INTERNATIONAL SYMPOSIUM ON PROCESS SYSTEMS ENGINEERING

&

EUROPEAN SYMPOSIUM ON COMPUTER AIDED PROCESS ENGINEERING

12TH

25TH

This page intentionally left blank

Edited by

Krist V. Gernaey, Jakob K. Huusom and Rafiqul Gani

Department of Chemical and Biochemical Engineering Technical University of Denmark DK-2800 Lyngby, Denmark

PART A

COMPUTER-AIDED CHEMICAL ENGINEERING, 37

INTERNATIONAL SYMPOSIUM ON PROCESS SYSTEMS ENGINEERING AND

EUROPEAN SYMPOSIUM ON COMPUTER AIDED PROCESS ENGINEERING

25TH

12TH

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1B, UK

Copyright © 2015 Elsevier B.V. All rights reserved

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier's Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress

ISBN (Part A): 978-0-444-63578-5 ISBN (Set): 978-0-444-63429-0 ISSN: 1570-7946

For information on all Elsevier publicationsvisit our web site at store.elsevier.com

Printed and bound in Great Britain

14 15 16 17 10 9 8 7 6 5 4 3 2 1

Contents

Preface xixCommittees xxiLocal Organising Committee xxvSponsors xxvii

Plenary PapersRecent advances in mathematical programming techniques for the optimization of process systems under uncertaintyIgnacio E. Grossmann, Robert M. Apap, Bruno A. Calfa, PabloGarcia-Herreros, Qi Zhang 1A multidisciplinary hierarchical framework for the design of consumer centered chemical productsKa M. Ng 15Multi-Level design of process systems for efficient chemicals production and energy conversionKai Sundmacher 25Keynote PapersPSE tools for process intensificationPhilip Lutze 35Towards the integration of process design, control and scheduling: Are we getting closer?Efstratios N. Pistikopoulos, Nikolaos A. Diangelakis, Amit M. Manthanwar 41Industrially applied PSE for problem solving excellenceAntoon J. B. ten Kate 49Sustainable production of liquid fuelsJonathan P. Raftery, M. N. Karim 55Industrial perspectives on the deployment of scheduling solutionsIiro Harjunkoski 63Overview of smart factory studies in petrochemical industryDefang Li, Baihua Jiang, Hansheng Suo, Ya Guo 71A PSE approach to patient-individualized physiologically-basedpharmacokinetic modelingRoberto Andrea. Abbiati, Gaetano Lamberti, Anna Angela. Barba,Mario Grassi, Davide Manca 77Modeling and optimization of continuous pharmaceutical manufacturing processesAmanda Rogers, Marianthi Ierapetritou 85

vi Contents

Process technology licensing: An interface of engineering and business Andreas Bode, Jose Castro-Arce, Bernd Heida, Carsten Henschel, Achim Wechsung, Justyna Wojcicka 93 Simple rules for economic plantwide control Vladimiros Minasidis, Sigurd Skogestad, Nitin Kaistha 101 Mixed-Integer fractional programming: Models, algorithms, and applications in process operations, energy systems, and sustainability Fengqi You 109 Advances and challenges in modelling of processing of lipids Bent Sarup 117 A perspective on PSE in fermentation process development and operation Krist V. Gernaey 123 Sustainable production and consumption: A decision-support framework integrating environmental, economic and social sustainability Adisa Azapagic 131 Control of reaction systems via rate estimation and feedback linearization Diogo Rodrigues, Julien Billeter, Dominique Bonvin 137 Modeling the fixed-bed Fischer-Tropsch reactor in different reaction media Rehan Hussain, Jan H. Blank, Nimir O. Elbashir 143 Contributed Papers T-0: PSE-CAPE and Education Process simulators: What students forget when using them, their limitations, and when not to use them Joseph A. Shaeiwitz, Richard Turton 149 Learning to solve mass balance problems through a web-based simulation environment Alexandros Koulouris, Dimitrios Vardalis 155 Model predictive control of post-combustion CO2 capture process integrated with a power plant Evgenia D. Mehleri, Niall Mac Dowell, Nina F. Thornhill 161 A framework to structure operational documents for chemical processes Hiroshi Osaka, Yuji Naka, Tetsuo Fuchino 167 Teaching sustainable process design using 12 systematic computer-aided tasks Deenesh K. Babi 173

Contents vii

Contributed Papers T-1: Modelling, Numerical Analysis and Simulation Optimization of chemical processes using surrogate models based on a kriging interpolation Natalia Quirante, Juan Javaloyes, Rubén Ruiz-Femenia, José A. Caballero 179 Global sensitivity analysis for a model of B-Cell chronic lymphocytic leukemia disease trajectories Symeon Savvopoulos, Ruth Misener, Nicki Panoskaltsis, Efstratios N. Pistikopoulos, Athanasios Mantalaris 185 Modelling and simulation of complex nonlinear dynamic processes using data based models: application to photo-fenton process Ahmed Shokry, Francesca Audino, Patricia Vicente, Gerard Escudero, Montserrat Perez Moya, Moises Graells, Antonio Espuña 191 A meshfree maximum entropy method for the solution of the population balance equation Menwer Attarakih, Abdelmalek Hasseine, Hans-Jörg Bart 197 Modelling the hydrodynamics of bubble columns using coupled OPOSPM-maximum entropy method Menwer Attarakih, Ferdaous Al-Slaihat, Mark W. Hlawitschkac, Hans-Jörg Bart 203 Process simulation of a 420MW gas-fired power plant using Aspen Plus Bao-Hong Li, Nan Zhang, Robin Smith 209 Optimal blending study for the commercial gasoline Cristian Patrascioiu, Bogdan Doicin, Grigore Stamatescu 215 Population balance model for enzymatic depolymerization of branched starch Christoph Kirse, Heiko Briesen 221 Analysis of the transfer of radical co-polymerization systems from semi-batch to continuous plants Thilo Goerke, Sebastian Engell 227 Thermodynamic calculations for systems biocatalysis Rohana Abu, Maria T. Gundersen, John M. Woodley 233 Parallel computation method for solving large scale equation-oriented models Yannan Ma, Jinzu Weng, Zhijiang Shao, Xi Chen, Lingyu Zhu, Yuhong Zhao 239 Dynamic investment appraisal: Economic analysis of mobile production concepts in the process industry G. Bas, T. E. Van der Lei 245

