Fig. 12 (see Section S2 of
the Supplementary Material for further details). Multiple operating
modes can be considered for each particular task, which determine
the required processing units and their synchronization. Moreover,
the following qualitative decisions can be made:
1. Solution or suspension copolymerization in the reaction (stage
1);
2. Organic or aqueous solvent in solution copolymerization (stage
1);
3. Selection of separation (stage 2);
4. Recirculation of the solvent recovered in the separation (stage 2)
to the copolymerization reaction (stage 1) in a subsequent batch;
5. Selection of washing and ltration (stage 3);
6. Recirculation of the washing water of the ltration (stage 3) to
the copolymerization reaction (stage 1);
7. Operating mode in the polymer wash and ltration (stage 7), i.e.
single unit F71 or series F71 followed by F72 ;
8. Operating mode in the solvent recovery (stage 8), i.e. single unit
C81 or series C81 followed by C82 ;
9. Recirculation of the solvent recovered in the separation (stage
8) to the copolymerization reaction (stage 1) or the repulping
(stage 4).
The resulting SEN superstructure representing all the possible
process stages and structural alternatives for acrylic ber produc-
tion is presented in Fig. 13.
7.1. Optimization problem
The goal is to produce 5 t of acrylic ber composed of 85% of
acrylonitrile (AN) and 15% of vinyl acetate (VA) in bulk format min-
imizing the total operational and capital cost. Also, a maximum
deviation of the 0.025 in the composition should be fullled to
guarantee polymer quality constraints. A single-product campaign
is assumed to produce the total amount of nal product in a time
horizon of 144 h. Particularly, the problem tackles the optimization
of the batch production process and the batch plant according to
the following objective function:
minimize  = Cost M1 + Cost M2 + Cost I + Cost S
udyn (t), ustat (39)
uint , uBool +Cost a + Cost p + Cost waste ,
where M1 refers to AN monomer, M2 to VA monomer, I to
initiator azobisisobutyronitrile (AIBN), and S to organic solvent
dimethylformamide (DMF), aqueous solvent sodium thiocyanate
(NaSCN(aq)), or suspension medium water. udyn (t) represents
j
the control proles of input and output ow rates (Fin1,k (t),
j j j
Fin2,k (t), Fout,k (t)) and cooling temperature (
tion and suspension copolymerization reactors j {R11 , R12 },
which follow three operations k {1, 2, 3} load, hold, and unload,
respectively. It also represents the control proles of the heat
j
(Qheat,k (t)) supplied to the evaporator j {E2 }, which comprises
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 305

Fig. 13. Superstructure of the acrylic ber example, composed of: solution and suspension polymerization reactors R11 and R12 , evaporator E2 , washing and ltering units F3 ,
F5 , F71 and F72 , repulping unit R4 , spinneret S6 , distillation columns C81 and C82 , and buffer tanks T2 , T3 , T4 , T81 and T84 . Equipment elements in black represent the addressed
primary polymerization process for bulk polymer production. Equipment elements in gray represent the secondary part of the process for spun production.

three operations k {1, 2, 3} load, unload of distillate, and unload
of both fractions, respectively. ustat denotes the composition of raw
j
materials in the reactors input stages (cc,in1,k , c {M1 , M2 , I}, k
{1, 2}) and the duration of batch operations (tl , l) and uint stands
for the capacity (Sizej ) of processing units j {R11 , R12 , E2 } and the
number of batches NB.
Regarding the logical decisions uBool , these include the selection
of separation stage (Zi , i {2}), solution or suspension technologies
j
in copolymerization reactors (V ,  {solu, susp}, j {R11 , R12 }),
organic or aqueous solvent in solution copolymerization reac-
j
tor (Sc , c {DMF, NaSCN(aq)}, j {R11 }), and solvent recovery and
reuse after the separation (R2,1 ). These decisions are related to
the selection of procesing units (Yj ), such as polymerization reac-
tors j {R11 , R12 }, separation unit j {E2 }, and recirculation buffer
tank j {T2 }. Finally, the multiple terms of the objective func-
tion  refer to: (i) the expenses of raw material, i.e. monomers
M1 and M2 , initiator I, and solvent or suspension medium S
(Cost M1 , Cost M2 , Cost I , Cost S ), (ii) equipment amortization (Costa ),
(iii) total processing cost (Costp ), and (iv) waste disposal costs
(Costwaste ). The denition of these terms and the corresponding
economic parameters can be found in Section S2 of the Supple-
mentary Material, (Eq. (S39) and Table S11).
The MLDO model additionally includes disjunctive equations
and logical propositions to relate the mentioned qualitative
alternatives, as well as multistage models representing potential
units. In particular, the process stages that have a critical impact
on the cost function are the copolymerization reaction deter-
mining the selectivity and production yield and the separation
stage governing the potential recovery of unreacted monomer,
solvent, and suspension medium. Their process performance is rep-
resented by dynamic models with discrete events and is controlled
by dynamic feed-forward control trajectories and duration of batch
operations. The model and process parameters are available in Sec-
tion S2 of the Supplementary Material.
7.2. Problem solution
The MLDO problem is solved through a direct-simultaneous
strategy. The dynamic model is fully discretized using eight nite
elements and three collocation points in normalized Legendre
roots. Piece-wise constant functions are used to dene the pro-
les of control variables. Moreover, given the low inuence of
dynamic proles in operations with material transference observed
in example 1, the dynamic models are approximated in load and
unload operations, namely stages k {1, 3} in reactors j {R11 ,
R12 } and evaporator j {E2 }. The resulting MINLP model is com-
putational intense for global optimizers due to the presence of
non-convex terms, even if it is moderate in size, as it can be
observed in Table 3. Therefore, the problem is solved with the OA
solver DICOPT in GAMS v.23.8.2 (Rosenthal, 2012), using CONOPT
3.15D and CPLEX 12.4 optimizers in the NLP and MILP sub-problems
respectively.
Several IFSs are used to initiate the search procedure from dif-
ferent initial points. Specically, the starting points correspond
to several subsystems of the integrated problem and are gener-
ated by xing the Boolean variables that dene specic structural
alternatives of the process. The subsystems are determined by the
following combinations of technologies and solvent:
306 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