viii Contents

Prediction of heat capacity of ionic liquids based on COSMO-RS S -profileYongsheng Zhao, Ying Huang, Xiangping Zhang, Suojiang Zhang 251OPOSSIM: A population balance-SIMULINK module for modelling coupled hydrodynamics and mass transfer in liquid extraction equipmentMenwer Attarakih, Samer Al-Zyod, Mark Hlawitschke, Hans-Jörg Bart 257CFD-DEM simulation of a fluidized bed crystallization reactorKristin Kerst, Luis Medeiros de Souza, Antje Bartz, Andreas Seidel-Morgenstern, Gàbor Janiga 263A modelling, simulation, and validation framework for the distributed management of large-scale processing systemsShaghayegh Nazari, Christian Sonntag, Goran Stojanovski, Sebastian Engell 269Performance analysis and optimization of the biomass gasification and Fischer-Tropsch integrated process for green fuel ProductionsKarittha Im-orb, Lida Simasatitkul, Amornchai Arpornwichanop 275Dynamic behavior adjustment of 1, 3-propanediol fermentation processHao Jiang, Nan Zhang, Jinsong Zhao, Tong Qiu, Bingzhen Chen 281Modelling and simulation of pressure swing adsorption (PSA) processes for post-combustion carbon dioxide (CO2) capture from flue gasGeorge N. Nikolaidis, Eustathios S. Kikkinides, Michael C. Georgiadis 287Validation of a functional model for integration of safety into process system designWu, J., Lind, M., Zhang, X., Jørgensen, S. B., Sin, G 293A dynamic method for computing thermodynamic equilibria in process simulationAlexander Zinser, Kongmeng Ye, Liisa Rihko-Struckmann, Kai Sundmacher 299Dynamics and operation analysis of the PHB (polyhydroxybutyrate) fermentation Moises González-Contreras, Omar Anaya-Reza, Mauricio Sales-Cruz,Teresa Lopez-Arenas 305Retrofitting of concentration plants using global sensitivity analysisFreddy Lucay, Luis A. Cisternas, Edelmira D. Gálvez 311Differential-algebraic approach to solve steady-state two-phase flow drift-flux model with phase changeRodrigo G. D. Teixeira, Argimiro R. Secchi, Evaristo C. Biscaia Jr. 317

Contents ix

Model-based design of experiments for the identification of kinetic models in microreactor platformsFederico Galvanin, Enhong Cao, Noor Al-Rifai, Asterios Gavriilidis,Vivek Dua 323Application of derivative - free estimator for semi batch autocatalytic esterification reactor: Comparison study of unscented Kalman filter, divided difference Kalman filter and cubature Kalman filter F. S. Rohman, S. Abdul Sata, N. Aziz 329Modeling and parameter estimation of coke combustion kinetics in a glycerol catalytic conversion reactorMinghai Lei, François Lesage, M. Abderrazak Latifi, Serge Tretjak 335Behavior of heavy metals during gasification of phytoextraction plants: thermochemical modellingMarwa SAID, Laurent CASSAYRE, Jean-Louis DIRION, AngeNZIHOU, Xavier JOULIA 341Multi-objective optimization for the production of fructose in a Simulated Moving Bed ReactorEdwin Zondervan, Bram van Duin, Nikola Nikacevic, Jan Meuldijk 347A modeling framework for optimal design of renewable energy processes under market uncertaintyAryan Geraili, Jose A. Romagnoli 353A molecular reconstruction feed characterization and CAPE OPEN implementation strategy to develop a tool for modeling HDT reactors for light petroleum cutsCésar G. Pernalete, Jasper van Baten, Juan C. Urbina, José F. Arévalo 359Techno-economic analysis of ethanol-selective membranes for corn ethanol-water separationAdam Kelloway, Michael Tsapatsis, Prodromos Daoutidis 365Equation-oriented modeling of multi-stream heat exchanger in air separation unitsLiuzhen Jiang, Kai Zhou, Lingyu Zhu 371Proposal of a new pathway for microalgal oil production and its comparison with conventional methodSofia Chaudry, Parisa A. Bahri, Navid R. Moheimani 377Superstructure development, simulation and optimization of desalination systems using Aspen Custom modelerSidra N. Malik, Parisa A. Bahri, Linh T. T. Vu 383Microalgae growth determination using modified breakage equation modelErgys Pahija, Yu Zhang, Maojian Wang, Yi Zhu, Chi Wai Hui 389

x Contents

A new strategy for the simulation of gas pipeline network based on system topology identificationZengzhi Du, Chunxi Li, Wei Sun, Jianhong Wang 395Modelling and optimization of a heat integrated gasification processYi Zhu, Adetoyese Olajire Oyedun, Maojian Wang, Ergys Pahija, ChiWai Hui 401Analyzing and modeling ethylene cracking process with complex networks approachZhou Fang, Tong Qiu, Bingzhen Chen 407Optimization and economic evaluation of bioethanol recovery and purification processes involving extractive distillation and pressure swing adsorptionY. Y. Loy, X. L. Lee, G. P. Rangaiah 413Data reconciliation in reaction systems using the concept of extentsSriniketh Srinivasan, Julien Billeter, Shankar Narasimhan, Dominique Bonvin 419Development of a generic model of a Ruthenium reactorNorbertin Nkoghe Eyeghe, Carl Sandrock, Carel Van Dam 425Dynamic modelling and experimental validation of a pilot-scaletubular continuous reactor for the autohydrolysis of lignocellulosic materialsC. González-Figueredo, A. Sánchez, G. Díaz, F. Rodríguez, R. Flores, M. A. Ceballos, R. Puente, H. A. Ruiz 431First-principles model diagnosis in batch systems by multivariate statistical modelingNatascia Meneghetti, Pierantonio Facco, Sean Bermingham, David Slade, Fabrizio Bezzo, Massimiliano Barolo 437Integrated analysis of an evaporation and distillation bioethanol industrial system using direct and indirect heatingRodrigo O. Silva, Vandré C. Tiski, Rafael O. Defendi, Lucas B. Rocha,Oswaldo C. M. Lima, Laureano Jiménez, Luiz Mario M. Jorge 443Reformulating the minimum eigenvalue maximization in optimal experiment design of nonlinear dynamic biosystemsDries Telen, Nick Van Riet, Filip Logist, Jan Van Impe 449A framework for modular modeling of the diesel engine exhaust gas cleaning systemAndreas Åberg, Thomas K. Hansen, Kasper Linde, Anders K. Nielsen, Rune Damborg, Anders Widd, Jens Abildskov, Anker D. Jensen, Jakob K. Huusom 455

Contents xi

An approximate modelling method for industrial L-lysine fermentation processHangzhou Wang, Faisal Khan, Bo Chen, Zongmei Lu 461Model-based prediction and experimental validation of viscosities of soap emulsionsDaniel M. Macías-Pelayo, Pedro A. Alonso-Dávila, Alfonso Martínez-Villalobos, Alicia Román-Martínez 467A novel quantisation-based integration method for ODEsVassilios S. Vassiliadis, Fabio Fiorelli, Harvey Arellano-Garcia 473Modelling and parameter estimation of enzymatic biodiesel synthesisPriscila S. Sabaini, Thais Fabiana C. Salum, Rossano Gambetta,Fabricio Machado 479Analysis of two alternatives to produce ethylene from shale gasAndrea P. Ortiz-Espinoza, Mahmoud M. El-Halwagi, Arturo Jiménez-Gutiérrez 485A crude oil econometric model for PSE applicationsDavide Manca, Valentina Depetri, Clément Boisard 491Simulation study of temperature distribution in the thermal drying oven for a lacquer coating processPaisan Kittisupakorn, Patsarawan Lipikanjanakul 497Outlier treatment for improving parameter estimation of group contribution based models for upper flammability limitJérôme Frutiger, Jens Abildskov, Gürkan Sin 503Integration and optimization of an air separation unit (ASU) in an IGCC plantMaojian Wang, Adetoyese Olajire Oyedun, Ergys Pahija, Yi Zhu,Guilian Liu, Chi Wai Hui 509Mathematical modeling of an industrial delayed coking unitClaudio N. Borges, Maria A. Mendes, Rita M. B. Alves 515Post-combustion CO2 capture with sulfolane based activated alkanolamine solventSukanta K. Dash, Bikash K. Mondal, Amar N. Samanta, Syamalendu S. Bandyopadhyay 521Modeling and sensitivity analysis of a medium-temperature gas cleaning process of biogenous synthesis gasMichaela Fraubaum, Heimo Walter 527Exergy analysis of monoethylene glycol (MEG) recovery systemsAlexandre M. Teixeira, José Luiz de Medeiros, Ofélia Q. F. Araújo 533Production of biodiesel via enzymatic palm oil ethanolysis: Kinetic studyShayane P. Magalhães, Fernando L. P. Pessoa, Tito L. M. Alves 539

xii Contents

Integration of retrofitted coal-fired power plant with CCS: Power de-rate minimizationJinjoo An, Ung Lee, Jaeheum Jung, Chonghun Han 545CUDA-optimized cellular automata for diffusion limited processesAndrey Kolnoochenko, Natalia Menshutina 551Effect of ship motion on amine absorber with structured-packing for CO2 removal from natural gas: An approach based on porous medium CFD modelDung A. Pham, Young-Il Lim, Hyunwoo Jee, Euisub Ahn, Youngwon Jung 557Model-based analysis and efficient operation of a glucose isomerization reactor plantEmmanouil Papadakis, Ulrich Madsen, Sven Pedersen, Krist V. Gernaey, John M. Woodley, Rafiqul Gani 563pyIDEAS: An open-source python package for model analysisTimothy Van Daele, Stijn Van Hoey, Ingmar Nopens 569A numerical procedure for model identifiability analysis applied to enzyme kineticsTimothy Van Daele, Stijn Van Hoey, Krist V. Gernaey, Ulrich Krühne,Ingmar Nopens 575Integrated simulation platform of chemical processes based on virtual reality and dynamic modelNa Luo, Xiaoqiang Wang, Feng Van, Zhen-Cheng Ye, Feng Qian 581OsmoseLua - An integrated approach to energy systems integration with LCIA and GISMin-Jung Yoo, Lindsay Lessard, Maziar Kermani, François Maréchal 587Incremental kinetic identification based on experimental data from steady-state plug flow reactorsNirav Bhatt, Srividhya Visvanathan 593Nonlinear fuzzy identification of batch polymerization processesNádson N. M. Lima, Lamia Zuniga Linan, Delba N. C. Melo, Flavio Manenti, Rubens Maciel Filho, Marcelo Embiruçu, Maria R. Wolf Maciel 599Modeling dissolution of solids based on cellular automata with changing sizes of cellsSviatoslav I. Ivanov, Irina A. Tiptsova, Natalia V. Menshutina 605Data analysis and modelling of a fluid catalytic cracking unit (FCCU) for an implementation of real time optimizationJuan D. Reyes, Adriana L. Rodriguez, Carlos A. M. Riascos 611