Table 3
Features of the MINLP models implemented in GAMS in the acrylic ber example, obtained using eight nite elements and three collocation points in the discretization step:
IFS solutions with problem subsystems 1a, 1c, 2b, and 3b, and whole MINLP problem initialized with IFS 3.

No. equations No. continuous variables No. discrete variables Non-zero elements Non-linear terms Solution time (s)

IFS 1 (subsystem 1a)

IFS 3 (subsystem 1c)
16,860 8285 4 66,608 15,996 85
IFS 5 (subsystem 2b)
IFS 7 (subsystem 3b)
MINLP with IFS 3 16,860 8285 14 66,608 15,996 338

R11 , S R11 =
(1) Solution polymerization and organic solvent, i.e. Vsolu
DMF
true;
(2) Solution polymerization and aqueous solvent, i.e.
R11 , S R11
Vsolu = true;
NaSCN(aq)
R12 = true.
(3) Suspension polymerization, i.e. Vsusp
Additionally, these alternatives are evaluated without the
recirculation of distillate from the evaporator E2 (R2,1 = false in sub-
systems 1a, 2a, 3a) and with recirculation (R2,1 = true in 1b, 2b, 3b).
In the specic case of solution polymerization with organic sol-
vent, there exists the possibility of not having separation stage too
(Z2 = false in subsystem 3c). The remaining Booleans are computed
through the logical propositions of the problem. Overall, seven sets
of xed logical variables are used to generate IFSs. The computa-
tional performance to calculate some of these IFSs is summarized
in Table 3.
7.3. Results and discussion
7.3.1. Trade-off among process stages
The diagram of Fig. 14 presents the optimal objective function
of each subsystem in front of the optimal solution of the inte-
grated MINLP problem for comparison. The optimal solution among
the three polymerization and solvent alternatives 1, 2, or 3 in the
reaction stage depends on decisions associated with other process
stages like the recirculation choice. Suspension polymerization is
chosen when there is no recirculation of the distillate (subsystem
3a), whereas solution polymerization with an organic solvent is
the best alternative otherwise (subsystem 1b). At the same time,
the selection of separation stage depends on decisions made in
the reaction, like the polymerization temperature and monomer
dosage, which determine the achieved monomer conversion, and
the solvent selected. The only possibility for disregarding the sep-
aration is to utilize solution polymerization with DMF achieving
a minimum conversion (subsystem 1c). The proposed modeling
approach integrates all these degrees of freedom and permits the
simultaneous evaluation of the trade-offs between process stages
to obtain the optimal solution.
7.3.2. Cost balance
The KPIs and cost contributions of the subsystem solutions and
the integrated problem are summarized in Table 4. Essentially,
the most signicant weights are equipment amortization and raw
material expenses, which are considerably higher than the cooling
water costs in the reaction and the heating costs in the separation.
The combined expenditure of monomers M1 and M2 is the most
relevant one, followed by organic and aqueous solvent costs in R11 .
Therefore, the possibility of recirculating solvent in subsystems 1b
and 2b encompasses important raw material savings without loss
of the reaction effectiveness. Nevertheless, despite the raw material
savings obtained with recirculation, the optimal solution gives pri-
ority to the total cost minimization by eliminating the equipment
investment for the evaporator E2 and the recirculation storage T2 .
7.3.3. System adaptability
Comparing these results with preliminary calculations with
xed batch size, it was observed how the results presented here
have closer values of the objective function for the different sub-
systems, thanks to the consideration of the number of batches
as a degree of freedom. However, this similarity does not rep-
resent homogeneous optimal solutions but the adaptability of
the model to provide favorable solutions by evaluating the deci-
sions trade-offs. Notably, previous research on exible plant design
(Moreno-Benito et al., 2014) enhanced the batch plant exibility by
combining process synthesis decisions, like dynamic control pro-
les, with plant allocation decisions, like batch sizing.