Contents xiii

A hybrid discrete/continuous dynamic model of trayed tower hydraulics David Pinilla-García, Santos Galán 617 Application of the lagrangian cfd approach to modelling of crystallization in stirred batch reactors using the smoothed particle hydrodynamics method Dragan D. Nikolic, Brian P. de Souza, Patrick J. Frawley 623 Model reduction in visual modelling Heinz A. Preisig 629 Automatic reconstruction and generation of structured hexahedral mesh for non-planar bifurcations in vascular network Mahsa Ghaffari, Chih-Yang Hsu, Andreas A. Linninger 635 Developing Surrogate Models via Computer Based Experiments Mandar N. Thombre, Heinz A. Preisig and Misganaw B. Addis 641 Systematic development of kinetic models for systems described by linear reaction schemes Carolina S. Vertis, Nuno M. C. Oliveira, Fernando P. M. Bernardo 647 Rigorous modeling, simulation and optimization of a dividing wall batch reactive distillation column: A comparative study Edna Soraya Lopez-Saucedo, Juan Gabriel Segovia-Hernandez, Ignacio E. Grossmann and Salvador Hernandez-Castro 653 Theoretical modeling of (non) reactive residue curve maps for TAME synthesis system using MATLAB – SIMULIS thermodynamics communication facilitie’sM. M. Ceau escu, Jordi Bonet-Ruiz, V. Ple u P. Iancu, A. E. Bonet-Ruiz 659 Alternative prediction models for data scarce environment Ali Al-Shanini, Arshad Ahmad, Faisal Khan, Olagoke Oladokum, Shadia Husna Mohdd Nor 665 Multi-objective optimisation of atmospheric crude distillation system operations based on bootstrap aggregated neural network models Funmilayo N. Osuolale, Jie Zhang 671 Optimization studies through simulation of a methanol/water/glycerol distillation column José Palmeira, João M. Silva, Henrique A. Matos 677 Simulation of a 3D bioprinted human vascular segment Nogueira JA., Lara VF., Marques TS., Oliveira DS., Mironov V., da Silva JV., Rezende RA. 683

xiv Contents

Modeling fixed-bed multicomponent adsorption as a step to achieve ultra-low sulfur dieselTristán Esparza-Isunza, Felipe López-Isunza 689Experimental and CFD simulation studies of circulating fluidized bed riser in the fast fluidization regimeMukesh Upadhyay, Myung Won Seo, Nam Sun Nho, Jong-Ho Park 695Application of new electrolyte model to phase transfer catalyst (PTC) systemsSun Hyung Kim, Amata Anantpinijwatna, Jeong-Won Kang,Mauricio Sales-Cruz, Rafiqul Gani 701A novel rigorous mathematical programming approach to construct phenomenological modelsVassilios S. Vassiliadis, Yian Wang, Harvey Arellano-Garcia, Ye Yuan 707Dynamic simulation of a batch aqueous two-phase extraction process for -amylaseNehal Patel, Daniel Bracewell, Eva Sorensen 713Contributed Papers T-2: Mathematical Programming (Optimization)A framework for hybrid multi-parametric model-predictive control with application to intravenous anaesthesiaIoana Nascu, Richard Oberdieck, Efstratios N. Pistikopoulos 719Dynamic chance-constrained optimization under uncertainty on reducedparameter setsDavid Müller, Erik Esche, Sebastian Werk, Günter Wozny 725Optimal design of thermal membrane distillation networksRamon González-Bravo, Fabricio Nápoles-Rivera, José María Ponce-Ortega, Medardo Serna-Gonzalez, Mahmoud M. El-Halwagi 731Multicolumn-multicut cross decomposition for stochastic mixed-integer linear programmingEmmanuel Ogbe, Xiang Li 737Efficient ant colony optimization (EACO) for solvent selection using computer aided molecular designBerhane H. Gebreslassie, Urmila M. Diwekar 743Optimisation of process parameters with simultaneous consideration of energy efficiency measuresTimo Bohnenstaedt, Kristina Zimmermann, Georg Fieg 749Optimization of split fractions and cleaning schedule management in heat exchanger networksJian Du, Jie Fan, Linlin Liu, Jilong Li, Yu Zhuang, Quinwei Meng 755

Contents xv

A cost targeting method for studying investment on heat exchanger networks for collection of industrial excess heatMatteo Morandin, Lina Eriksson 761Ellipsoidal arithmetic for multivariate systemsM. E. Villanueva, J. Rajyaguru, B. Houska, B. Chachuat 767Reduced model trust region methods for embedding complex simulations in optimization problemsJohn P. Eason, Lorenz T. Biegler 773Optimization of LNG supply chainAlice Bittante, Raine Jokinen, Frank Pettersson, Henrik Saxén 779Metaheuristic techniques for the optimal design of NGL pipeliningPaola P. Oteiza, Martín C. De Meio Reggiani, Diego A. Rodriguez,Valentina Viego, Nélida B. Brignole 785Deterministic global dynamic optimisation using interval analysisCarlos Perez-Galvan, I. David L. Bogle 791Separation process optimization under uncertainty by chance constraint programming with recourseLi Sun, Huajie Zhang 797Optimal operating policies for synthesizing tailor made gradient copolymersCecilia Fortunatti, Bruno Mato, Adriana Brandolin, Claudia Sarmoria,Mariano Asteasuain 803Degeneracy hunter: An algorithm for determining irreducible sets of degenerate constraints in mathematical programsAlexander W. Dowling, Lorenz T. Biegler 809Dynamic multi-objective optimization of batch chromatographic separation processesA. Holmqvist, F. Magnusson, B. Nilsson 815An adaptive multi-objective differential evolution algorithm for solving chemical dynamic optimization problemsXu Chen, Wenli Du, Feng Qian 821Optimal operation of a pyrolysis reactorAysar T. Jarullah, Shemaa A. Hameed, Zina A. Hameed, I.M. Mujtaba 827Representation of the convex envelope of bilinear terms in a reformulation framework for global optimisationAndreas Lundell, Tapio Westerlund 833Interactive multi-objective decision-support for the optimization of nonlinear dynamic (bio)chemical processes with uncertaintyMattia Vallerio, Jan Hufkens, Jan Van Impe, Filip Logist 839

xvi Contents

Superstructure optimisation of a water minimization network with an embedded multi-contaminant electrodialysis modelChiedza D. Nezungai, Thokozani Majozi 845Deterministic global optimization and transition statesDimitrios Nerantzis, Claire S. Adjiman 851A metaheuristic for solving large-scale two-stage stochastic mixed 0-1programs with a time stochastic dominance risk averse strategySusana Baptista, Ana P. Barbosa-Póvoa, Laureano Escudero, Maria I. Gomes, Celeste Pizarro 857Optimized production of multilayered monodisperse polymer nanoparticlesBrahim Benyahia, M. A. Latifi, C. Fonteix, F. Pla 863Systematic design of chemical reactors with multiple stages via multi-objective optimization approachMohd Nazri Mohd Fuad, Mohd Azlan Hussain 869Synthesis and design of integrated process and water networksZainatul B. Handani, Alberto Quaglia, Rafiqul Gani 875Optimization of high-density polyethylene process based on molecular weight distribution and chemical composition distribution under uncertaintyJiayuan Kang, Xi Chen, Zhijiang Shao 881A systematic approach for targeting zero liquid discharge in industrial parksZakarya A. Othman, Patrick Linke, Mahmoud El-Halwagi 887Decomposition techniques for the real-time optimization of a propylene production unitA. M. Acevedo P., J.E.A. Graciano, Fabio D.S. Liporace, A.S. Vianna Jr., Galo A.C. Le Roux 893An approach to deal with non-convex models in real-time optimization with modifier adaptationMaximiliano Garcia, Juan Pablo Ruiz, Marta Basualdo 899A robust minimax Semidefinite Programming formulation for optimal design of experiments for model parameterisationBelmiro P.M. Duarte, Guillaume Sagnol, Nuno M.C. Oliveira 905Design of a multi-contaminant water allocation network using multi-objective optimizationSofía De-León Almaraz, Marianne Boix, Catherine Azzaro-Pantel,Ludovic Montastruc, Serge Domenech 911

Contents xvii

Simulation and optimization of the ethane cracking process to produce ethyleneDaison Y. Caballero, Lorenz T. Biegler, Reginaldo Guirardello 917Study of performance of a novel stochastic algorithm based on Boltzmann distribution (BUMDA) coupled with self-adaptive handling constraints technique to optimize chemical engineering processR. Murrieta-Dueñas, J. Cortez-González, A. Hernández-Aguirre, R.Gutiérrez-Guerra, S. Hernandez, J. G. Segovia-Hernández 923Dynamic modelling and optimal design of the solid-phase reactive chromatographic separation system for biomass saccharification via acid hydrolysisPakkapol Kanchanalai, Matthew J Realff, Yoshiaki Kawajiri 929

Krist V. Gernaey, Jakob K. Huusom and Rafiqul Gani (Eds.), 12th International Symposium on Process

Systems Engineering and 25th European Symposium on Computer Aided Process Engineering.

31 May - 4 June 2015, Copenhagen, Denmark. c© 2015 Elsevier B.V. All rights reserved.

Rigorous Modeling, Simulation and Optimization ofa Dividing Wall Batch Reactive Distillation Column:a comparative studyEdna Soraya Lopez-Saucedoa, Juan Gabriel Segovia-Hernandeza, Ignacio E. Grossmannb

and Salvador Hernandez-Castroa

aDepartment of Chemical Engineering; Universidad de Guanajuato; Guanajuato, MexicobDepartment of Chemical Engineering; CMU, Pittsburgh, USAelopez@andrew.cmu.edu

AbstractA model and solution strategies are investigated for the optimization of a Dividing Wall Batch

Reactive Distillation Column (DWBRC). In order to accomplish this objective, we describe a

dynamic model that involves tray-by-tray calculations for the time varying column profiles. In

order to compare the simultaneous solution of the system of differential and algebraic equations,

two different approaches are used: equation oriented based on orthogonal collocation implemented

in GAMS, and control vector parameterization (CVP) as implemented in gPROMS.