7.3.4. System response

The optimal processing scheme corresponds to solution poly-
merization without separation, illustrated in Fig. 15, and is
characterized by the control and processing variable proles rep-
resented in Fig. 16. The optimal temperature trajectory in the
polymerization reactor follows a monotonically increasing func-
tion, whereas the M1 dosage features a local maximum in the
second half of the process that guarantees the desired polymer
composition of 85% of M1 . Thanks to the dosage prole and the
long task duration (i.e. 18.4 h) the process generates slight excess
of monomers and fullls the minimum monomer conversion of the
75% required for this processing structure.
This solution is compared to an equivalent optimization with
halved unitary amortization costs. In this case, the optimal solu-
tion is suspension polymerization with recirculation (i.e. VsuspR12 =

Fig. 14. Total cost for each subsystem and the integrated problem with and without
recirculation and avoiding separation stage (i.e. R2,1 = true, R2,1 = false, Z2 = false) in the
acrylic ber example. In black, integrated solution.
true, R2,1 = true), since the raw material savings are prioritized in
front of amortization costs, which are larger than in the original
economic scenario. The control and processing variable proles,
illustrated in Fig. 17, are characterized by a reduction of the
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 307

Table 4
KPIs and individual economic weights in the acrylic ber example considering: subsystem alternatives 1ac, 2ab, 3ab, and the whole MINLP problem. Items in bold
correspond to the total cost minimization objective.

Subsystems
Integrated
1a 1b 1c 2a 2b 3a 3b

Polymerization type Solution Solution Suspension Solution

Solvent type DMF NaSCN(aq) DMF
Process structure (Z2 ,R2,1 ) (1,0) (1,1) (0,0) (1,0) (1,1) (1,0) (1,1) (0,0)
Equipment size [m3 ] 1.5 1.5 2 2.5 2 1.5 1.5 2
No. batches 8 9 6 6 9 17 22 6
Batch size [kg/batch] 625 556 833 833 556 294 227 833
Total processing time [h] 119.9 134.9 89.9 89.9 86.2 144.0 144.0 89.9
Batch proc. time [h/batch] 18.0 18.2 18.3 19.9 11.9 11.6 8.4 18.4
Batch cycle time [h/batch] 16.7 16.9 18.3 18.5 10.6 10.4 7.2 18.4
Total cost [D ] 46,919 41,057 38,614 44,432 42,904 41,567 42,187 38,620
R11 or R12 Amortization 12,234 12,234 14,083 15,716 14,083 12,234 12,234 14,083
Water consumption cost 113 114 113 106 104 372 369 113
Monomer M1 cost 12,048 13,750 8,568 8,412 16,233 10,418 13,151 8,683
Monomer M2 cost 5,082 6,448 2,267 1,619 8,526 1,637 2,013 2,154
Initator I cost 3,355 2,876 7,449 2,258 1,114 1,653 2,145 7,296
Solvent/suspension S cost 1,192 692 2,824 732 236 99 129 2,981
E2 Amortization 8,368 8,368 10,750 9,633 8,368 8,368
Energy cost 202 257 268 452 655 977
Waste disposal cost 1,016 1,130 1,262 1,530 1,980 2,571
T2 Amortization 3,309 3,309 3,309
T4 Amortization 3,309 3,309 3,309 3,309 3,309 3,309 3,309 3,309
Cost per batch [D /batch] 5,865 4,562 6,436 7,405 4,767 2,445 1,918 6,437
Monomer conversion 0.44 0.37 0.75 0.83 0.29 0.71 0.56 0.75

Fig. 15. Optimal structure and logical variables in the acrylic ber example.