Keywords: Reactive batch distillation, dynamic optimization, Dividing Wall Columns

1. Introduction

Reactive Batch Distillation Columns (RBC) have been studied as a promising technology due to

its dual functionality: separation and reaction. Modeling, simulation, and optimization of batch

distillation processes rely on dynamic models. A number of different solution approaches for this

kind of systems, described by a set of differential and algebraic equations (DAEs) have been pro-

posed in the literature. One of these approaches has been developed by Biegler (1984), in which

the dynamic optimal control problem is approximated by a finite dimensional nonlinear program

(NLP) through the discretization of all variables using finite elements with orthogonal colloca-

tion points. This equation oriented approach can then be solved with GAMS (General Algebraic

Modeling System 24.2.2) as an NLP problem to simultaneously perform the optimization while

converging the DAEs. The other solution method is the Control Vector Parametization (CVP) pro-

posed by Vassiliadis et al. (1994) which relies on the iterative solution of DAEs in the space of the

control variables in order to perform the optimization with a Successive Quadratic Programming

(SQP) method.

In this study we propose the optimal design and operation of a Dividing Wall Batch Reactive

Distillation Column (DWBRC). We propose an esterification reaction for the production of ethyl

acetate (as the distillate product). This study investigates with the two solution approaches how

the parameters such as vapor flow rate (V , kmol/hr) and optimal control variable reflux ratio(RR = R/V ) are to be adjusted to maximize the productivity in the column for a given productspecification. First, we provide the problem statement followed by the proposed mathematical

model in section 3. The model was solved using two different approaches presented in section

4 for the production of ethyl acetate. The column configuration and operational conditions are

Edna Soraya Lopez-Saucedo et al.

presented in section 5. The results are shown in section 6 followed by the conclusions in section

7.

2. Problem statement

R1

V1

V2

Reflux Drum

Liquid distributor

Vapordistributor

D

Dividing Wall

R2

x2j+1,i

y2j-1,i

RV

R2

V2

R1

V1 V

2R2

RV

Internal Trays j

Reboilerstage 1

Condenserstage NT

j

Figure 1: Batch Reactive Distillation

Column (BRC)

In a general form the problem can be stated as follows:

� Given a feed consisting of a mixture of NC com-ponents, the column configuration, and product purityspecification for a key distillate component. The goal isto maximize an objective function by manipulating thecolumn reflux ratio RR(t) and vapor flowrate V to purifya given mixture until an NC pure component is obtained(inside some pre-specified tolerance). �The specific dynamic optimization problem can then be

described as:

Given: Column configuration, feed mixture,

vapor flow rate, product purity and

batch time.

Determine: Optimal reflux ratio.

To maximize: The amount of distillate product.

Subject to: Equality and inequality constraints.

The reflux ratio RR(t) is considered the control variable in the optimization problem. A general

profit function P that combines the minimum time and the maximum distillate problem is given

by Mujtaba (2004) and is used in this study:

P=C1D−C2MB0

tB+ tS−C3 (1)

where P is the profit ($/hr), D is the amount of distillate product (kmol), C1 is the sales value ofthe distillate product ($/kmol), MB0 is the initial raw material charge (kmol), C2 is the cost of rawmaterial ($/kmol),C3 is the fixed operating cost (energy, wages, depreciation, etc., $/hr), tB is thebatch time (hrs) and tS is the set up time (charging and cleaning time between batches, hrs). Inmathematical terms, the optimization problem can be represented as:

maxRR(t) D

s.t. Dynamic process equations (equality constraints)

xproduct(t)≥xdesired (inequality constraint)

Bounds on QREB and Dproduct (inequality constraint)

3. Dynamic Process Equations

In order to solve the optimization problem shown in the previous section, it is necessary to develop

a rigorous model to successfully predict the behavior of the variables with respect to time. Two

basic assumptions are applied in the formulation of the model:

1. The vapor phase holdup is assumed to be negligible compared to the liquid phase holdup oneach plate.

2. Chemical reactions in the vapor phase are neglected.

654

Rigorous Modeling, Simulation and Optimization of a Batch Reactive DistillationColumn: a comparative study

The proposed set of differential and algebraic equations (DAEs), can be decomposed into differentequations: mass balances, energy balances, equilibrium equations (chemical, physical and thermo-dynamic) and other equations such as reaction rate, summation of compositions, etc. The set ofequations that constitutes the proposed model is presented in the set of Equations 2-19, which arederived from the distillation column on Figure 1. The heat of reaction in the energy balance equa-tions is omitted because heat of formation at the standard conditions is used as a base for enthalpycalculations. The notation for the variables is given in Figure 2. The stages are numbered frombottoms to top of the batch column, stage 1 being the reboiler and condenser the stage 10. Moredetails about the column are given in Section 5.

Total mass balancesReboiler: j = 1n

dMBdt

=−D+Δn1MB (2)

Distribuitors: j = NT −1 and Tray 2dMj

dt= R1j+1+R

2j −Rj+Vj+1−Vj+Δn jMj (3)

Internal Trays: j = 2 and NT −2dM1

j

dt= R1j+1−R1j +V 1j−1−V 1j +Δn1jM

1j (4)

Component mass balancesReboiler: j = 1

MBdxB,idt

= R2x2,i−RBxB,i−VB(xB,i− yB,i)+ rB,iMB−ΔniMB (5)

Distribuitors: j = 2and j = NT −1d(Mjx j,i)

dt= R1j+1x

1j+1,i+R

2j+1x

2j+1,i−Rjx j,i+Vj−1y j−1,i−Vjy j,i+ r jiMj (6)

Internal Trays (section 1 and 2): j = Distributor+1, ...,NT −2d(M1

j x1j,i)

dt= R1j+1x

1j+1,i−R1j x1j,i+V 1j−1,iy j−1,i−V 1j y1j,i+ r1jiM1

j (7)

d(M2j x2j,i)

dt= R2j+1x

2j+1,i−R2j x2j,i+V 2j−1,iy j−1,i−V 2j y2j,i+ r2jiM2

j (8)

Condenser: j = NT

d(MNT xNT,i)dt

=VNT−1yNT−1,i− (Vj+ΔnNTMNT )xNT,i+ rNT,iMNT (9)

Energy balanceReboiler: j = 1

0= QREB+R2hL2 −RBhLB+VB(hLB−hVB) (10)

Distribuitors: j = NT −1 and Tray 2R1j+1h

L1j+1−R2j+1hL2j+1−RjhLj +Vj−1hVj−1−VjhVj = 0 (11)

Internal Trays: j = Distributor+1, ...,NT −2R1j+1h

L1j+1−R1j hL1j +V 1j−1h

V1j−1−V 1j hV1j = 0 (12)

Condenser: j = NT

QCOND−VNT−1hVNT−1+(VNT−1+ΔnNTMNT )hLNT = 0 (13)

655

Edna Soraya Lopez-Saucedo et al.

Equilibrium relationship and vapor and liquid summations

y j,i = Kj,ix j,i and y1,2j,i = Kj,ix1,2j,i , Σy j,i = 1 and x j,i = 1 (14)

Vapor-Liquid Equilibrium constant

Kj,i = Kj,i(x j,i,Tj,Pj) and K1,2j,i = K1,2j,i (x1,2j,i ,T

1,2j ,Pj) (15)

Enthalpy

hLj,i = hLj,i(x j,i,Tj,Pj) and hL1,2j,i = hL1,2j,i (x1,2j,i ,T

1,2j ,Pj) (16)

hVj,i = hVj,i(y j,i,Tj,Pj) and hV1,2j,i = hV1,2j,i (y1,2j,i ,T

1,2j ,Pj) (17)

Reaction and Reaction Terms

AcOH + EtOH ↔ EtOAc + H2O(acetic acid) (ethanol) (ethyl acetate) (water)

(18)

Δn j = Σr j,i when r j,i = r j,i(k j,i,x j,i) (19)

j and i are the number of trays and components, respec-tively

x j,i is the liquid mole fraction for tray j and component iy j,i is the vapor mole fraction for tray j and component iD is the distillate flowrate

Rj is the liquid flowrate in tray jVj is the vapor flowrate in tray jMB and MNT are the reboiler and distillate holdup, respec-tively

QREB is the energy consumed in the reboilerQCOND is the energy consumed in the condenserhLj is the liquid enthalpy in tray jhVj is the vapor enthalpy in tray jsuperscript 1 and 2 represent left side (section 1) and rightside (section 2) of the dividing wall batch column, respec-tively

Figure 2: Notation used on the dynamic model

4. Solution approachesIn order to determine the optimal solution of the dynamic model presented in the previous section,

we summarize below some of the issues that arise in the two approaches used in this study.

4.1. Solving the optimization problem by an Equation Oriented Approach

In this approach the set of DAEs (Equations 2 - 19) is discretized into a set of algebraic equations

by applying finite elements and the orthogonal collocation method developed by Cuthrell and

Biegler (1987). These equations are then used in a large-scale NLP model. The use of finite

elements and collocation points provides more flexibility but the error in the discretization cannot

be easily controlled. The proposed DAE system involves a complex set of equations that leads

to an index 2 problem. To be solved, the index should be reduced to 1 by differentiating the

equation ∑NCi=1 x j,i = 1. This leads to a new algebraic equation that substitutes the internal trays

mass balances given in Equation 7 for one component.

4.2. Solving the optimization problem by Control Vector Parameterization Approach

To formulate the optimal control problem as a reduced NLP problem, the control variable RR(t)

is approximated by a finite dimensionally equation. The time interval is divided into a finite

number of subintervals, each involving a finite number of parameters for the control variables to

be optimized. This new problem is subjected to the constraints of the model and can be solved

using a Succesive Quadratic Programming (SQP) algorithm. This approach has the advantage

of providing a direct control of the discretization error by adjusting the size and order of the

integration steps using integration techniques.

656

Rigorous Modeling, Simulation and Optimization of a Batch Reactive DistillationColumn: a comparative study

5. Case study

Table 1: Operational conditions for the

batch reactive column: acetic acid esteri-

ficationNumber of Trays

(including reboiler and condenser) 10

Feed, MB0 , kmol 100

Vapor, V, kmol/h Variable

Vapor and Liquid Distributors β1 = 0.70F1 = 0.70

Composition Feed, x0i , mole fractionAcetic Acid 0.45

Ethanol 0.45

Ethyl Acetate 0

Water 0.10

Activity coefficients (NRTL), αiAcetic Acid 0.98

Ethanol 0.99

Ethyl Acetate 2.3

Water 2.5

Column Holdup, Mj , kmol

Condenser 0.283

Internal trays 0.071

Column Pressure, P, bar

Condenser 1.05

Internal Trays 1.12-1.08

Reboiler 1.2

Reaction equations given in Mujtaba (2004)

r = k1C1C2 - k2C3C4k1 = 4.76x10−4k1 = 1.63x10−4

The reaction-separation is carried out using a 10 tray

Dividing Wall Batch Reactive Column (DWBRC).