monomer dosage along the second stage of R12 , due to a higher
input of monomer in the rst load operation, coming from raw
material and recirculation ow rates. There is also a higher excess
of monomer M1 to guarantee the desired composition. This excess
is recovered in the distillate fraction in E2 . Even though the poly-
merization time is shorter in this case, with a cycle time of 9.1 h,
the solution exhibits a polymerization reaction signicantly longer
than the separation. Consequently, the consideration of more poly-
merization reactors with a parallel out-of-phase conguration
would be challenging to reduce the cycle time in case of desiring
larger production rates.
To conclude, the major strength of the proposed method-
ology is the holistic evaluation of the decision criteria. The
optimization model considers the trade-offs between the vari-
ous process stages within the complete process, as well as the
interactions between synthesis and allocation degrees of free-
dom. Additionally, it could expedite the analysis of processing
alternatives according to diverse processing, economic, or sustain-
able production drivers. The primary copolymerization stage for
acrylic ber production has been solved successfully, so it could be
challenging to consider the secondary polymerization process to
exploit the methodology. Additionally, further degrees of freedom
could be incorporated into the model, like the use of several reac-
tors and their arrangement in parallel out-of-phase conguration
to allow the reduction of the cycle time and increase the production
in a limited time horizon.
308 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
Fig. 16. Control and process variable proles of the polymerization reactor R11 for
the optimal solution in the acrylic ber example: (1) ow rates/dosage and temper-
ature and (2) molar compositions and reaction volume.
Fig. 17. Control and process variable proles of the polymerization reactor R12 and the separator E2 for the optimal solution with reduced amortization costs in the acrylic
ber example: (a1) ow rates/dosage and temperature, (b1) vapor ow, and (a2b2) molar compositions and volumes.
8. Conclusions and future directions
This work introduces a modeling strategy to tackle batch process
development and improvement at production scale, pursuing opti-
mal solutions that enhance process efciency and guarantee the
best utilization of available equipment and new investments. With
this purpose, multiple degrees of freedom associated with process
synthesis and plant allocation sub-problems are solved simulta-
neously, while taking into account the plant design and retrot.
The principal novelty of this strategy is the synchronization of
unit procedures with dynamic proles as a function of alternative
processing schemes in an integrated optimization problem. From a
modeling perspective, signicant progress has been made to coor-
dinate coexisting multistage and single-stage models through the
combination of hybrid discrete/continuous models, dynamic opti-
mization, and mixed-logic modeling, in a formulation that can be
handled by current optimization tools. Overall, the resulting MLDO
model manages the problem complexity using mixed-logic equa-
tions, while its reformulation into a MINLP problem enables its
optimization through well-established search algorithms.
The notable improvement of plant adaptability gained through
the presented approach justies and stimulates the research in this
direction. The complex mathematical implications and the risk of
obtaining mathematically intractable problems can be addressed
with appropriated solution strategies like the use of initial fea-
sible solutions and the denition of proper variable bounds. The
promising results of the numerical example show improvements
between the 23% and the 121% in the objective function compared
to the use of xed recipes in all the optimization scenarios iiv of
the Denbigh example. Also, the rapid evaluation of diverse objec-
tive functions proves the potential of the proposed strategy to adapt
master recipes according to in-time needs, as they could be changes
in economic scenarios or decision criteria, and to analyze the sys-
tem response in front of uncertain parameters.
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311 309

This strategy also enables the quantication of interactions Table 5

Basic logical operators expressed using logical and algebraic equations (Raman and
between synthesis and allocation sub-problems. A larger inu-
Grossmann, 1991), where Ad , d {1, . . ., N} are Boolean variables and ad , d {1, . . .,
ence on structural decisions with constant set-points optimization N} are the equivalent binaries.
is identied in the Denbigh example in comparison to the opti-
Logical operator Logical expression Algebraic equation
mization of dynamic proles with xed equipment conguration.
Both cases render improvements over the xed recipe of the 24% Conjunction and () A1 A2 AN a1 1, a2 1, . . ., aN 1
and the 12% respectively. In contrast, the consideration of dynamic Disjunction or () A1 A2 AN a1 + a2 + + aN 1
exclusive or () A1  A2   AN a1 + a2 + + aN = 1
ow rate proles in the operations with material transference has a
Implication () A1 A2 1 a1 + a2 1
minor impact, since most of the corresponding stages are character- or A1 A2 or a1 a2 0
ized by constant proles in the optimal solutions of the numerical Equivalence () (A1 A2 ) (A2 A1 ) a1 a2 0, a2 a1 0
cases tested. or(A1 A2 ) (A2 A1 ) or a1 = a2
Finally, the acrylic ber example shows the trade-offs among
decisions in neighboring process stages. Therein, the inuence of
the distillate recirculation on reaction decisions like the selection 3. Distributing the or operator () over the and operator ()
of the polymerization technology and solvent is illustrated. Like- recursively by using the following equivalence:
wise, reaction decisions like the temperature and dosage proles,
raw monomer concentration, and processing times determine the (A1 A2 ) A3 (A1 A3 ) (A2 A3 ). (A.4)
possibility of omitting separation stage if high monomer conver-
sions are achieved. The large advantage of the presented integrated Table 5 summarizes the relation between logical and algebraic
strategy is its potential to assess all processing alternatives simul- expressions, according to Raman and Grossmann (1991).
taneously and avoid methodologies based on the identication
of challenging solutions and their independent solution, with the
Appendix B. Application of orthogonal collocation on nite
probability of missing optimal solutions.
elements
Future work can be focused on the incorporation of additional
degrees of freedom in the problem, like out-of-phase parallel oper-
Orthogonal collocation on nite elements (Cuthrell and Biegler,
ation and multi-product/multi-purpose production campaigns. The
1989) is used to perform the full discretization of the time-
modular representation of unit procedures in the presented strat-
continuous model. The collocation points used in the numerical
egy should allow the direct extension of the formulation to achieve
examples are calculated with a shifted Legendre polynomial P() of
this purpose. Equally, the renement of solution strategies is highly
order M = 3, which is dened as:
relevant to propitiate the use of global optimization solvers and to
face problems with a larger number of process stages and prod- P() = 20  3 30  2 + 12  1. (B.1)
ucts. In this regard, the study of adaptive approaches that adjust the
number of nite elements in the discretization of the dynamic mod- The roots of the above polynomial provide the normalized loca-
els according to the selected batch times is suggested. The solution tions  m = {0.1127, 0.5000, 0.8873} of the collocation points m {1,
of this problem using hybrid approaches that combine determinis- . . ., M}, in addition to the boundary points located at  0 = 0 and
tic solvers with evolutionary algorithms is another promising area  M+1 = 1. Then, Lagrange polynomials are used for the approxima-
of research. tion of the state and control variables. In particular, the polynomials
of the differential z(t), algebraic y(t), and control udyn (t) variables,
and derivatives z(t) have the following form:
Acknowledgements