The operational conditions are given in Table 1. An

amount of 100 kmol is fed into the reboiler at the

start of the operation with the next composition in

mole fraction: 0.45 acetic acid, 0.45 ethanol and

0.10 water. The distillate product must achieve a pu-rity of at least 0.70 in mole fraction of ethyl acetate

in the distillate with a fixed batch time of 1 hour. The

performance is modeled assuming a constant molar

holdup for each tray (a total column holdup of 4% of

the initial feed where 50% is taken as the condenser

hold up and the rest is equally divided in the plates).

Constant relative volatilities, calculated using NRTL

are used for the various components for modeling the

phase equilibrium. The initial values for the system

variables are calculated at total reflux in the steady

state. The reaction zone extends from tray 1 to 9.

The production of ethyl acetate by the esterification

of acetic acid with ethanol is accomplished by the

stoichiometric equation 18.

6. ResultsThe optimization problem is solved by discretizing

the differential equations using the two approaches

presented in section 4 with the next specifications:

the CVP approach implemented in gPROMS and the

equation oriented approach with 10 finite elements

and 3 collocation points implemented in GAMS

(24.2.2) using CONOPT as the NLP solver. The ob-

jective is to maximize the productivity by converting

the maximum profit problem in a maximum produc-

tivity problem when C1 = 1 and C2 = C3 = tS = 0 in Equation 1. The reflux ratio is used as the

control variable. Five different cases with different vapor flowrates were studied. All examples

were solved on a Workstation with 8 GB RAM memory and Intel�CoreTM i7 CPU (2.20 GHz).

The CVP approach results in a system of 530 equations and 600 variables. The optimization

problem is solved in 40 seconds. From the results presented in Table 2(a). The fixed batch time is

1.0 hour, while the optimal reflux ratio is piecewise as shown in Figure 3(c) for all vapor flowrates.It is clear that the maximum amount of distillate product Dethylacetate (kmol) is achieved for themaximum vapor flowrate V = 90 kmol/hr. The equation oriented approach results in a system of

532 equations and 698 variables. The optimization problem is solved in 318 seconds. The results

are presented in Table 2(b). As in the CVP approach, the batch time is 1.0 hour. The maximumamount of distillate productDethylacetate (kmol) is achieved for the maximum vapor flowrateV = 80

kmol/hr. In both approaches, the amount of distillate and the energy consumption were directlyproportional to increases in vapor flowrate. Therefore, any higher purity would require higher

energy consumption. A comparison between the calculated reboiler energy shows that the EOA

needs a better initial point for the temperatures in the trays. The accumulated distillate composition

profiles for the maximum distillate product for the two approaches are shown in Figures 3(a) and

3(b).

657

Edna Soraya Lopez-Saucedo et al.

Table 2: Results for the productivity maximization

(a) Results for the CVP approach

Vapor QREB DEtOAc(kmol/hr) (kJ) (kmol)

50 47.40 10.0060 56.88 12.3270 66.48 14.1680 76.13 15.8390 86.13 16.85

(b) Results for the EOA approach

Vapor QREB DEtOAc(kmol/hr) (kJ) (kmol)

50 78.43 10.3260 94.14 12.0970 109.78 13.2780 125.18 12.33

0.0 0.2 0.4 0.6 0.8 1.00.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Dis

tilla

te c

ompo

sitio

n pr

ofile

s in

mol

e fra

ctio

n

Time (hrs)

acetic acid ethanol ethyl acetate water

(a) CVP approach

0.0 0.5 1.00.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Acetic Acid Ethanol Ethyl acetate Water

Com

posi

tion

prof

iles

in th

e di

still

ate

(mol

e fra

ctio

n)

Time (hrs)

(b) EOA

0.0 0.2 0.4 0.6 0.8 1.0

0.5

0.6

0.7

0.8

0.9

1.0

Ref

lux

ratio

(R/V

)

Time (hrs)

Vaporflowrate(kmol/hr)

50 60 70 80 90

(c) Optimal reflux ratio for the CVP

approach

Figure 3: Distillate composition profiles for the production of ethyl acetate in a Dividing Wall

Batch Reactive Distillation Column for the CVP and EOA approaches and optimal piecewise

reflux ratio for the production of ethyl acetate in a Dividing Wall Batch Reactive Distillation

Column for the CVP approach

7. Conclusion

In this work, a model for the optimization of a dividing wall batch reactive distillation column

when the esterification of ethanol using acetic acid to produce ethyl acetate is studied. A maximum

productivity optimization problem is solved under fixed distillate product purity (ethyl acetate

concentration higher than 0.70 in mole fraction). The results show that the problem is solved

using the two different approaches: the finite elements with collocation points implemented in

GAMS (24.2.2), and the control vector parametization implemented in gPROMS (2004) with no

major differences on the calculated variables values. These differences are due to the discretization

error carried out during the discretization. Also, the results show that significantly more CPU time

is required with the EOA approach.

ReferencesBiegler, L. T., 1984. Solution of dynamic optimization problems by succesive quadratic programming and orthogonal

collocation. Comput. Chem. Eng. 8, 243–248.

Cuthrell, J. E., Biegler, L. T., 1987. On the optimization of differential-algebraic process systems. AIChE J 33, 257.

Mujtaba, I. M., 2004. Batch Distillation Design and Operation. Series on Chemical Engineering Vol. 3, University of

Bradford, UK.

Vassiliadis, V. S., Sargent, R. W. H., Pantelides, C. C., 1994. Solution of a class of multistage dynamic optimization

problems. part i-algorithmic framework. Ind. Eng. Chem. Res. 33, 2115–2123.

658