M 
M
The authors would like to thank the nancial support received z(t) = ze,m m (t),
z(t) = ze,m  m (t),
from the Spanish Ministry of Economy and Competitiveness and m=0 m=0
the European Regional Development Fund (research project SIG- (B.2)
M

M
dyn
ERA DPI2012-37154-C02-01 and FPU Research Fellowship). Marta y(t) = ye,m
m (t), udyn (t) = ue,m
m (t),
Moreno would also like to thank Prof. Luis Puigjaner for the fruitful m=1 m=1
discussions held during the development of this work.
dyn
where ze,m , ye,m , and ue,m represent the approximations of the dif-
ferential, algebraic, and control variables in nite elements e {1,
Appendix A. Reformulation of logical propositions into
. . ., Ne } and collocation or boundary points m {0, 1, . . ., M, M + 1},
linear constraints
M = 3 with normalized locations  m = {0, 0.1127, 0.5000, 0.8873, 1}.
m (t) and
m (t) represent the basis functions dened by:
The operations for formulating the conjunctive normal form
(CNF) of a logical proposition were formalized by Clocksin and 
M  M
Mellish (1981). Considering Boolean variables Ad , d {1, 2, 3}, these (t te,m ) (t te,m )
m (t)= ,
m (t)= . (B.3)
operations are: (te,m te,m ) (te,m te,m )
m = 0, m = 1,
m =
/ m m =
/ m
1. Replacing the implication by its equivalent disjunction:
The polynomials of differential variables z(t) are one order
A1 A2 A1 A2 ; (A.1) higher Mth than the polynomials of algebraic y(t) and control
variables udyn (t), with order (M 1)th. The purpose is to dene
2. Moving the negation inward by applying DeMorgans Theorem: boundary conditions of the approximation of differential variables
ze,0 in each nite element e, thanks to the symmetry properties
(A1 A2 ) A1 A2 , (A.2) of the method. This is necessary to enforce their continuity from
previous element e 1 in time te . Thus, accurate initial conditions
(A1 A2 ) A1 A2 , (A.3) ze,0 = z(te1 ) = ze1,M+1 can be calculated through the evaluation of
310 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311

the rst polynomial in Eq. (B.2) at the nal point  M+1 = 1 of each Chu, Y., You, F., 2014. Integrated scheduling and dynamic optimization of network
nite element: batch processes. In: American Control Conference (ACC), pp. 50245029,
http://dx.doi.org/10.1109/ACC.2014.6858593, ISSN 0743-1619.

M Clocksin, W.F., Mellish, C.S., 1981. Programming in Prolog. Springer-Verlag, New
ze,0 = ze1,m m (M+1 = 1), e {2, . . ., Ne }. (B.4)
m=0
Finally, control variables are approximated to a piecewise con-
dyn dyn
stant (PWC) function by establishing the relation ue,m = ue .
Appendix C. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.compchemeng.
2016.07.030.
References
Ahmad, B.S., 1997. Synthesis of Batch Processes with Integrated Solvent Recovery
(Ph.D. thesis). Massachusetts Institute of Technology.
Ali, S.A., 1999. Synthesis of Batch Processing Schemes for the Production of
Pharmaceuticals and Specialty Chemicals (Ph.D. thesis). Massachusetts
Institute of Technology.
Allgor, R.J., Barton, P.I., 1999. Mixed-integer dynamic optimization I: problem
formulation. Comput. Chem. Eng. 23 (45), 567584.
Allgor, R., Evans, L., Barton, P., 1999. Screening models for batch process
development: part I. Design targets for reaction/distillation networks. Chem.
Eng. Sci. 54 (19), 41454164.
Bajaj, P., Sen, K., Bahrami, S.H., 1996. Solution polymerization of acrylonitrile with
vinyl acids in dimethylformamide. J. Appl. Polym. Sci. 59 (10), 15391550.
Balas, E., 1985. Disjunctive programming and a hierarchy of relaxations for discrete
optimization problems. SIAM J. Algebraic Discrete Methods 6 (3), 466486.
Banimostafa, A., Papadokonstantakis, S., Hungerbhler, K., 2012. Retrot design of
a pharmaceutical batch process improving green process chemistry &
engineering principles. In: Karimi, I.A., Srinivasan, R. (Eds.), 11th International
Symposium on Process Systems Engineering. Computer Aided Chemical
Engineering, vol. 31. Elsevier, pp. 11201124.
Bansal, V., Perkins, J.D., Pistikopoulos, E.N., 2002. A case study in simultaneous
design and control using rigorous, mixed-integer dynamic optimization
models. Ind. Eng. Chem. Res. 41 (4), 760778.
Barbosa-Pvoa, A.P., 2007. A critical review on the design and retrot of batch
plants. Comput. Chem. Eng. 31 (7), 833855.
Barton, P.I., Pantelides, C.C., 1994. Modeling of combined discrete/continuous
processes. AIChE J. 40 (6), 966979.
Barton, P.I., Allgor, R.J., Feehery, W.F., Galan, S., 1998. Dynamic optimization in a
discontinuous world. Ind. Eng. Chem. Res. 37 (3), 966981.
Barton, P.I., Ahmad, B.S., Cheong, W., Tolsma, J., 1999. Synthesis of batch processes
with integrated solvent recovery. In: Sikdar, S.K., Diwekar, U. (Eds.), In: Tools
and Methods for Pollution Prevention, vol. 62. Kluwer Academic Publishers,
Dordrecht, pp. 205231.
Beaumont, N., 1990. An algorithm for disjunctive programs. Eur. J. Oper. Res. 48
(3), 362371.
Bemporad, A., Morari, M., 1999. Control of systems integrating logic, dynamics, and
constraints. Automatica 35 (3), 407427.
Bhatia, T., Biegler, L.T., 1996. Dynamic optimization in the design and scheduling of
multiproduct batch plants. Ind. Eng. Chem. Res. 35 (7), 22342246.
Binder, T., Blank, L., Bock, H.G., Bulirsch, R., Dahmen, W., Diehl, M., Kronseder, T.,
Marquardt, W., Schlder, J.P., von Stryk, O., 2001. Introduction to Model Based
Optimization of Chemical Processes on Moving Horizons. Springer-Verlag,
Berlin, pp. 295339.
Bryson, A.E., Ho, Y.-C., 1975. Applied Optimal Control: Optimization, Estimation
and Control. John Wiley & Sons, Inc.
Bumann, A.A., Papadokonstantakis, S., Fischer, U., Hungerbhler, K., 2011.
Investigating the use of path ow indicators as optimization drivers in batch
process retrotting. Comput. Chem. Eng. 35 (12), 27672785.
Capn-Garca, E., Bojarski, A.D., Espuna, A., Puigjaner, L., 2011. Multiobjective
optimization of multiproduct batch plants scheduling under environmental
and economic concerns. AIChE J. 57 (10), 27662782.
Capn-Garca, E., Guilln-Goslbez, G., Espuna, A., 2013. Integrating process Moreno-Benito, M., Espuna,
dynamics within batch process scheduling via mixed-integer dynamic
optimization. Chem. Eng. Sci. 102, 139150.
Carvalho, A., Matos, H.A., Gani, R., 2009. Design of batch operations: systematic
methodology for generation and analysis of sustainable alternatives. Comput.
Chem. Eng. 33 (12), 20752090.
Cavin, L., Fischer, U., Glover, F., Hungerbhler, K., 2004. Multi-objective process
design in multi-purpose batch plants using a Tabu Search optimization
algorithm. Comput. Chem. Eng. 28 (4), 459478.
Chakraborty, A., Linninger, A.A., 2002. Plant-wide waste management. 1. Synthesis
and multiobjective design. Ind. Eng. Chem. Res. 41 (18), 45914604.
Charalambides, M.S., Shah, N., Pantelides, C.C., 1995. Synthesis of batch
reaction/distillation processes using detailed dynamic models. Comput. Chem.
Eng. 19 (S1), 167174.
York.
Corsano, G., Aguirre, P.A., Iribarren, O.A., Montagna, J.M., 2004. Batch fermentation
networks model for optimal synthesis, design, and operation. Ind. Eng. Chem.
Res. 43 (15), 42114219.
Corsano, G., Montagna, J.M., Iribarren, O.A., Aguirre, P.A., 2007. Heuristic method
for the optimal synthesis and design of batch plants considering mixed
product campaigns. Ind. Eng. Chem. Res. 46 (9), 27692780.
Cuthrell, J., Biegler, L., 1989. Simultaneous optimization and solution methods for
batch reactor control proles. Comput. Chem. Eng. 13 (12), 4962.
Denbigh, K., 1958. Optimum temperature sequences in reactors. Chem. Eng. Sci. 8
(1-2), 125132.
Duran, M., Grossmann, I.E., 1986. An outer-approximation algorithm for a class of
mixed-integer nonlinear programs. Math. Program. 36, 307339.
Ferrer-Nadal, S., Capn-Garca, E., Mendez, C.A., Puigjaner, L., 2008. Material
transfer operations in batch scheduling. A critical modeling issue. Ind. Eng.
Chem. Res. 47 (20), 77217732.
Floudas, C.A., Grossmann, I.E., 1994. Algorithmic approaches to process synthesis:
logic and global optimization. In: Biegler, L.T., Doherty, M.F. (Eds.), Foundations
of Computer-Aided Process Design. , pp. 198221.
Frankl, K., Beenken, J., Marquardt, W., 2012. Integrated scheduling and control of
continuous-time blending processes. In: Karimi, I.A., Srinivasan, R. (Eds.), 11th
International Symposium on Process Systems Engineering. Computer Aided
Chemical Engineering, vol. 31. Elsevier, pp. 10901094.
Furman, K.C., Jia, Z., Ierapetritou, M.G., 2007. A robust event-based continuous time
formulation for tank transfer scheduling. Ind. Eng. Chem. Res. 46 (26),
91269136.
Gani, R., Papaeconomou, I., 2005. Conceptual, design and synthesis of batch
processes. In: Chemical Industries. CRC Press, pp. 4382.
Geoffrion, A.M., 1972. Generalized Benders decomposition. J. Optim. Theory Appl.
10, 237260.
Giudici, R., 2000. Polymerization reaction engineering: a personal overview of the
state of art. Latin Am. Appl. Res. 30, 351356.
Goldfein, M.D., Zyubin, B.A., 1990. Kinetics and mechanism of the processes of
preparing bre-forming polymers based on acrylonitrile review. Polym. Sci.
U.S.S.R. 32 (11), 21452166.
Grossmann, I.E., Guilln-Goslbez, G., 2010. Scope for the application of
mathematical programming techniques in the synthesis and planning of
sustainable processes. Comput. Chem. Eng. 34 (9), 13651376.
Grossmann, I., Westerberg, A., Biegler, L., 1987. Retrot design of processes. In:
Reklaitis, G., Spriggs, H. (Eds.), Proceedings of the First International
Conference on Foundations of Computer Aided Process Operations. , ISBN 0
444 98925 0, pp. 403442.
Grossmann, I.E., 1985. Mixed-integer programming approach for the synthesis of
integrated process owsheets. Comput. Chem. Eng. 9 (5), 463482.
Grossmann, I.E., 2002. Review of nonlinear mixed-integer and disjunctive
programming techniques. Optim. Eng. 3, 227252.
Halim, I., Carvalho, A., Srinivasan, R., Matos, H.A., Gani, R., 2011. A combined
heuristic and indicator-based methodology for design of sustainable chemical
process plants. Comput. Chem. Eng. 35 (8), 13431358.
Iribarren, O., Malone, M., Salomone, H., 1994. A heuristic approach for the design of
hybrid batch-continuous processes. Chem. Eng. Res. Des. 72 (A3), 295306.
Iribarren, O.A., Montagna, J.M., Vecchietti, A.R., Andrews, B., Asenjo, J.A., Pinto, J.M.,
2004. Optimal process synthesis for the production of multiple recombinant
proteins. Biotechnol. Prog. 20 (4), 10321043.
Iribarren, O.A., 1985. Batch Chemical Processes Design (Ph.D. thesis). University of
Massachusetts.
Jain, S., Kim, J.-K., Smith, R., 2013. Process synthesis of batch distillation systems.
Ind. Eng. Chem. Res. 52 (24), 82728288.
Kondili, E., Pantelides, C., Sargent, R., 1993. A general algorithm for short-term
scheduling of batch operations 1. MILP formulation. Comput. Chem. Eng. 17
(2), 211227, ISSN 0098-1354.
Linninger, A.A., Chakraborty, A., 1999. Synthesis and optimization of waste
treatment owsheets. Comput. Chem. Eng. 23 (10), 14151425.
Linninger, A.A., Stephanopoulos, E., Ali, S.A., Han, C., Stephanopoulos, G., 1995.
Generation and assessment of batch processes with ecological considerations.
Comput. Chem. Eng. 19 (Suppl. 1), 713.
Luus, R., 1990. Application of dynamic programming to high-dimensional
nonlinear optimal control problems. Int. J. Control 52 (1), 239250.
A., Puigjaner, L., 2014. Flexible batch process and plant
design using mixed-logic dynamic optimization: single-product plants. Ind.
Eng. Chem. Res.
Moreno-Benito, M., Dombayci, C., Espuna, A., Puigjaner, L., 2015. Integrated
process and plant design optimisation of industrial scale batch systems:
addressing the inherent dynamics through stochastic and hybrid approaches.
Chem. Eng. Trans. 45, 17891794.
Mosat, A., Cavin, L., Fischer, U., Hungerbhler, K., 2008. Multiobjective
optimization of multipurpose batch plants using superequipment class
concept. Comput. Chem. Eng. 32 (3), 512529.
Nemhauser, G.L., Wolsey, L.A., 1988. Integer and Combinatorial Optimization. John
Wiley & Sons, Inc.
Neuman, C., Sen, A., 1973. A suboptimal control algorithm for constrained
problems using cubic splines. Automatica 9 (5), 601613.
 M. Moreno-Benito et al. / Computers and Chemical Engineering 94 (2016) 287311
Nie, Y., Biegler, L.T., Wassick, J.M., 2012. Integrated scheduling and dynamic
optimization of batch processes using state equipment networks. AIChE J. 58
(11), 34163432.
Nie, Y., 2014. Integration of Scheduling and Dynamic Optimization: Computational
Strategies and Industrial Applications (Ph.D. thesis). Carnegie Mellon
University.
Oldenburg, J., Marquardt, W., 2008. Disjunctive modeling for optimal control of
hybrid systems. Comput. Chem. Eng. 32 (10), 23462364, ISSN 0098-1354.
Oldenburg, J., Marquardt, W., Heinz, D., Leineweber, D.B., 2003. Mixed-logic
dynamic optimization applied to batch distillation process design. AIChE J. 49
(11), 29002917.
Pantelides, C.C., 1994. Unied frameworks for the optimal process planning and
scheduling. In: Proceedings of the 2nd Conference on the Foundations of
Computer Aided Operations, p. 253.
Papaeconomou, I., 2005. Integration of Synthesis and Operational Design of Batch
Processes (Ph.D. thesis). Department of Chemical Engineering, Technical
University of Denmark.
Prata, A., Oldenburg, J., Kroll, A., Marquardt, W., 2008. Integrated scheduling and
dynamic optimization of grade transitions for a continuous polymerization
reactor. Comput. Chem. Eng. 32 (3), 463476.
Raman, R., Grossmann, I.E., 1991. Relation between MILP modelling and logical
inference for chemical process synthesis. Comput. Chem. Eng. 15 (2),
7384.
Reklaitis, G., 1989. Progress and issues in computer-aided batch process design. In:
3rd International Conference on Foundations of Computer-Aided Process
Design, Snowmass, Colorado, July 0914, pp. 241275.
Rippin, D.W.T., 1983. Simulation of single- and multiproduct batch chemical plants
for optimal design and operation. Comput. Chem. Eng. 7 (3), 137156.
Robinson, J.D., Loonkar, Y.R., 1972. Minimizing capital investment for
multi-product batch-plants. Process Technol. 17 (11), 861.
Rosenthal, R.E., 2012. GAMS: A Users Guide. Scientic Press.
Sahinidis, N.V., 1996. BARON: a general purpose global optimization software
package. J. Glob. Optim. 8 (2), 201205.
311
Sargent, R.W.H., Sullivan, G.R., 1978. The development of an efcient optimal
control package. In: Stoer, J. (Ed.), Optimization Techniques. Lecture Notes in
Control and Information Sciences, vol. 7. Springer-Verlag, Berlin Heidelberg,
pp. 158168.
Sharif, M., Shah, N., Pantelides, C.C., 1999. Design of integrated batch processes
with discrete and continuous equipment sizes. Comput. Chem. Eng. 23
(Suppl.), S117S120.
Sharif, M., Samsatli, N.J., Shah, N., 2000. Abstract design in the development of
pharmaceutical processes. In: Pierucci, S. (Ed.), European Symposium on
Computer Aided Process Engineering 10. Computer Aided Chemical
Engineering, vol. 8. Elsevier, pp. 685690.
Simon, L.L., Osterwalder, N., Fischer, U., Hungerbhler, K., 2008. Systematic retrot
method for chemical batch processes using indicators, heuristics, and process
models. Ind. Eng. Chem. Res. 47 (1), 6680.
Smith, E., Pantelides, C., 1995. Design of reaction/separation networks using
detailed models. Comput. Chem. Eng. 19 (Suppl. 1), 8388.
Srinivasan, B., Palanki, S., Bonvin, D., 2003. Dynamic optimization of batch
processes I. Characterization of the nominal solution. Comput. Chem. Eng. 27
(1), 126.
Stein, O., Oldenburg, J., Marquardt, W., 2004. Continuous reformulations of
discrete-continuous optimization problems. Comput. Chem. Eng. 28 (10),
19511966.
Stephanopoulos, G., Reklaitis, G.V., 2011. Process systems engineering: from Solvay
to modern bio- and nanotechnology. A history of development, successes and
prospects for the future. Chem. Eng. Sci. 66 (19), 42724306.
Trkay, M., Grossmann, I.E., 1998. Tight mixed-integer optimization models for the
solution of linear and nonlinear systems of disjunctive equations. Comput.
Chem. Eng. 22 (9), 12291239.
Tsang, T., Himmelblau, D., Edgar, T., 1975. Optimal control via collocation and
nonlinear-programming. Int. J. Control 21 (5), 763768.
Vassiliadis, V.S., Sargent, R.W.H., Pantelides, C.C., 1994. Solution of a class of
multistage dynamic optimization problems. 1. Problems without path
constraints. Ind. Eng. Chem. Res. 33 (9), 21112122.